**Meet the editor**

Dr. Dushanova's research interests are in motor neurophysiology, pathophysiology of Parkinson's disease and in the development of approaches for diagnostics. Her works span human and non-human primate research, computational modeling and simulations. She received her MS degree from Sofia University, Bulgaria, Predoctoral fellow by Prof. Pfurtscheller at Department of

Medical Informatics, Ludwig Boltzmann Institute for Medical Informatics and Neuroinformatics, Technical University Graz, Austria, PhD from Institute of Neurobiology, Bulgarian Academy of Sciences and studied neurophysiology under Prof. J.P. Donoghue in the Neuroscience Department of Brown University, RI USA. Assoc. Prof. J. Dushanova's research is in the field of movement disorders and she has been involved in electroencephalography research and practice since 1994.

Contents

**Preface IX** 

**Part 1 Biomarkers for Preclinical Diagnosis of PD 1** 

Rosella Ciurleo and Placido Bramanti

**by Electrophysiological Methods 27** 

Chapter 3 **Brain Event - Related Oscillations in Parkinsonian** 

Chapter 4 **Extraction of Single-Trial Post-Movement MEG** 

Chapter 5 **Developing an MRI-Based Biomarker for** 

Chapter 7 **Minor and Trace Elements in Cerebrospinal** 

Chapter 8 **Language Processing in Parkinson's** 

Po-Lei Lee, Yu-Te Wu and Jen-Chuen Hsieh

**Early Diagnosis of Parkinson's Disease 115**  Jorge E. Quintero, Xiaomin Wang and Zhiming Zhang

Chapter 6 **Neuroimaging in Manganese-Induced Parkinsonism 131** 

**Fluid of Parkinson's Patients – Suggestions** 

**Disease Patients Without Dementia 165**  Katrien Colman and Roelien Bastiaanse

Margherita Speziali and Michela Di Casa

**After a Critical Review of the Analytical Data 149** 

Chapter 2 **Diagnosis of Parkinson's Disease** 

Juliana Dushanova

Yangho Kim

Chapter 1 **Early Marker for the Diagnosis of Parkinson's Disease 3**  Silvia Marino, Pietro Lanzafame, Silvia Guerrera,

Elena Lukhanina, Irina Karaban and Natalia Berezetskaya

**Patients During Discrimination Task Conditions 59** 

**Beta Synchronization in Normal and Parkinson's Patient Using ICA-Based Spatiotemporal Approach 85** 

## Contents

#### **Preface XIII**


Contents VII

Chapter 18 **Potentials of Telehealth Devices**

**Part 4 Invasive Methods Examine** 

Sun Ha Paek

Chapter 23 **Electrical Stimulation**

Lilian Beijer and Toni Rietveld

Chapter 19 **Rehabilitation Versus no Intervention –**

Chapter 20 **Improving Transfer of Parkinson's Disease** 

Chapter 21 **Human Central Nervous System Circuits** 

Naoki Tani and Youichi Saitoh

**Only a Continued Intensive Program** 

Jesús Seco Calvo and Inés Gago Fernández

**for Speech Therapy in Parkinson's Disease 379**

**Conducted Statistically Significant Improvements Motor Skills in Parkinson's Disease Patients 403** 

**Patients – Sit-to-Stand Motion Assistance 419** Yoshiyuki Takahashi, Osamu Nitta and Takashi Komeda

**Examined in Patients with Parkinson's Disease Using the Electrodes Implanted for Deep Brain Stimulation 435** 

**STN DBS in Patients with Advanced Parkinson's Disease 469**

**of Primary Motor Cortex for Parkinson's Syndrome 493**

Adrian J. Cassar Gheiti, Joseph F. Baker and Kevin J. Mulhall

Joao Costa, Francesc Valldeoriola and Josep Valls-Sole

**in Patients with Parkinson's Disease 433**

Chapter 22 **Estimation of Electrode Position with Fused Images of Preoperative MRI and Postoperative CT Using the Mutual Information Technique After** 

Chapter 24 **Joint Replacement Surgery in Parkinson's Disease 509**

Chapter 9 **Novel Methods to Evaluate Symptoms in Parkinson's Disease – Rigidity and Finger Tappin 191**  Takuyuki Endo, Masaru Yokoe, Harutoshi Fujimura and Saburo Sakoda

#### Chapter 10 **Relevance of Aerodynamic Evaluation in Parkinsonian Dysarthria 207**  Sarr Mamadou Moustapha, Ghio Alain, Espesser Robert, Teston Bernard, Dramé Moustapha and Viallet François

	- **Part 3 Multidisciplinary Cognitive Rehabilitation 255**

Chapter 18 **Potentials of Telehealth Devices for Speech Therapy in Parkinson's Disease 379**  Lilian Beijer and Toni Rietveld

VI Contents

**Part 2 Novel Methods** 

Chapter 10 **Relevance of Aerodynamic**

Chapter 9 **Novel Methods to Evaluate Symptoms**

Takuyuki Endo, Masaru Yokoe, Harutoshi Fujimura and Saburo Sakoda

**to Evaluate the Symptoms in Parkinson's Disease 189**

**in Parkinson's Disease – Rigidity and Finger Tappin 191** 

**Evaluation in Parkinsonian Dysarthria 207**

Chapter 11 **Objective Evaluation of the Severity of Parkinsonism Using** 

Weidong Pan, Yoshiharu Yamamoto and Shin Kwak

Ryoichi Hayashi, Takeshi Hayashi, Junpei Aizawa,

Chapter 12 **Postural Control While Sitting and Its Association**

**Part 3 Multidisciplinary Cognitive Rehabilitation 255** 

**Transfer Training and Sports Therapy 257**  I. Reuter, S. Mehnert, M. Oechsner and M. Engelhardt

**for Neurological Diseases Management – The Case of Parkinson's Disease 287** Laura Pastor-Sanz, Mario Pansera, Jorge Cancela,

Chapter 15 **An Investigation into the Impact of Parkinson's Disease** 

Jessica Davies, Hoe Lee and Torbjorn Falkmer

Chapter 16 **Effects of a Multimodal Exercise Program on Clinical, Functional Mobility and Cognitive Parameters of Idiopathic Parkinson's Disease Patients 339** L.T.B. Gobbi, F.A. Barbieri, R. Vitório, M.P. Pereira and C. Teixeira-Arroyo on behalf of the PROPARKI Group

Chapter 17 **Rehabilitation of Patients Suffering from**

Gilles Orgeret

Matteo Pastorino and María Teresa Arredondo Waldmeyer

**Parkinson's Disease by Normotensive Therapy 353**

**upon Decision Making Ability and Driving Performance 309** 

Chapter 13 **Cognitive Rehabilitation in Parkinson's Disease Using Neuropsychological Training,**

Hiroaki Nagase and Shinji Ohara

Chapter 14 **Mobile Systems as a Challenge** 

Sarr Mamadou Moustapha, Ghio Alain, Espesser Robert, Teston Bernard, Dramé Moustapha and Viallet François

**Power-Law Temporal Auto-Correlation of Activity 225**

**with Risk of Falls in Patients with Parkinson's Disease 239**

	- **Part 4 Invasive Methods Examine in Patients with Parkinson's Disease 433**

Preface

We have based this book on the philosophy that one has control on how Parkinson's disease affects the life of a person if the disease is battled with the belief that through support and maximization of personal strengths, life will continue to have meaning and richness. The battle begins with the partnership between the physicians and the patient. By understanding as much as possible about Parkinson's disease, medications, and non-drug therapies, a patient can actively participate in the health-care decisions that have an unique effect on the self. If a person has Parkinson's disease (PD), it does not necessarily mean that one will experience all possible symptoms. We have

organized this book in sections, each one dealing with a different aspect of PD.

who take health care of people with PD and their families.

degree of the symptoms varies depending on the neurologist.

The idea was to edit a book that will reflect a professional wisdom, extended to the experiences of scientists from different fields. We hope that a positive attitude of the colleagues who have contributed to the expertise of this book will serve as a powerful instrument against this disease and that this handbook will also be useful for clinicians

Parkinson's disease is a progressive neurologic disorder affecting 1 in 100 people over the age of 50. Typically the diagnosis is made in the sixth or seventh decade of life, with approximately 7% of people diagnosed before the age of 40. It has been well documented that PD is also characterized by a long preclinical phase from the onset of dopamine neuron loss to the onset of motor symptoms. In the recent years more light was thrown on non-motor symptoms such as autonomic dysfunction, pain and cognitive decline. Approximately 10% of subjects older than 60 years are in the prediagnostic phase of PD and exhibit the pathological hallmarks of PD, like Lewy bodies and neuronal loss at the substantia nigra (SN), without showing the motor signs during life time that allow the diagnosis of PD. There is enough evidence to show that cognitive deficits affect the quality of life. Thus, there is an urgent need to develop imaging modality to screen individuals who may be in the preclinical phase of PD for earlier diagnosis and treatment to slow down or even stop progression of the disease. Parkinsonian symptoms such as tremor, rigidity, akinesia, postural instability and facial expression are perceived subjectively, and therefore understanding the

The neuroimaging examination could help to evaluate PD subjects in the preclinical stage. Conventional MR imaging, as well as different advanced MRI techniques,

## Preface

We have based this book on the philosophy that one has control on how Parkinson's disease affects the life of a person if the disease is battled with the belief that through support and maximization of personal strengths, life will continue to have meaning and richness. The battle begins with the partnership between the physicians and the patient. By understanding as much as possible about Parkinson's disease, medications, and non-drug therapies, a patient can actively participate in the health-care decisions that have an unique effect on the self. If a person has Parkinson's disease (PD), it does not necessarily mean that one will experience all possible symptoms. We have organized this book in sections, each one dealing with a different aspect of PD.

The idea was to edit a book that will reflect a professional wisdom, extended to the experiences of scientists from different fields. We hope that a positive attitude of the colleagues who have contributed to the expertise of this book will serve as a powerful instrument against this disease and that this handbook will also be useful for clinicians who take health care of people with PD and their families.

Parkinson's disease is a progressive neurologic disorder affecting 1 in 100 people over the age of 50. Typically the diagnosis is made in the sixth or seventh decade of life, with approximately 7% of people diagnosed before the age of 40. It has been well documented that PD is also characterized by a long preclinical phase from the onset of dopamine neuron loss to the onset of motor symptoms. In the recent years more light was thrown on non-motor symptoms such as autonomic dysfunction, pain and cognitive decline. Approximately 10% of subjects older than 60 years are in the prediagnostic phase of PD and exhibit the pathological hallmarks of PD, like Lewy bodies and neuronal loss at the substantia nigra (SN), without showing the motor signs during life time that allow the diagnosis of PD. There is enough evidence to show that cognitive deficits affect the quality of life. Thus, there is an urgent need to develop imaging modality to screen individuals who may be in the preclinical phase of PD for earlier diagnosis and treatment to slow down or even stop progression of the disease. Parkinsonian symptoms such as tremor, rigidity, akinesia, postural instability and facial expression are perceived subjectively, and therefore understanding the degree of the symptoms varies depending on the neurologist.

The neuroimaging examination could help to evaluate PD subjects in the preclinical stage. Conventional MR imaging, as well as different advanced MRI techniques, including magnetic resonance spectroscopy, diffusion-weighted and diffusion tensor imaging are helpful to distinguish PD from atypical or secondary PD. The secondary symptoms of Parkinson's disease may affect some but not all people. These include changes in speech and swallowing, bowel and bladder function, fatigue, mood and memory, sexual function, and sleep. Speech of patients with Parkinson's disease (PD) is characterized as hypokinetic dysarthria, manifesting itself by low volume, diminished voice quality, flattened prosody and deteriorated articulation. Patients with mild cognitive deficits have a higher risk for developing dementia. The interventions in early stages of cognitive decline might slow down the progress of cognitive deficits or even prevent the development of dementia. Brain event-related oscillations are one of the promising candidates explaining the neural mechanisms at central nervous system in Parkinson's patients and give a useful tool for detecting subtle abnormalities of the cognitive processes.The sustained yield of a multidisciplinary training programme on cognitive function of PD patients with mild cognitive deficits is to support patients to manage their daily life better and to become more self-confident. When symptoms first appear, they are very mild and sometimes intermittent. As the disease progresses the symptoms become more pronounced and more persistent. The primary symptoms of Parkinson's disease are tremor, rigidity, bradykinesia (slowness), and impaired balance but not all symptoms need to be present to make the diagnosis. Tremor is most prominent when a person is sitting quietly, and improves or disappears when the person is using their arms or legs. That is why Parkinson tremor is described as a resting tremor. Rigidity is stiffness of muscles. As the disease progresses, the stiffness is perceived as a cramp or tired, aching muscles. This rigidity is associated with the flexion posture characteristic of Parkinson's disease. Rigidity responds well to a combination of medications and a vigorous stretching program. Bradykinesia means slowness of movement. Fine movements, such as writing, becomes more clumsy. Another manifestation of bradykinesia is loss of associated movements. These are movements outside of one's awareness: blinking the eyes, facial expression, swallowing, swinging the arms, and changes of posture. The overall appearance is one of unusual stillness when a person is sitting quietly. Bradykinesia also affects voluntary movement. Medication can be very helpful in improving bradykinesia. Impaired balance occurs because of a change in postural reflexes. These are the reflexes that facilitate rapid changes in the center of balance when walking or standing. A person will notice a feeling of unsteadiness and in later stages of disease may have a tendency to list to one side or the other and may even fall backwards or forwards. Medication is less helpful for balance problems than for rigidity or bradykinesia. Physical therapy can be very helpful in teaching safety maneuvers, balance exercises, and providing consultation regarding ambulation aids.

Preface XI

by approximately 80% the symptoms of Parkinson's disease emerge. Over time the loss of dopamine-producing cells continues and symptoms become more severe. The current debate is whether nerve cell loss is something that has a genetic link occurring slowly over time or happens suddenly after being exposed to a toxic substance. It may be a combination of these two theories. It has been suggested that some people are genetically predisposed to developing Parkinson's disease and therefore more

Parkinson's disease progresses slowly. The rate of progression varies, making it difficult to give patients and families definitive information that can provide them the comfort of knowing what to expect and how to prepare for their future. The studies that have attempted to describe the prognosis of Parkinson's disease have been hard to interpret because drug therapy treats the symptoms. Drug treatment, specifically levodopa therapy, slows the onset of disability. Disability will eventually occur and increase while on therapy but this is usually because of the emergence of new

People should consider how Parkinson's symptoms might alter performance and think of ways to modify their work environment or their type of work. It is also important to remember that current research offers tangible possibilities that could change the course of Parkinson's disease. Gene therapy, surgical therapies, rehabilitation, and drugs that delay progression hold great hope for changing the disability of Parkinson's disease. The training tasks were allocated to different categories: concentration, strategy, improvement of orientation, planning, use of mnemonic devices. Impulse control, decision processes, listening training and memory, and also a special programme with the aim to learn motor sequences, dual tasking, orientation in a room. The motor training helped patients to deal with their cognitive problems; many patients are afraid of the cognitive deficits and rather try to hide than to approach the problems. Physical activity improves cognitive functions especially executive functions which are important for daily living. Sport is as well an activity to get the partners involved. Social aspects are important for the continuation of the training and as well important to prevent depression. Depression and social isolation are closely associated with cognitive decline. The new ICT systems are able to evaluate the variation of symptoms along whole the time the patient is wearing the sensors and provide useful information to physicians in order to allow them make decisions more accurate, more efficient and quicker. Long-term multimodal exercise programs can improve both motor and cognitive impairments in people with PD, which could have

a broader impact on quality of life than specific exercise interventions.

therapy.

Predating the diagnosis of PD and identifying subjects at risk will be one important goal for future research aimed to postpone the onset of the disease by neuroprotective

Part I of the book first gives an introduction to "prediagnostic" phase of PD or early markers for the diagnosis of Parkinson's disease as non-motor symptoms such as

susceptible to the potential damage of a toxic exposure.

symptoms that do not respond to levodopa.

Parkinsonism is the loss of dopamine-producing cells in an area of the brain stem called the substantia nigra. These nerve cells project fibers to areas deep in the brain called the basal ganglia. The function of the neurochemical, dopamine, is to allow nerve impulses to run smoothly along these fibers and transmit messages to muscles of the body producing normal movement. When the supply of dopamine is decreased by approximately 80% the symptoms of Parkinson's disease emerge. Over time the loss of dopamine-producing cells continues and symptoms become more severe. The current debate is whether nerve cell loss is something that has a genetic link occurring slowly over time or happens suddenly after being exposed to a toxic substance. It may be a combination of these two theories. It has been suggested that some people are genetically predisposed to developing Parkinson's disease and therefore more susceptible to the potential damage of a toxic exposure.

X Preface

including magnetic resonance spectroscopy, diffusion-weighted and diffusion tensor imaging are helpful to distinguish PD from atypical or secondary PD. The secondary symptoms of Parkinson's disease may affect some but not all people. These include changes in speech and swallowing, bowel and bladder function, fatigue, mood and memory, sexual function, and sleep. Speech of patients with Parkinson's disease (PD) is characterized as hypokinetic dysarthria, manifesting itself by low volume, diminished voice quality, flattened prosody and deteriorated articulation. Patients with mild cognitive deficits have a higher risk for developing dementia. The interventions in early stages of cognitive decline might slow down the progress of cognitive deficits or even prevent the development of dementia. Brain event-related oscillations are one of the promising candidates explaining the neural mechanisms at central nervous system in Parkinson's patients and give a useful tool for detecting subtle abnormalities of the cognitive processes.The sustained yield of a multidisciplinary training programme on cognitive function of PD patients with mild cognitive deficits is to support patients to manage their daily life better and to become more self-confident. When symptoms first appear, they are very mild and sometimes intermittent. As the disease progresses the symptoms become more pronounced and more persistent. The primary symptoms of Parkinson's disease are tremor, rigidity, bradykinesia (slowness), and impaired balance but not all symptoms need to be present to make the diagnosis. Tremor is most prominent when a person is sitting quietly, and improves or disappears when the person is using their arms or legs. That is why Parkinson tremor is described as a resting tremor. Rigidity is stiffness of muscles. As the disease progresses, the stiffness is perceived as a cramp or tired, aching muscles. This rigidity is associated with the flexion posture characteristic of Parkinson's disease. Rigidity responds well to a combination of medications and a vigorous stretching program. Bradykinesia means slowness of movement. Fine movements, such as writing, becomes more clumsy. Another manifestation of bradykinesia is loss of associated movements. These are movements outside of one's awareness: blinking the eyes, facial expression, swallowing, swinging the arms, and changes of posture. The overall appearance is one of unusual stillness when a person is sitting quietly. Bradykinesia also affects voluntary movement. Medication can be very helpful in improving bradykinesia. Impaired balance occurs because of a change in postural reflexes. These are the reflexes that facilitate rapid changes in the center of balance when walking or standing. A person will notice a feeling of unsteadiness and in later stages of disease may have a tendency to list to one side or the other and may even fall backwards or forwards. Medication is less helpful for balance problems than for rigidity or bradykinesia. Physical therapy can be very helpful in teaching safety maneuvers, balance exercises, and providing consultation regarding ambulation aids.

Parkinsonism is the loss of dopamine-producing cells in an area of the brain stem called the substantia nigra. These nerve cells project fibers to areas deep in the brain called the basal ganglia. The function of the neurochemical, dopamine, is to allow nerve impulses to run smoothly along these fibers and transmit messages to muscles of the body producing normal movement. When the supply of dopamine is decreased Parkinson's disease progresses slowly. The rate of progression varies, making it difficult to give patients and families definitive information that can provide them the comfort of knowing what to expect and how to prepare for their future. The studies that have attempted to describe the prognosis of Parkinson's disease have been hard to interpret because drug therapy treats the symptoms. Drug treatment, specifically levodopa therapy, slows the onset of disability. Disability will eventually occur and increase while on therapy but this is usually because of the emergence of new symptoms that do not respond to levodopa.

People should consider how Parkinson's symptoms might alter performance and think of ways to modify their work environment or their type of work. It is also important to remember that current research offers tangible possibilities that could change the course of Parkinson's disease. Gene therapy, surgical therapies, rehabilitation, and drugs that delay progression hold great hope for changing the disability of Parkinson's disease. The training tasks were allocated to different categories: concentration, strategy, improvement of orientation, planning, use of mnemonic devices. Impulse control, decision processes, listening training and memory, and also a special programme with the aim to learn motor sequences, dual tasking, orientation in a room. The motor training helped patients to deal with their cognitive problems; many patients are afraid of the cognitive deficits and rather try to hide than to approach the problems. Physical activity improves cognitive functions especially executive functions which are important for daily living. Sport is as well an activity to get the partners involved. Social aspects are important for the continuation of the training and as well important to prevent depression. Depression and social isolation are closely associated with cognitive decline. The new ICT systems are able to evaluate the variation of symptoms along whole the time the patient is wearing the sensors and provide useful information to physicians in order to allow them make decisions more accurate, more efficient and quicker. Long-term multimodal exercise programs can improve both motor and cognitive impairments in people with PD, which could have a broader impact on quality of life than specific exercise interventions.

Predating the diagnosis of PD and identifying subjects at risk will be one important goal for future research aimed to postpone the onset of the disease by neuroprotective therapy.

Part I of the book first gives an introduction to "prediagnostic" phase of PD or early markers for the diagnosis of Parkinson's disease as non-motor symptoms such as mood disorders, olfactory, vegetative, sensory or neuropsychological signs may be noticed by the patients or physicians in advance of motor signs reflecting the dysfunction of dopaminergic or non-dopaminergic neurons. Several procedures in *chapter 1* have been proposed to identify subjects in early stages of PD as the saccadic eye movements to investigate and quantify motor impairments in PD, and a new vision-based nonintrusive eye tracker as a possible tool for supporting the diagnosis of PD in association with levodopa test. A more complex relationship between brain iron changes and disease state in PD has revealed different metabolic patterns and a reduced capacity of the macromolecules in brain tissue to exchange magnetization with the surrounding water molecules was found in the substantia nigra pars compacta, substantia nigra pars reticulate, red nucleus in PD. Diffusion Weighted Imaging and a statistical parametric mapping localized significant increases of diffusivity in the region of both olfactory tracts in patients with PD compared to healthy controls. This observation is in line with the well-established clinical finding of hyposmia in early PD. Olfactory dysfunction may be considered a reliable marker of PD. Objective olfaction tests are olfactory-evoked potentials or functional magnetic resonance imaging (fMRI). These techniques have been used to assess the severity of olfactory dysfunction and its correlation with cerebral changes in studies carried out in early PD patients. Effective differential diagnosis of Parkinson's disease needs in informative indexes that objectively reflect the functional state of the extrapyramidal system. Such informative diagnostic indexes in PD are surface electromyograms (EMGs) and brain evoked potentials (*Chap. 2*). The fractal analysis of EMG is sensitive to neuromuscular status. The contingent negative variation is a sensitive indicator for the objective evaluation of the severity of PD and to quantify the efficacy of the therapy. Deviations of the auditorily elicited brain oscillatory responses at specific frequencies in association with sensorimotor and cognitive processes for the Parkinson's patients compared to healthy controls are a evidence for disturbances in the temporal and regional integration of these frequency components and the relationships between cortical and the basal ganglia circuits in parkinsonism (*Chap. 3*). A method based on independent component analysis (ICA) and the use of a template-based correlation approach to extract Rolandic beta rhythm from magnetoencephalographic measurements of right finger lifting are presented in *chapter 4.* The objectives of the *chapter 5* are the use of neuroimaging such as magnetic resonance imaging, positron emission tomography or single-photon emission computed tomography in manganese-induced Parkinsonism and the assessment of the neural correlates of manganese-induced memory impairment in response to a subclinical dysfunction of working memory network in welders with chronic manganese exposure. *Chapter 6* has been trying to introduce in developing of MRI-based imaging biomarkers for early detection of PD. The imaging method as pharmacological MRI can detect functional deficiency of the nigrostriatal system. The purpose of the *chapter 7* is to give a systematic and exhaustive depiction of element levels (Cu, Fe, Zn, Cr, Be, Cd, Cr, Hg, Se, Si, V, Mn, Cu, Co, Pb , Ni) in the cerebrospinal fluid of PD patients and paired controls and to verify the influence on the results of number, age, gender of the subjects and health conditions with regard to clinical variables as duration and Preface XIII

severity of the disease and pharmacological therapies. The *chapter 8* considers the underlying mechanisms of the subtle language impairments in non-demented PD patients. Recently functional imaging in PD patients has begun to add information to the underlying nature of the language impairments in PD, both comprehension and

Part II concentrates on the novel methods to evaluate symptoms in Parkinson's disease. This contribution in the book deals with the sensing systems identifying rigidity or spasticity and the nature of abnormal finger tapping in Parkinson' s disease, and show Parkinsonian symptoms as a system error in software of repetitive movement (*Chap. 9*). Among Parkinsonian signs, speech impairment represent an important disabling symptom able to lead towards a significant reduction of oral communication. The principal methods for PD speech evaluation will be reviewed briefly in the *chapter 10* prior to the presentation of the use and relevance of aerodynamic parameters for Parkinsonin dysarthria evaluation. The analytical method based on power-law temporal auto-correlation of physical activity collected by an actigraph device enables evaluation of the severity of Parkinsonism with sufficient sensitivity and reliability, and is useful for the evaluation of efficacy of therapy for Parkinsonism (*Chap. 11*). Postural control while sitting with or without arm raising and its association with risk

Part III focuses on multidisciplinary cognitive rehabilitation in Parkinson`s disease. A study on cognitive training in Parkinson`s disease and compared different types of cognitive training is discussed in *chapter 13*. The studies assess the effect and sustained yield of a multidisciplinary training programme on cognitive function of PD patients with mild cognitive deficits. *Chapter 14* proposes modules known as systems for support the decision-making which allow doctors and clinicians be more agile in the decision-making process improving the time that they need and the quality of information that they have. The researches explore the impact of impaired decision making ability upon the driving performance of people with PD (*Chap. 15*). The effectiveness of a long-term multimodal exercise program based on the improvement of the functional capacity components in improving clinical parameters, functional mobility and cognitive function in people with PD was demonstrated in *chapter 16*. The research areas with respect to the development, implementation and evaluation of telehealth applications for speech therapy in dysarthric speakers with Parkinson's disease are addressed in *chapter 17*. The relevant and the long-term efficiency of a physiotherapy program in PD are demonstrated in *chapter 18*. Normotensive therapy is a specific rehabilitation to control of postures and it uses of a supporting elastic expander while exercising and a traumatic normalization of the tissues, and allow mobility to be restored without any pain (*Chap. 19*). A sit-to-stand assistance system is developed, and used to assist individuals with Parkinson's disease who could or could

Part IV deals with practical issues that are critical for the invasive methods examined in patients with Parkinson's disease. A review of the use of deep brain stimulation

production deficits at the word, sentence and discourse level.

of falls in patients with Parkinson disease was discussed in *chapter 12*.

not stand up without help (*Chap. 20*).

severity of the disease and pharmacological therapies. The *chapter 8* considers the underlying mechanisms of the subtle language impairments in non-demented PD patients. Recently functional imaging in PD patients has begun to add information to the underlying nature of the language impairments in PD, both comprehension and production deficits at the word, sentence and discourse level.

XII Preface

mood disorders, olfactory, vegetative, sensory or neuropsychological signs may be noticed by the patients or physicians in advance of motor signs reflecting the dysfunction of dopaminergic or non-dopaminergic neurons. Several procedures in *chapter 1* have been proposed to identify subjects in early stages of PD as the saccadic eye movements to investigate and quantify motor impairments in PD, and a new vision-based nonintrusive eye tracker as a possible tool for supporting the diagnosis of PD in association with levodopa test. A more complex relationship between brain iron changes and disease state in PD has revealed different metabolic patterns and a reduced capacity of the macromolecules in brain tissue to exchange magnetization with the surrounding water molecules was found in the substantia nigra pars compacta, substantia nigra pars reticulate, red nucleus in PD. Diffusion Weighted Imaging and a statistical parametric mapping localized significant increases of diffusivity in the region of both olfactory tracts in patients with PD compared to healthy controls. This observation is in line with the well-established clinical finding of hyposmia in early PD. Olfactory dysfunction may be considered a reliable marker of PD. Objective olfaction tests are olfactory-evoked potentials or functional magnetic resonance imaging (fMRI). These techniques have been used to assess the severity of olfactory dysfunction and its correlation with cerebral changes in studies carried out in early PD patients. Effective differential diagnosis of Parkinson's disease needs in informative indexes that objectively reflect the functional state of the extrapyramidal system. Such informative diagnostic indexes in PD are surface electromyograms (EMGs) and brain evoked potentials (*Chap. 2*). The fractal analysis of EMG is sensitive to neuromuscular status. The contingent negative variation is a sensitive indicator for the objective evaluation of the severity of PD and to quantify the efficacy of the therapy. Deviations of the auditorily elicited brain oscillatory responses at specific frequencies in association with sensorimotor and cognitive processes for the Parkinson's patients compared to healthy controls are a evidence for disturbances in the temporal and regional integration of these frequency components and the relationships between cortical and the basal ganglia circuits in parkinsonism (*Chap. 3*). A method based on independent component analysis (ICA) and the use of a template-based correlation approach to extract Rolandic beta rhythm from magnetoencephalographic measurements of right finger lifting are presented in *chapter 4.* The objectives of the *chapter 5* are the use of neuroimaging such as magnetic resonance imaging, positron emission tomography or single-photon emission computed tomography in manganese-induced Parkinsonism and the assessment of the neural correlates of manganese-induced memory impairment in response to a subclinical dysfunction of working memory network in welders with chronic manganese exposure. *Chapter 6* has been trying to introduce in developing of MRI-based imaging biomarkers for early detection of PD. The imaging method as pharmacological MRI can detect functional deficiency of the nigrostriatal system. The purpose of the *chapter 7* is to give a systematic and exhaustive depiction of element levels (Cu, Fe, Zn, Cr, Be, Cd, Cr, Hg, Se, Si, V, Mn, Cu, Co, Pb , Ni) in the cerebrospinal fluid of PD patients and paired controls and to verify the influence on the results of number, age, gender of the subjects and health conditions with regard to clinical variables as duration and

Part II concentrates on the novel methods to evaluate symptoms in Parkinson's disease. This contribution in the book deals with the sensing systems identifying rigidity or spasticity and the nature of abnormal finger tapping in Parkinson' s disease, and show Parkinsonian symptoms as a system error in software of repetitive movement (*Chap. 9*). Among Parkinsonian signs, speech impairment represent an important disabling symptom able to lead towards a significant reduction of oral communication. The principal methods for PD speech evaluation will be reviewed briefly in the *chapter 10* prior to the presentation of the use and relevance of aerodynamic parameters for Parkinsonin dysarthria evaluation. The analytical method based on power-law temporal auto-correlation of physical activity collected by an actigraph device enables evaluation of the severity of Parkinsonism with sufficient sensitivity and reliability, and is useful for the evaluation of efficacy of therapy for Parkinsonism (*Chap. 11*). Postural control while sitting with or without arm raising and its association with risk of falls in patients with Parkinson disease was discussed in *chapter 12*.

Part III focuses on multidisciplinary cognitive rehabilitation in Parkinson`s disease. A study on cognitive training in Parkinson`s disease and compared different types of cognitive training is discussed in *chapter 13*. The studies assess the effect and sustained yield of a multidisciplinary training programme on cognitive function of PD patients with mild cognitive deficits. *Chapter 14* proposes modules known as systems for support the decision-making which allow doctors and clinicians be more agile in the decision-making process improving the time that they need and the quality of information that they have. The researches explore the impact of impaired decision making ability upon the driving performance of people with PD (*Chap. 15*). The effectiveness of a long-term multimodal exercise program based on the improvement of the functional capacity components in improving clinical parameters, functional mobility and cognitive function in people with PD was demonstrated in *chapter 16*. The research areas with respect to the development, implementation and evaluation of telehealth applications for speech therapy in dysarthric speakers with Parkinson's disease are addressed in *chapter 17*. The relevant and the long-term efficiency of a physiotherapy program in PD are demonstrated in *chapter 18*. Normotensive therapy is a specific rehabilitation to control of postures and it uses of a supporting elastic expander while exercising and a traumatic normalization of the tissues, and allow mobility to be restored without any pain (*Chap. 19*). A sit-to-stand assistance system is developed, and used to assist individuals with Parkinson's disease who could or could not stand up without help (*Chap. 20*).

Part IV deals with practical issues that are critical for the invasive methods examined in patients with Parkinson's disease. A review of the use of deep brain stimulation (DPS) electrodes for externally controlled recording or stimulation at the level of basal ganglia nuclei discussed the physiological observations underlying the engagement of structures affected by the electrical field around the electrode in motor and sensory functions (*Chap. 21*). Based on the information from the fused images of preoperative MRI-postoperative computer tomography, it was emphasized that the documentation of the electrode position by using mutual information technique after unilateral or bilateral STN stimulation provides useful information for the prediction of the surgical outcome (*Chap. 22*). Fusion-image-based programming and reprogramming of the stimulator parameters using the visual information of the location of the electrode contacts propose to find the best sites and the best stimulation parameters for the advanced PD patients treated with deep brain stimulation of subthalamic nucleus (STN-DBS). The electrical motor cortex stimulation can improve axial symptom of PD, which is difficult even with STN-DBS, and is less surgical invasiveness than deep brain stimulation can make the surgery safer for the patients with advanced age or sever brain atrophy *(Chap. 23*). The last contribution in the book deals with the rational arthroplasty surgery and multidisciplinary approaches for the patients with Parkinson's disease (*Chap. 24*).

We would like to thank all the people who supported the preparation of this book, who contributed to the book, and in particular also to all who made the book possible due to their positive evaluations of the proposals for this book.

> **Dr. Juliana Dushanova**  Institute of Neurobiology, Bulgarian Academy of Sciences, Sofia, Bulgaria

XIV Preface

Parkinson's disease (*Chap. 24*).

(DPS) electrodes for externally controlled recording or stimulation at the level of basal ganglia nuclei discussed the physiological observations underlying the engagement of structures affected by the electrical field around the electrode in motor and sensory functions (*Chap. 21*). Based on the information from the fused images of preoperative MRI-postoperative computer tomography, it was emphasized that the documentation of the electrode position by using mutual information technique after unilateral or bilateral STN stimulation provides useful information for the prediction of the surgical outcome (*Chap. 22*). Fusion-image-based programming and reprogramming of the stimulator parameters using the visual information of the location of the electrode contacts propose to find the best sites and the best stimulation parameters for the advanced PD patients treated with deep brain stimulation of subthalamic nucleus (STN-DBS). The electrical motor cortex stimulation can improve axial symptom of PD, which is difficult even with STN-DBS, and is less surgical invasiveness than deep brain stimulation can make the surgery safer for the patients with advanced age or sever brain atrophy *(Chap. 23*). The last contribution in the book deals with the rational arthroplasty surgery and multidisciplinary approaches for the patients with

We would like to thank all the people who supported the preparation of this book, who contributed to the book, and in particular also to all who made the book possible

> **Dr. Juliana Dushanova**  Institute of Neurobiology, Bulgarian Academy of Sciences,

> > Sofia, Bulgaria

due to their positive evaluations of the proposals for this book.

**Part 1** 

**Biomarkers for Preclinical Diagnosis of PD** 

## **Part 1**

## **Biomarkers for Preclinical Diagnosis of PD**

**1** 

*Messina Italy* 

**Early Marker for the Diagnosis** 

Silvia Marino, Pietro Lanzafame, Silvia Guerrera,

Parkinson's disease (PD) is a progressive disorder with a relentless neuronal cell loss in several brain areas and nuclei notably in the substantia nigra (SN). The course of this neuronal loss is still unclear and may be highly variable in different PD patients and at

At present, no treatment has proven to influence this progressive course of the disease by

One potential reason for the lack of neuroprotective effects of various agents, which have been highly effective in animal experiments, is the fact that the neurodegenerative process has already substantially proceeded when the diagnosis is established on the basis of widely accepted diagnostic criteria for PD: when the patients fulfill the clinical criteria of PD, 60– 70% of neurons of the SN are degenerated and the striatal dopamine content is reduced by

The "preclinical" phase may give the incorrect impression of patients exhibiting no clinical signs or symptoms of the incipient disease. Conversely, it is known that motor signs develop insidiously and minor signs of asymmetric hypokinesia may be detected years before the diagnosis of PD can be established. In addition, non-motor symptoms such as mood disorders, olfactorial, vegetative, sensory or neuropsychological signs may be noticed by the patients or physicians in advance of motor signs reflecting the dysfunction of

Therefore, the term "early" or "prediagnostic" phase of PD would more appropriately characterize this stage of the disease. The clinical impression of autonomic, olfactorial and affective symptoms preceding motor signs of PD are in line with the findings demonstrating that neuronal alteration, with regard to Lewy body formation, occurs first in the dorsal vagal nucleus, the olfactory bulb, the raphe and coeruleus nuclei before entering the SN. According to neuropathological findings, it is suggested that approximately 10% of subjects older than 60 years are in the "prediagnostic" phase of PD. These subjects exhibit the pathological hallmarks of PD, like Lewy bodies and neuronal loss at the SN, without showing the motor signs during life time that allow the diagnosis of PD. In only 10% of this group with so-called "incidental Lewy body disease", neuronal loss will proceed reaching

the degree where motor symptoms are distinct enough to allow the diagnosis of PD.

80%, suggesting that the remaining neurons of the SN are also altered.

**1. Introduction** 

different phases of the disease.

protecting neurons or by postponing cell death.

dopaminergic or non-dopaminergic neurons.

**of Parkinson's Disease** 

Rosella Ciurleo and Placido Bramanti *IRCCS Centro Neurolesi "Bonino-Pulejo"* 

## **Early Marker for the Diagnosis of Parkinson's Disease**

Silvia Marino, Pietro Lanzafame, Silvia Guerrera, Rosella Ciurleo and Placido Bramanti *IRCCS Centro Neurolesi "Bonino-Pulejo" Messina Italy* 

## **1. Introduction**

Parkinson's disease (PD) is a progressive disorder with a relentless neuronal cell loss in several brain areas and nuclei notably in the substantia nigra (SN). The course of this neuronal loss is still unclear and may be highly variable in different PD patients and at different phases of the disease.

At present, no treatment has proven to influence this progressive course of the disease by protecting neurons or by postponing cell death.

One potential reason for the lack of neuroprotective effects of various agents, which have been highly effective in animal experiments, is the fact that the neurodegenerative process has already substantially proceeded when the diagnosis is established on the basis of widely accepted diagnostic criteria for PD: when the patients fulfill the clinical criteria of PD, 60– 70% of neurons of the SN are degenerated and the striatal dopamine content is reduced by 80%, suggesting that the remaining neurons of the SN are also altered.

The "preclinical" phase may give the incorrect impression of patients exhibiting no clinical signs or symptoms of the incipient disease. Conversely, it is known that motor signs develop insidiously and minor signs of asymmetric hypokinesia may be detected years before the diagnosis of PD can be established. In addition, non-motor symptoms such as mood disorders, olfactorial, vegetative, sensory or neuropsychological signs may be noticed by the patients or physicians in advance of motor signs reflecting the dysfunction of dopaminergic or non-dopaminergic neurons.

Therefore, the term "early" or "prediagnostic" phase of PD would more appropriately characterize this stage of the disease. The clinical impression of autonomic, olfactorial and affective symptoms preceding motor signs of PD are in line with the findings demonstrating that neuronal alteration, with regard to Lewy body formation, occurs first in the dorsal vagal nucleus, the olfactory bulb, the raphe and coeruleus nuclei before entering the SN.

According to neuropathological findings, it is suggested that approximately 10% of subjects older than 60 years are in the "prediagnostic" phase of PD. These subjects exhibit the pathological hallmarks of PD, like Lewy bodies and neuronal loss at the SN, without showing the motor signs during life time that allow the diagnosis of PD. In only 10% of this group with so-called "incidental Lewy body disease", neuronal loss will proceed reaching the degree where motor symptoms are distinct enough to allow the diagnosis of PD.

Early Marker for the Diagnosis of Parkinson's Disease 5

Oculomotor movements are controlled by many brain areas including the cerebral cortex, basal ganglia, brain stem and cerebellum. PD is a condition of degeneration of dopaminergic neurons in the substantia nigra pars compacta, resulting in progressive basal ganglia dysfunction. Because important eye movement pathways travel through the basal ganglia, aspects of oculomotor movement control should be impaired by the disease progression. This study showed that deficit in Pursuit Ocular Movements (POM) also occurs in patients

The authors used a vision-based non-intrusive eye tracker. The developed interface provides the patients a visual stimulation. This system was able to measure, analyze and record the

The subjects were seated at 60 cm from the scene monitor, in front of the camera, on a chair which could be raised or lowered so that the subject's eyes were at the same height as the PC monitor, when the visual stimulus was administered on. The subjects were asked to

Fig. 1. Video-based eye-tracking functional scheme: subject looks at the screen (laptop or desktop PC) and all mechanical and electronic supports help to perform a real-time

The results of the study confirm that POM are clearly impaired in patients with de novo PD. The same authors (Marino el al., 2010) studied the POM by using the same vision-based non-intrusive eye tracker, in patients with suspected PD, before and after L-Dopa

All patients had a positive test demonstrated by the improvement of UPDRS motor subscore, after L-Dopa administration, and as a new finding, by the improvement of POM. A plausible explanation is that the improvement of horizontal eye displacement gain was induced by the dopaminergic action of L-Dopa. Some newly diagnosed PD patients have been shown to improve POM after L-Dopa treatment and this suggested the possibility of dopaminergic control of ocular movements, particularly smooth pursuit and saccades.

with non-advanced PD and is closely correlated with clinical stage and motor scores.

resulting horizontal eye movements.

perform the test three times.

acquisition of eye movements.

administration.

It would be of great interest with respect to research and treatment to identify those subjects at risk i) to initiate neuroprotective treatment earlier, giving them a better base to act and ii) to define the causes of more rapid neuronal loss and disease progression in those patients with "incidental Lewy body disease" who will cross the threshold of critical neuronal loss at the SN and develop PD.

The duration of the early or prediagnostic period remains unknown. The duration of this phase of PD was estimated to last from a few years up to several decades before the first symptoms are noticed by the patients.

Several procedures have been proposed to identify subjects in early ("prediagnostic") stages of PD. In the following we present some instrumental approaches to identify patients in the early stages of PD.

One set of simple behavioural tasks that may provide insight into the neural control of response suppression uses saccadic eye movements to investigate and quantify motor impairments in PD. The study of ocular movements has been increasingly used to detect subtle pathological modifications, caused by a wide variety of neurological diseases.

A recent method, a new vision-based, nonintrusive, eye tracker, previously described in de novo PD patients (Marino et al., 2007), was proposed as a possible tool for supporting the diagnosis of PD in association with levodopa test, as an add-on to the Unified Parkinson Disease Rating Scale (UPDRS) score (Marino et al., 2010).

In addition, conventional MR Imaging (cMRI), as well as different advanced MRI techniques, including magnetic resonance spectroscopy (MRS), magnetization transfer imaging (MTI), diffusion-weighted and diffusion tensor imaging (DWI/DTI) are helpful to distinguish PD from atypical or secondary PD, especially in early stage of disease where a differentiation of these conditions is not easy.

Objective olfaction tests, such as olfactory-evocated potentials or functional magnetic resonance imaging, can be used to assess the severity of olfactory dysfunction, an early clinical feature of PD, its correlation with cerebral changes, and then the risk of developing PD in asymptomatic subjects.

## **2. Analysis of pursuit ocular movements in Parkinson's disease by using a video-based eye tracking system**

Patients with PD characteristically have difficulty initiating movements (akinesia). When movements are initiated, they are of low velocity (bradykinesia) and reduced amplitude (hypokinesia). In addition, patients with PD are unable to sustain repetitive motor action. When they attempt to open or close the hand rapidly or tap the foot on the ground, the movement rapidly decreases in amplitude and slows in speed until it ceases. This disability is easily appreciated in the progressive micrographia of the handwriting of PD patients.

Research in the past 30 years has established that PD impairs control of eye movements. Voluntary saccades, such as self-paced, predictive and remembered saccades, are hypometric, multistep, of reduced velocity and of increased duration. Visually guided saccades are normal. Advanced PD is known to be associated with reduced ocular smooth pursuit gain (Bares et al., 2003; Lekwuwa et al., 1999). This has been explained in terms of advanced PD affecting other structures outside the basal ganglia.

A recent study (Marino et al., 2007) described a new eye movement measurement and analysis system which was developed for generating a set of visual stimuli paradigms and which is able to measure, analyze and record the resulting horizontal eye movements.

It would be of great interest with respect to research and treatment to identify those subjects at risk i) to initiate neuroprotective treatment earlier, giving them a better base to act and ii) to define the causes of more rapid neuronal loss and disease progression in those patients with "incidental Lewy body disease" who will cross the threshold of critical neuronal loss at

The duration of the early or prediagnostic period remains unknown. The duration of this phase of PD was estimated to last from a few years up to several decades before the first

Several procedures have been proposed to identify subjects in early ("prediagnostic") stages of PD. In the following we present some instrumental approaches to identify patients in the

One set of simple behavioural tasks that may provide insight into the neural control of response suppression uses saccadic eye movements to investigate and quantify motor impairments in PD. The study of ocular movements has been increasingly used to detect

A recent method, a new vision-based, nonintrusive, eye tracker, previously described in de novo PD patients (Marino et al., 2007), was proposed as a possible tool for supporting the diagnosis of PD in association with levodopa test, as an add-on to the Unified Parkinson

In addition, conventional MR Imaging (cMRI), as well as different advanced MRI techniques, including magnetic resonance spectroscopy (MRS), magnetization transfer imaging (MTI), diffusion-weighted and diffusion tensor imaging (DWI/DTI) are helpful to distinguish PD from atypical or secondary PD, especially in early stage of disease where a

Objective olfaction tests, such as olfactory-evocated potentials or functional magnetic resonance imaging, can be used to assess the severity of olfactory dysfunction, an early clinical feature of PD, its correlation with cerebral changes, and then the risk of developing

**2. Analysis of pursuit ocular movements in Parkinson's disease by using a** 

Patients with PD characteristically have difficulty initiating movements (akinesia). When movements are initiated, they are of low velocity (bradykinesia) and reduced amplitude (hypokinesia). In addition, patients with PD are unable to sustain repetitive motor action. When they attempt to open or close the hand rapidly or tap the foot on the ground, the movement rapidly decreases in amplitude and slows in speed until it ceases. This disability is easily appreciated in the progressive micrographia of the handwriting of PD patients. Research in the past 30 years has established that PD impairs control of eye movements. Voluntary saccades, such as self-paced, predictive and remembered saccades, are hypometric, multistep, of reduced velocity and of increased duration. Visually guided saccades are normal. Advanced PD is known to be associated with reduced ocular smooth pursuit gain (Bares et al., 2003; Lekwuwa et al., 1999). This has been explained in terms of

A recent study (Marino et al., 2007) described a new eye movement measurement and analysis system which was developed for generating a set of visual stimuli paradigms and which is able to measure, analyze and record the resulting horizontal eye movements.

subtle pathological modifications, caused by a wide variety of neurological diseases.

the SN and develop PD.

early stages of PD.

symptoms are noticed by the patients.

Disease Rating Scale (UPDRS) score (Marino et al., 2010).

advanced PD affecting other structures outside the basal ganglia.

differentiation of these conditions is not easy.

PD in asymptomatic subjects.

**video-based eye tracking system** 

Oculomotor movements are controlled by many brain areas including the cerebral cortex, basal ganglia, brain stem and cerebellum. PD is a condition of degeneration of dopaminergic neurons in the substantia nigra pars compacta, resulting in progressive basal ganglia dysfunction. Because important eye movement pathways travel through the basal ganglia, aspects of oculomotor movement control should be impaired by the disease progression. This study showed that deficit in Pursuit Ocular Movements (POM) also occurs in patients with non-advanced PD and is closely correlated with clinical stage and motor scores.

The authors used a vision-based non-intrusive eye tracker. The developed interface provides the patients a visual stimulation. This system was able to measure, analyze and record the resulting horizontal eye movements.

The subjects were seated at 60 cm from the scene monitor, in front of the camera, on a chair which could be raised or lowered so that the subject's eyes were at the same height as the PC monitor, when the visual stimulus was administered on. The subjects were asked to perform the test three times.

Fig. 1. Video-based eye-tracking functional scheme: subject looks at the screen (laptop or desktop PC) and all mechanical and electronic supports help to perform a real-time acquisition of eye movements.

The results of the study confirm that POM are clearly impaired in patients with de novo PD. The same authors (Marino el al., 2010) studied the POM by using the same vision-based non-intrusive eye tracker, in patients with suspected PD, before and after L-Dopa administration.

All patients had a positive test demonstrated by the improvement of UPDRS motor subscore, after L-Dopa administration, and as a new finding, by the improvement of POM.

A plausible explanation is that the improvement of horizontal eye displacement gain was induced by the dopaminergic action of L-Dopa. Some newly diagnosed PD patients have been shown to improve POM after L-Dopa treatment and this suggested the possibility of dopaminergic control of ocular movements, particularly smooth pursuit and saccades.

Early Marker for the Diagnosis of Parkinson's Disease 7

markers. Proton spectroscopy presents the problem that metabolites at millimolar concentration must be detected in the presence of a background water signal that is present at about 100 molar. For this reason solvent-suppression techniques have been combined with localization schemes to produce spatially localized solvent-suppressed spectra. The two most commonly used localization methods are STEAM and PRESS. These methods can be implemented as single-voxel and multi-voxels methods. With the single voxel localization, the signal is acquired from a single brick-shaped volume of various sizes (minimum volume 2-3 cm3). The multi-voxel or MR spectroscopic imaging (MRSI) or Chemical Shift Imaging (CSI) approach generates individual spectra from multiple voxels at the same time (minimum volume 0.5-1 cm3). Single-voxel spectroscopy detects the signal from a single region during one measurement, whereas MRS imaging, using additional phase-encoding pulses, obtains the signal from multiple regions at the same time and provides the information of spatial distribution of major cerebral metabolites. The spatial information in MRI is done in 2-D for one or more slices and can generate low-resolution images for each metabolite by integration of the MR signals from each voxel (Ross & Bluml, 2001). The possibility to acquire the spectra from 2D multi-voxel allows to study the metabolite distribution of a large area of the brain with the advantage of identifying more anatomical and functional details. Most importantly, collecting data from many different adjacent regions simultaneously reduces the potential for systemic errors that can affect

sequential measurements and thus results in more accurate repeated studies.

suggesting an important, unknown role in brain metabolism (Clark, 1998).

viewed as a marker of membrane turnover or inflammation in 1H MRS studies.

(Figure 2) (Bonavita et al., 1999; Lin et al., 2005).

The metabolites detectable with 1H-MRS include the prominent resonances of *N*acetylaspartate (NAA), choline-containing compounds (Cho), creatine + phosphocreatine (Cr), myo-inositol (mI), lactate (Lac), and a variety of other resonances that might not be evident depending on type and quality of spectra as well as on the pathological condition

NAA, which resonates at 2.02 parts per million (ppm), represents the largest proton metabolic concentration in the human brain after water. Indeed the concentration of NAA reaches on the order of 10 μmol/g. NAA is widely interpreted as a neuronal marker and implicated in several neuronal processes, mitochondrial functioning and osmoregulation. NAA synthesis occurs in mitochondria and requires acetyl-CoA and L-aspartic acid as substrates. NAA has been proposed to serve as a mitochondrial shuttle of acetyl-CoA used for fatty acid synthesis. NAA undergoes dramatic increase during brain development and significant decrease during lesion progression in various neurodegenerative diseases,

The Cho peak (3.2 ppm) represents a combination of several choline-containing compounds, including free Cho, phosphorylcholine and glycerophosphorylcholine, and to a small extent acetylcholine. Free Cho acts as a precursor to acetylcholine, while glycerophosphorylcholine is a product of breakdown of membrane phosphatidylcholine and acts as an osmoregulator. The concentration of Cho is relatively low on the order of 0.5 to 1.5 μmol/g and can be altered in normal aging and many focal inflammatory diseases. The Cho peak is often

The concentration of total Cr is estimated on the order of 8 to 9 μmol/g and is approximately 20% higher in human gray matter than white matter. In 1H-MRS, the resonance at 3.03 ppm represents total Cr and PCr supplies phosphate for conversion of ADP to ATP in creatine kinase reaction. Indeed these metabolites buffer the energy use and energy storage of cells. The level of total Cr mainly remains constant in many neuronal

The POM methodology could be considered as a not invasive, objective and repetitive method (and these conditions could be an advantage with respect to only UPDRS examination) to support the clinical evaluation.

This method could be considered as a possible tool for supporting the diagnosis of PD in association with levodopa test, as an add-on to the UPDRS score. These results showed that this vision-based eye tracker can be used as reliable indices of disease severity in early and suspected PD patients.

## **3. Diagnosis of Parkinson's disease by using MR techniques**

PD in its early stages can easily be mistaken for any number of disorders. Indeed PD is most likely to be confused with various Atypical Parkinsonian Disorders (APDs) such as Progressive Supranuclear Palsy (PSP), Multiple-System Atrophy (MSA), especially the Parkinson variant of MSA (MSA-P), and Corticobasal Degeneration (CBD).

A differentiation of these clinical entities, each characterized by completely different natural histories, may be challenging, particularly in the early stages of the disease, where overlapping clinical signs lead to a high rate of misclassification. However, a differentiation between APDs and PD, that may make easier early diagnosis, is crucial for determining the prognosis and choosing a treatment strategy.

Magnetic Resonance Imaging (MRI) plays an important role in the differential diagnosis in PD. Conventional MRI (cMRI) and advanced MRI techniques, including proton magnetic resonance spectroscopy, diffusion-weighted and diffusion tensor imaging and magnetization transfer imaging, are helpful to distinguish PD from atypical or secondary PD.

## **3.1 Magnetic Resonance Spectroscopy**

MRS is a non-invasive technique that can be used to measure the concentrations of different low-molecular weight chemicals. The technique is based on the same physical principles as MRI, i.e. the detection of energy exchanges between external magnetic fields and specific nuclei within atoms. MRS is the more modern version of Nuclear Magnetic Resonance which over the past five decades has evolved from a technique used in chemistry to determine the structure of molecules to a method with which to probe the metabolism of cells, tissues, intact animals and humans (Allen, 1990; Avison et al., 1986).

MRS has been demonstrated in vivo for different nuclei, including 1H, 31P, 13C, 15N, 19F and 23Na. While most of these nuclei are very difficult to detect, 1H and 31P are available in the human brain in significant concentration and have the appropriate physical configuration to be detected by MRS. For instance, 31P-MRS has been the first to be applied to medicine in vivo, and can be used to evaluate brain energy metabolism by directly and non-invasively measuring of Adenosine Triphosphate (ATP), Phosphocreatine (PCr) or Inorganic Phosphate (Pi) concentrations. While 31P-MRS was the first spectroscopic technique to be applied in vivo, the main nucleus studied today in neurospectroscopy is 1H, which provides information on markers of neurons, myelin, energy metabolism and other metabolically active compounds.

1H-MRS detects very small differences in the frequencies of proton resonances from comparatively large volumes (1 ml or more) of brain tissue. The frequency of the resonance is affected by its local chemical environment, while the amplitude reflects its concentration. As such, 1H-MRS is able to provide a measurement of certain proton-containing chemical

The POM methodology could be considered as a not invasive, objective and repetitive method (and these conditions could be an advantage with respect to only UPDRS

This method could be considered as a possible tool for supporting the diagnosis of PD in association with levodopa test, as an add-on to the UPDRS score. These results showed that this vision-based eye tracker can be used as reliable indices of disease severity in early and

PD in its early stages can easily be mistaken for any number of disorders. Indeed PD is most likely to be confused with various Atypical Parkinsonian Disorders (APDs) such as Progressive Supranuclear Palsy (PSP), Multiple-System Atrophy (MSA), especially the

A differentiation of these clinical entities, each characterized by completely different natural histories, may be challenging, particularly in the early stages of the disease, where overlapping clinical signs lead to a high rate of misclassification. However, a differentiation between APDs and PD, that may make easier early diagnosis, is crucial for determining the

Magnetic Resonance Imaging (MRI) plays an important role in the differential diagnosis in PD. Conventional MRI (cMRI) and advanced MRI techniques, including proton magnetic resonance spectroscopy, diffusion-weighted and diffusion tensor imaging and magnetization transfer imaging, are helpful to distinguish PD from atypical or secondary

MRS is a non-invasive technique that can be used to measure the concentrations of different low-molecular weight chemicals. The technique is based on the same physical principles as MRI, i.e. the detection of energy exchanges between external magnetic fields and specific nuclei within atoms. MRS is the more modern version of Nuclear Magnetic Resonance which over the past five decades has evolved from a technique used in chemistry to determine the structure of molecules to a method with which to probe the metabolism of

MRS has been demonstrated in vivo for different nuclei, including 1H, 31P, 13C, 15N, 19F and 23Na. While most of these nuclei are very difficult to detect, 1H and 31P are available in the human brain in significant concentration and have the appropriate physical configuration to be detected by MRS. For instance, 31P-MRS has been the first to be applied to medicine in vivo, and can be used to evaluate brain energy metabolism by directly and non-invasively measuring of Adenosine Triphosphate (ATP), Phosphocreatine (PCr) or Inorganic Phosphate (Pi) concentrations. While 31P-MRS was the first spectroscopic technique to be applied in vivo, the main nucleus studied today in neurospectroscopy is 1H, which provides information on markers of neurons, myelin, energy metabolism and other metabolically

1H-MRS detects very small differences in the frequencies of proton resonances from comparatively large volumes (1 ml or more) of brain tissue. The frequency of the resonance is affected by its local chemical environment, while the amplitude reflects its concentration. As such, 1H-MRS is able to provide a measurement of certain proton-containing chemical

**3. Diagnosis of Parkinson's disease by using MR techniques** 

Parkinson variant of MSA (MSA-P), and Corticobasal Degeneration (CBD).

cells, tissues, intact animals and humans (Allen, 1990; Avison et al., 1986).

examination) to support the clinical evaluation.

prognosis and choosing a treatment strategy.

**3.1 Magnetic Resonance Spectroscopy** 

suspected PD patients.

PD.

active compounds.

markers. Proton spectroscopy presents the problem that metabolites at millimolar concentration must be detected in the presence of a background water signal that is present at about 100 molar. For this reason solvent-suppression techniques have been combined with localization schemes to produce spatially localized solvent-suppressed spectra. The two most commonly used localization methods are STEAM and PRESS. These methods can be implemented as single-voxel and multi-voxels methods. With the single voxel localization, the signal is acquired from a single brick-shaped volume of various sizes (minimum volume 2-3 cm3). The multi-voxel or MR spectroscopic imaging (MRSI) or Chemical Shift Imaging (CSI) approach generates individual spectra from multiple voxels at the same time (minimum volume 0.5-1 cm3). Single-voxel spectroscopy detects the signal from a single region during one measurement, whereas MRS imaging, using additional phase-encoding pulses, obtains the signal from multiple regions at the same time and provides the information of spatial distribution of major cerebral metabolites. The spatial information in MRI is done in 2-D for one or more slices and can generate low-resolution images for each metabolite by integration of the MR signals from each voxel (Ross & Bluml, 2001). The possibility to acquire the spectra from 2D multi-voxel allows to study the metabolite distribution of a large area of the brain with the advantage of identifying more anatomical and functional details. Most importantly, collecting data from many different adjacent regions simultaneously reduces the potential for systemic errors that can affect sequential measurements and thus results in more accurate repeated studies.

The metabolites detectable with 1H-MRS include the prominent resonances of *N*acetylaspartate (NAA), choline-containing compounds (Cho), creatine + phosphocreatine (Cr), myo-inositol (mI), lactate (Lac), and a variety of other resonances that might not be evident depending on type and quality of spectra as well as on the pathological condition (Figure 2) (Bonavita et al., 1999; Lin et al., 2005).

NAA, which resonates at 2.02 parts per million (ppm), represents the largest proton metabolic concentration in the human brain after water. Indeed the concentration of NAA reaches on the order of 10 μmol/g. NAA is widely interpreted as a neuronal marker and implicated in several neuronal processes, mitochondrial functioning and osmoregulation. NAA synthesis occurs in mitochondria and requires acetyl-CoA and L-aspartic acid as substrates. NAA has been proposed to serve as a mitochondrial shuttle of acetyl-CoA used for fatty acid synthesis. NAA undergoes dramatic increase during brain development and significant decrease during lesion progression in various neurodegenerative diseases, suggesting an important, unknown role in brain metabolism (Clark, 1998).

The Cho peak (3.2 ppm) represents a combination of several choline-containing compounds, including free Cho, phosphorylcholine and glycerophosphorylcholine, and to a small extent acetylcholine. Free Cho acts as a precursor to acetylcholine, while glycerophosphorylcholine is a product of breakdown of membrane phosphatidylcholine and acts as an osmoregulator.

The concentration of Cho is relatively low on the order of 0.5 to 1.5 μmol/g and can be altered in normal aging and many focal inflammatory diseases. The Cho peak is often viewed as a marker of membrane turnover or inflammation in 1H MRS studies.

The concentration of total Cr is estimated on the order of 8 to 9 μmol/g and is approximately 20% higher in human gray matter than white matter. In 1H-MRS, the resonance at 3.03 ppm represents total Cr and PCr supplies phosphate for conversion of ADP to ATP in creatine kinase reaction. Indeed these metabolites buffer the energy use and energy storage of cells. The level of total Cr mainly remains constant in many neuronal

Early Marker for the Diagnosis of Parkinson's Disease 9

aerobic conditions. Increases of Lac in the brain are often accompanied by decreased intracellular pH and high-energy phosphates. The proposed role of Lac is a source of energy for neurons and the transport of Lac plays an essential role in the concept of metabolic

The concentration changes of all metabolites detected by 1H-MRS and 31P-MRS could help to evaluate PD subjects in the "preclinical" stages, especially in early differential diagnosis. 1H-MRS of striatal structures might differentiate PD from APDs by virtue of reduced NAA/Cr ratios in MSA but not PD. 1H-MRS showed reduced NAA/Cr ratios in the lentiform nucleus in six of seven MSA-P cases, whereas normal levels of putaminal NAA

As compared to normal controls, in patients with PSP, CBD, and MSA, but not in those with PD, significant reduction of the NAA/Cr ratio in the frontal cortex was found (Abe et al., 2009). Patients with CBD showed significant reduction of the NAA/Cr ratio in the frontal cortex and putamen as compared to patients with PD and MSA. Patients with PSP showed a significant reduction of the NAA/Cr ratio in the putamen as compared with patients with PD and MSA. Patients with CBD showed clear asymmetry in the putamen as compared to controls and other patients (Abe et al., 2009). By application of 1H-MRSI statistically significant difference in regional patterns of the NAA/Cr and NAA/Cho ratios between patients with PD and those with CBD and between patients with PD and those with PSP

Other 1H-MRS examinations didn't show significant difference between the PD patients and the control subjects (Tedeschi et al., 1997), also in the striatum (Holshouser et al., 1995), in the putamen and globus pallidus (Federico et al., 1997), and in occipital lobe (Bowen et al., 1995). The NAA/Cho and NAA/Cr ratios were significantly reduced in the putamen and globus pallidus of MSA and the PSP patients, in which neuronal loss involves, compared with the control subjects (Federico et al., 1997). In another study Federico at al. showed that NAA/Cho peak ratio was significantly reduced in MSA and in PSP patients compared to PD patients and to control. Moreover the NAA/Cr peak ratio was significantly reduced in MSA, in PSP and in PD patients also compared to controls, but only in MSA compared to

Normal 1H-MRS data could suggest the clinical diagnosis of PD, whereas low striatal levels

However, further MRS studies have shown reduced NAA/Cr and NAA/Cho ratios in the lentiform nucleus not only in APD, but also in PD (Clarke & Lowry, 2001; Firbank et al.,

Technical factors such as MRS technique including different echo- time and relaxation-time, voxel sizes, field strength and pulse sequences used in the different studies, may be responsible for some of the variation of results seen in the published literature on 1H-MRS for the differential diagnosis of neurodegenerative parkinsonism (Clarke & Lowry, 2001; Firbank et al., 2002). The development of 1H-MRS at higher magnetic field strengths may lead 1H-MRS to a more important role as imaging tool in the differential diagnosis of

1H-MRS of the brain with high magnetic field at 3 Tesla has many advantages that, with respect to the well-established and technologically advanced 1.5 Tesla 1H-MRS, include better signal to noise ratio (SNR) and increased spectral, spatial and temporal resolution,

coupling between neurons and glia.

was found (Tedeschi et al., 1997).

PD patients (Federico et al., 1999).

parkinsonian disorders.

2002).

of NAA could suggest the diagnosis of MSA or PSP.

were found in eight of nine PD subjects (Davie et al., 1995).

Fig. 2. Chemical structure of main cerebral metabolites detected by 1H-MRS.

diseases. Thus, total Cr is often used as an internal reference (i.e., a denominator in metabolite signal ratio).

The mI (3.56 ppm) has been recognized as a cerebral osmolyte or an astrocyte marker due to its cellular specificity based on cell culture studies. mI is also been known as a breakdown product of myelin and precursor of inositol polyphosphate, an intracellular messenger. The concentration of mI is on the order of 5–10 μmol/g while one of its isomers, syllo-inositol, has substantially lower concentrations of on the order of less than 1 mol/g in the brain and remains relatively consistent in many diseases.

The Lac (1.3 ppm) is an end product of anaerobic glycolysis, thus increase in Lac concentrations often serves as an index of altered oxidative metabolism, i.e., in ischemia, hypoxia, and cancer. The concentration of Lac is on the order of about 1 μmol/g in normal

**NH**

**+**

**N CH3**

**Creatine (Cre)** 

**COO**

**-**

**N(CH3)3 +**

**Choline - containing compounds (Cho)** 

**OH**

**<sup>3</sup>CH OH**

**Lactate (Lac)**

**N(CH3)3** +

**COO-**

**<sup>2</sup>NH**

**O P O**

**O**

**O -**

Fig. 2. Chemical structure of main cerebral metabolites detected by 1H-MRS.

metabolite signal ratio).

remains relatively consistent in many diseases.

**COO**

**OH**

**OH**

**NH**

**N-Acetylaspartate (NAA)**

**OH OH OH**

**- OOC**

**CO <sup>3</sup>CH**

**OH**

**OH**

**Myo-inositol (mI)** 

**OH**

**-**

diseases. Thus, total Cr is often used as an internal reference (i.e., a denominator in

The mI (3.56 ppm) has been recognized as a cerebral osmolyte or an astrocyte marker due to its cellular specificity based on cell culture studies. mI is also been known as a breakdown product of myelin and precursor of inositol polyphosphate, an intracellular messenger. The concentration of mI is on the order of 5–10 μmol/g while one of its isomers, syllo-inositol, has substantially lower concentrations of on the order of less than 1 mol/g in the brain and

The Lac (1.3 ppm) is an end product of anaerobic glycolysis, thus increase in Lac concentrations often serves as an index of altered oxidative metabolism, i.e., in ischemia, hypoxia, and cancer. The concentration of Lac is on the order of about 1 μmol/g in normal aerobic conditions. Increases of Lac in the brain are often accompanied by decreased intracellular pH and high-energy phosphates. The proposed role of Lac is a source of energy for neurons and the transport of Lac plays an essential role in the concept of metabolic coupling between neurons and glia.

The concentration changes of all metabolites detected by 1H-MRS and 31P-MRS could help to evaluate PD subjects in the "preclinical" stages, especially in early differential diagnosis.

1H-MRS of striatal structures might differentiate PD from APDs by virtue of reduced NAA/Cr ratios in MSA but not PD. 1H-MRS showed reduced NAA/Cr ratios in the lentiform nucleus in six of seven MSA-P cases, whereas normal levels of putaminal NAA were found in eight of nine PD subjects (Davie et al., 1995).

As compared to normal controls, in patients with PSP, CBD, and MSA, but not in those with PD, significant reduction of the NAA/Cr ratio in the frontal cortex was found (Abe et al., 2009). Patients with CBD showed significant reduction of the NAA/Cr ratio in the frontal cortex and putamen as compared to patients with PD and MSA. Patients with PSP showed a significant reduction of the NAA/Cr ratio in the putamen as compared with patients with PD and MSA. Patients with CBD showed clear asymmetry in the putamen as compared to controls and other patients (Abe et al., 2009). By application of 1H-MRSI statistically significant difference in regional patterns of the NAA/Cr and NAA/Cho ratios between patients with PD and those with CBD and between patients with PD and those with PSP was found (Tedeschi et al., 1997).

Other 1H-MRS examinations didn't show significant difference between the PD patients and the control subjects (Tedeschi et al., 1997), also in the striatum (Holshouser et al., 1995), in the putamen and globus pallidus (Federico et al., 1997), and in occipital lobe (Bowen et al., 1995). The NAA/Cho and NAA/Cr ratios were significantly reduced in the putamen and globus pallidus of MSA and the PSP patients, in which neuronal loss involves, compared with the control subjects (Federico et al., 1997). In another study Federico at al. showed that NAA/Cho peak ratio was significantly reduced in MSA and in PSP patients compared to PD patients and to control. Moreover the NAA/Cr peak ratio was significantly reduced in MSA, in PSP and in PD patients also compared to controls, but only in MSA compared to PD patients (Federico et al., 1999).

Normal 1H-MRS data could suggest the clinical diagnosis of PD, whereas low striatal levels of NAA could suggest the diagnosis of MSA or PSP.

However, further MRS studies have shown reduced NAA/Cr and NAA/Cho ratios in the lentiform nucleus not only in APD, but also in PD (Clarke & Lowry, 2001; Firbank et al., 2002).

Technical factors such as MRS technique including different echo- time and relaxation-time, voxel sizes, field strength and pulse sequences used in the different studies, may be responsible for some of the variation of results seen in the published literature on 1H-MRS for the differential diagnosis of neurodegenerative parkinsonism (Clarke & Lowry, 2001; Firbank et al., 2002). The development of 1H-MRS at higher magnetic field strengths may lead 1H-MRS to a more important role as imaging tool in the differential diagnosis of parkinsonian disorders.

1H-MRS of the brain with high magnetic field at 3 Tesla has many advantages that, with respect to the well-established and technologically advanced 1.5 Tesla 1H-MRS, include better signal to noise ratio (SNR) and increased spectral, spatial and temporal resolution,

Early Marker for the Diagnosis of Parkinson's Disease 11

al., 2009). In the putamen and midbrain of both PD groups compared to control was found a bilateral reduction of high-energy phosphates such as ATP and PCr as final acceptors of energy from mitochondrial oxidative phosphorylation. In contrast, low-energy metabolites such as ADP and Pi were within normal ranges. Patients with early Parkinson's Disease, with clearly lateralized motor symptoms, exhibited a significant reduction of putamen highenergy phosphates in the less affected hemisphere with a less pronounced dopaminergic cell loss. Therefore, mitochondrial dysfunction is a rather early occurring and subsequently persistent event in the pathophysiology of dopaminergic degeneration in PD. These data strongly support the hypothesis that mitochondrial dysfunction is involved early in

In vivo MRS is increasingly utilized for the study of neurochemistry and cerebral energy metabolism in PD. Particularly, the recent technical advances of in vivo MRS including the availability of higher magnetic fields permitting improved spectral and spatial resolution, the development of a reliable method for absolute metabolite quantification, the development of various spectroscopic methods to enhance metabolite signal identification, and the application of combined 31P- and 1H-MRS can be use to examine the changes in neurochemical profile non–invasively and to achieve a differential diagnosis of PD versus other forms of parkinsonism, especially in early stages of disease when signs and symptoms of different forms of parkinsonism have greater overlap. However, several multicentre trials using a larger sample of patients, absolute quantification of tissue metabolite concentrations and a standardized technique are required to fully determine the place of

 In the early disease stages the clinical separation of atypical parkinsonism disorders (APD)s from PD carries a high rate of misdiagnosis. An early differentiation between APD and PD, each characterized by completely different natural histories, is crucial for determining the

The principles of MR imaging are based on the ubiquitous presence of hydrogen in body tissues and the spin of the hydrogen atom proton, which induces a small magnetic field. In general, T2- weighted sequences are sensitive to changes in tissue properties, including tissue damage, due to changes of the transverse magnetization or T2 decay. Neurodegenerative processes characterized by cell loss, increased age-related deposition of iron or other paramagnetic substances, and by astroglial reaction and microglial proliferation may lead to signal changes in affected brain areas, like the basal ganglia or infratentorial structures, in neurodegenerative parkinsonism (Duguid et al., 1986; Gupta et

Because cMRI is believed to be usually normal in patients with PD, while it frequently shows characteristic abnormalities in patients with APD, cMRI images takes a major part in excluding underlying pathologies such as vascular lesions, multiple sclerosis, brain tumors, normal pressure hydrocephalus, bilateral striopallidodentate calcinosis, and other potential, but rare, causes of symptomatic parkinsonism such as Wilson disease, manganese-induced parkinsonism, or different subtypes of neurodegeneration associated with brain iron

At 1,5 T, patients with advanced PD, and sometimes those with APD, may show distinct abnormalities of the substantia nigra, including signal increase on T2-weighted MR images, smudging of the hypointensity in the substantia nigra towards the red nucleus or signal loss

pathogenesis of PD and it may be used as early marker for this pathology.

MRS in early clinical differential diagnosis.

**3.2 Conventional Magnetic Resonance** 

prognosis and choosing a treatment strategy.

al., 2008; Hirsch et al., 2007; Wilms et al., 2007).

accumulation.

allowing the acquisition of high quality, easily quantifiable spectra in acceptable imaging times (Di Costanzo et al., 2007).

The increase SNR associated with higher magnetic fields permits shorter imaging times for a given spatial resolution, higher resolution for a given imaging time or the combination of both.

The spectral resolution is linearly correlated with the field strength and is about twice at 3 Tesla as compared to 1.5 Tesla. Clinical 1.5 Tesla scanners equipped with 1H-MRS packages allow the quantification of NAA, Cho, Cr and lactate at long echo-time, and further metabolites, such as mI and glutamate-glutamine (Gxl), at short echo-time. mI is a strongly coupled system and resonates at four chemical shift positions. At 1.5 Tesla, only the singlet component at 3.57 ppm is detected. However, at 3 Tesla this resonance is resolved into its components at 3.55 and 3.61 ppm. Therefore by increasing of spectral resolution and SNR, the quantification precision of mI is significantly better at 3 Tesla relative to 1.5 Tesla (Srinivasan et al., 2004).

Despite shorter T2 relaxation times and increased field inhomogeneity, the chemical shift doubling at 3 Tesla yields better spectral resolution. This is reflected by improved baseline separation of Cho and Cr, which are only 0.2 parts per million (ppm) apart, and by slightly better resolution of Glu/Gln region, between 2.05 and 2.5 ppm, at shorter TE.

Higher field strengths also lead to a flatter baseline that contributes to more reliable estimation of peak area and, hence, more precise quantification, in addition to a more accurate identification of each metabolite.

This has been shown by a recent study applying multiple regional single voxel 1H-MRS including putamen, pontine basis and cerebral white matter at 3 Tesla in 24 patients with MSA compared to 11 PD patients and 18 controls. Significant NAA/Cr reductions have been shown in the pontine basis of both patients with MSA-C (cerebellar ataxia variant of MSA) and MSA-P, while putaminal NAA/Cr was only reduced in the patients with MSA-P. Eight of the 11 MSA-P patients compared to none of the PD and control group were classified correctly by combining individual NAA/Cr reductions in the pontine basis and in the putamen. These results suggest that combined assessment of NAA/Cr in the pontine basis and putamen may be effective in differentiating MSA-P from PD in terms of the high specificity of reduced NAA/Cr in the pontine basis or in the putamen in patients with MSA-P (Watanabe et al., 2004).

Moreover, in these studies, the metabolite concentrations were expressed in terms of semiquantitative ratios such as NAA/Cr, NAA/Cho, Cho/Cr and mI/Cr. In relative quantification, one of the metabolite peaks measured is used as the concentration standard and serves as the denominator of the peak ratios. As a result, the total number of quantifiable metabolites is decreased by one. Furthermore, alterations in the peak ratio do not necessarily reflect a change in the concentration of the numerator. The alteration may be caused by change in the concentration of the numerator, the denominator, or both or may be due to changes in relaxation behavior. The assumption that the concentration of certain reference metabolites (e.g. total creatine, choline) remains constant may be incorrect under normal conditions, as well as in many pathologic states. It is therefore advisable to obtain concentration expressed in standard units (such as millimoles per kilogram wet weight) by applying absolute quantification.

Combined 31P- and 1H-MRSI at 3 Tesla measuring absolute adenosine diphophosphate (ADP), ATP, Cr and PCr concentrations in two well-defined cohorts of patients with early and advanced PD has been performed to evaluate brain energy metabolism (Hattingen et

allowing the acquisition of high quality, easily quantifiable spectra in acceptable imaging

The increase SNR associated with higher magnetic fields permits shorter imaging times for a given spatial resolution, higher resolution for a given imaging time or the combination of

The spectral resolution is linearly correlated with the field strength and is about twice at 3 Tesla as compared to 1.5 Tesla. Clinical 1.5 Tesla scanners equipped with 1H-MRS packages allow the quantification of NAA, Cho, Cr and lactate at long echo-time, and further metabolites, such as mI and glutamate-glutamine (Gxl), at short echo-time. mI is a strongly coupled system and resonates at four chemical shift positions. At 1.5 Tesla, only the singlet component at 3.57 ppm is detected. However, at 3 Tesla this resonance is resolved into its components at 3.55 and 3.61 ppm. Therefore by increasing of spectral resolution and SNR, the quantification precision of mI is significantly better at 3 Tesla relative to 1.5 Tesla

Despite shorter T2 relaxation times and increased field inhomogeneity, the chemical shift doubling at 3 Tesla yields better spectral resolution. This is reflected by improved baseline separation of Cho and Cr, which are only 0.2 parts per million (ppm) apart, and by slightly

Higher field strengths also lead to a flatter baseline that contributes to more reliable estimation of peak area and, hence, more precise quantification, in addition to a more

This has been shown by a recent study applying multiple regional single voxel 1H-MRS including putamen, pontine basis and cerebral white matter at 3 Tesla in 24 patients with MSA compared to 11 PD patients and 18 controls. Significant NAA/Cr reductions have been shown in the pontine basis of both patients with MSA-C (cerebellar ataxia variant of MSA) and MSA-P, while putaminal NAA/Cr was only reduced in the patients with MSA-P. Eight of the 11 MSA-P patients compared to none of the PD and control group were classified correctly by combining individual NAA/Cr reductions in the pontine basis and in the putamen. These results suggest that combined assessment of NAA/Cr in the pontine basis and putamen may be effective in differentiating MSA-P from PD in terms of the high specificity of reduced NAA/Cr in the pontine basis or in the putamen in patients with MSA-

Moreover, in these studies, the metabolite concentrations were expressed in terms of semiquantitative ratios such as NAA/Cr, NAA/Cho, Cho/Cr and mI/Cr. In relative quantification, one of the metabolite peaks measured is used as the concentration standard and serves as the denominator of the peak ratios. As a result, the total number of quantifiable metabolites is decreased by one. Furthermore, alterations in the peak ratio do not necessarily reflect a change in the concentration of the numerator. The alteration may be caused by change in the concentration of the numerator, the denominator, or both or may be due to changes in relaxation behavior. The assumption that the concentration of certain reference metabolites (e.g. total creatine, choline) remains constant may be incorrect under normal conditions, as well as in many pathologic states. It is therefore advisable to obtain concentration expressed in standard units (such as millimoles per kilogram wet weight) by

Combined 31P- and 1H-MRSI at 3 Tesla measuring absolute adenosine diphophosphate (ADP), ATP, Cr and PCr concentrations in two well-defined cohorts of patients with early and advanced PD has been performed to evaluate brain energy metabolism (Hattingen et

better resolution of Glu/Gln region, between 2.05 and 2.5 ppm, at shorter TE.

times (Di Costanzo et al., 2007).

(Srinivasan et al., 2004).

P (Watanabe et al., 2004).

applying absolute quantification.

accurate identification of each metabolite.

both.

al., 2009). In the putamen and midbrain of both PD groups compared to control was found a bilateral reduction of high-energy phosphates such as ATP and PCr as final acceptors of energy from mitochondrial oxidative phosphorylation. In contrast, low-energy metabolites such as ADP and Pi were within normal ranges. Patients with early Parkinson's Disease, with clearly lateralized motor symptoms, exhibited a significant reduction of putamen highenergy phosphates in the less affected hemisphere with a less pronounced dopaminergic cell loss. Therefore, mitochondrial dysfunction is a rather early occurring and subsequently persistent event in the pathophysiology of dopaminergic degeneration in PD. These data strongly support the hypothesis that mitochondrial dysfunction is involved early in pathogenesis of PD and it may be used as early marker for this pathology.

In vivo MRS is increasingly utilized for the study of neurochemistry and cerebral energy metabolism in PD. Particularly, the recent technical advances of in vivo MRS including the availability of higher magnetic fields permitting improved spectral and spatial resolution, the development of a reliable method for absolute metabolite quantification, the development of various spectroscopic methods to enhance metabolite signal identification, and the application of combined 31P- and 1H-MRS can be use to examine the changes in neurochemical profile non–invasively and to achieve a differential diagnosis of PD versus other forms of parkinsonism, especially in early stages of disease when signs and symptoms of different forms of parkinsonism have greater overlap. However, several multicentre trials using a larger sample of patients, absolute quantification of tissue metabolite concentrations and a standardized technique are required to fully determine the place of MRS in early clinical differential diagnosis.

#### **3.2 Conventional Magnetic Resonance**

 In the early disease stages the clinical separation of atypical parkinsonism disorders (APD)s from PD carries a high rate of misdiagnosis. An early differentiation between APD and PD, each characterized by completely different natural histories, is crucial for determining the prognosis and choosing a treatment strategy.

The principles of MR imaging are based on the ubiquitous presence of hydrogen in body tissues and the spin of the hydrogen atom proton, which induces a small magnetic field. In general, T2- weighted sequences are sensitive to changes in tissue properties, including tissue damage, due to changes of the transverse magnetization or T2 decay. Neurodegenerative processes characterized by cell loss, increased age-related deposition of iron or other paramagnetic substances, and by astroglial reaction and microglial proliferation may lead to signal changes in affected brain areas, like the basal ganglia or infratentorial structures, in neurodegenerative parkinsonism (Duguid et al., 1986; Gupta et al., 2008; Hirsch et al., 2007; Wilms et al., 2007).

Because cMRI is believed to be usually normal in patients with PD, while it frequently shows characteristic abnormalities in patients with APD, cMRI images takes a major part in excluding underlying pathologies such as vascular lesions, multiple sclerosis, brain tumors, normal pressure hydrocephalus, bilateral striopallidodentate calcinosis, and other potential, but rare, causes of symptomatic parkinsonism such as Wilson disease, manganese-induced parkinsonism, or different subtypes of neurodegeneration associated with brain iron accumulation.

At 1,5 T, patients with advanced PD, and sometimes those with APD, may show distinct abnormalities of the substantia nigra, including signal increase on T2-weighted MR images, smudging of the hypointensity in the substantia nigra towards the red nucleus or signal loss

Early Marker for the Diagnosis of Parkinson's Disease 13

Fig. 3. Diagrams illustrate magnetization transfer, that is, the exchange of longitudinal magnetization between restricted protons associated with rigid macromolecules and free water protons. (a)Diagram shows macromolecular protons (H), including hydration layer protons and free water protons. (b) Off-resonance irradiation (arrows) saturates the

immobile macromolecular protons (unsatured protons are designated H, while satured are designated H). (c) Saturation is transferred to hydration layer protons (…H). (d) Satured protons diffuse into the free water proton pool and decrease the signal from this pool.

Misregistration can occur if the subject moves between the two scans, but the M0 and Ms images must be in register; otherwise, artifacts appear at the edges of features in the calculated MTR image, with false MTR values. It is best to acquire the two images in an interleaved way (Barker et al., 1996; Inglese at al., 2001) although it is possible to register

when using inversion recovery MRI (Brooks, 2000; Rutledge et al., 1987; Savoiardo et al., 1994).

Biochemical studies have reported increased iron content in the substantia nigra pars compacta (SNc) in PD, with changes most marked in severe disease, suggesting that measurement of nigral iron content may provide an indication of the pathologic severity of the disease (Youdim et al.,1990). Iron accumulates in the brain as a function of age, primarily in the form of ferritin and particularly in oligodendrocytes, but also in neurons and microglia. The adult brain has a very high iron content, particularly in the basal ganglia. Brain iron concentration is highest in the globus pallidus, substantia nigra, red nucleus, caudate, and putamen. Abnormally elevated iron levels are evident in various neurodegenerative disorders, including PD where there is evidence of increased iron in the substantia nigra (Dexter et al.,1989; Sofic et al.,1988). Signal changes on T2-weighted images in the basal ganglia as well as in infratentorial structures have been reported for all APDs at 1.5 T, where they have been used as a differentiating criterion from PD. Furthermore, estimation of transverse relaxation in patients with PD, using a 1.5 Tesla whole body imaging system, showed shortened T2 values in substantia nigra, caudate and putamen in PD patients as compared to healthy controls (Antinoni et al., 1993). These data do suggest a potential utility of these measurements as a biomarker of disease progression.

#### **3.3 Magnetization Transfer Imaging**

Standard MR imaging detects signal only from hydrogen nuclei (protons) that are ''mobile'' (contained within a liquid); if a hydrogen atom is part of a molecule that is large and cannot move about freely, the signal from that hydrogen atom decays too quickly to be seen using a clinical MR imaging scanner. Such protons are found in large molecules (macromolecules), such as those of cell membranes and myelin. The mobile protons are in constant motion, however, and come into regular and intimate contact with the macromolecular protons, and the spin state (the proton magnetization state, which is measured with MR imaging) of the mobile protons can exchange with that of the macromolecular protons. This exchange of magnetization forms the basis of magnetization transfer imaging (Horsfield, 2005). Magnetization transfer is a physical phenomenon that results from interactions and exchanges between magnetized protons in water that are unrestricted in their molecular motion and those that are restricted because of their association with macromolecules. The latter have a much shorter T2 relaxation time and broader resonance, which makes it possible to selectively saturate their magnetization with an appropriate off-resonance pulse. The acquisition of two images, one obtained with the magnetization transfer saturation pulse turned on and the other with it turned off, can be used to generate a magnetization transfer ratio (MTR) image in which the signal intensity of each voxel is determined by the percent magnetization transfer in that voxel.

A MTR image is calculated from a pair of images acquired in an identical way, except that one has extra off-resonance RF pulses applied, which saturates the macromolecular magnetization pool. The MTR is calculated for every corresponding pair of pixels in the two images. If the intensity of the pixel in the image without saturation pulses is M0 and the corresponding intensity in the image with saturation pulses is Ms, the MTR is as follows:

MTR= [(M0 - Ms)/M0]\* 100%

when using inversion recovery MRI (Brooks, 2000; Rutledge et al., 1987; Savoiardo et al.,

Biochemical studies have reported increased iron content in the substantia nigra pars compacta (SNc) in PD, with changes most marked in severe disease, suggesting that measurement of nigral iron content may provide an indication of the pathologic severity of the disease (Youdim et al.,1990). Iron accumulates in the brain as a function of age, primarily in the form of ferritin and particularly in oligodendrocytes, but also in neurons and microglia. The adult brain has a very high iron content, particularly in the basal ganglia. Brain iron concentration is highest in the globus pallidus, substantia nigra, red nucleus, caudate, and putamen. Abnormally elevated iron levels are evident in various neurodegenerative disorders, including PD where there is evidence of increased iron in the substantia nigra (Dexter et al.,1989; Sofic et al.,1988). Signal changes on T2-weighted images in the basal ganglia as well as in infratentorial structures have been reported for all APDs at 1.5 T, where they have been used as a differentiating criterion from PD. Furthermore, estimation of transverse relaxation in patients with PD, using a 1.5 Tesla whole body imaging system, showed shortened T2 values in substantia nigra, caudate and putamen in PD patients as compared to healthy controls (Antinoni et al., 1993). These data do suggest a

potential utility of these measurements as a biomarker of disease progression.

Standard MR imaging detects signal only from hydrogen nuclei (protons) that are ''mobile'' (contained within a liquid); if a hydrogen atom is part of a molecule that is large and cannot move about freely, the signal from that hydrogen atom decays too quickly to be seen using a clinical MR imaging scanner. Such protons are found in large molecules (macromolecules), such as those of cell membranes and myelin. The mobile protons are in constant motion, however, and come into regular and intimate contact with the macromolecular protons, and the spin state (the proton magnetization state, which is measured with MR imaging) of the mobile protons can exchange with that of the macromolecular protons. This exchange of magnetization forms the basis of magnetization transfer imaging (Horsfield, 2005). Magnetization transfer is a physical phenomenon that results from interactions and exchanges between magnetized protons in water that are unrestricted in their molecular motion and those that are restricted because of their association with macromolecules. The latter have a much shorter T2 relaxation time and broader resonance, which makes it possible to selectively saturate their magnetization with an appropriate off-resonance pulse. The acquisition of two images, one obtained with the magnetization transfer saturation pulse turned on and the other with it turned off, can be used to generate a magnetization transfer ratio (MTR) image in which the signal intensity of each voxel is determined by the

A MTR image is calculated from a pair of images acquired in an identical way, except that one has extra off-resonance RF pulses applied, which saturates the macromolecular magnetization pool. The MTR is calculated for every corresponding pair of pixels in the two images. If the intensity of the pixel in the image without saturation pulses is M0 and the corresponding intensity in the image with saturation pulses is Ms, the MTR is as

MTR= [(M0 - Ms)/M0]\* 100%

**3.3 Magnetization Transfer Imaging** 

percent magnetization transfer in that voxel.

follows:

1994).

Fig. 3. Diagrams illustrate magnetization transfer, that is, the exchange of longitudinal magnetization between restricted protons associated with rigid macromolecules and free water protons. (a)Diagram shows macromolecular protons (H), including hydration layer protons and free water protons. (b) Off-resonance irradiation (arrows) saturates the immobile macromolecular protons (unsatured protons are designated H, while satured are designated H). (c) Saturation is transferred to hydration layer protons (…H). (d) Satured protons diffuse into the free water proton pool and decrease the signal from this pool.

Misregistration can occur if the subject moves between the two scans, but the M0 and Ms images must be in register; otherwise, artifacts appear at the edges of features in the calculated MTR image, with false MTR values. It is best to acquire the two images in an interleaved way (Barker et al., 1996; Inglese at al., 2001) although it is possible to register

Early Marker for the Diagnosis of Parkinson's Disease 15

These studies show that MTR analysis may be a useful technique for PD diagnosis and

DWI imaging visualizes the random movement of water molecules in the tissue by applying diffusion-sensitizing gradients to assess changes in diffusion magnitude and orientation of water molecules in tissue. Quantification of the diffusivity is achieved by applying diffusion-sensitizing gradients of different degrees in 3 orthogonal directions and calculating the apparent diffusion coefficient (ADC) for each direction. The ADC is very

decrease in MTR probably begins previously than the clinical onset of the disease.

Fig. 4. Image-based visualization of diffusion tensor data. Top: Sphere representing directional color encoding. Second row: 2D images of image based visualization (from left:

FA image, mean diffusivity image, and color-encoded image).

**3.4 Diffusion-Weighted Imaging** 

dependent on the direction of diffusion encoding.

them after acquisition. Two forms of data analysis have been used extensively for MTR images: region of interest (ROI) and histogram analysis. ROI analysis may be useful for elucidating the degree of tissue damage within individual lesions seen on T2-weighted scans or within anatomic regions associated with particular symptoms. ROI analysis, however, can be subject to operator bias, because the placement of regions is normally done manually. This could be overcome by first registering scans to an anatomic template and using ROIs defined on the template image. With MTR histogram analysis, a histogram of pixel MTR values is formed from the whole of the brain parenchyma; thus, focal damage and more widespread diffuse tissue damage are reflected in changes to the shape of the histogram, with a general shift toward lower MTR values as the density of macromolecules is reduced with demyelination or axonal loss. Extraction of the brain parenchyma, using the same procedures that are used in atrophy measurements, is a necessary preprocessing step. After normalization (to remove any effect of the absolute brain size), the MTR histogram can be characterized by several simple statistics, such as the histogram peak position, the peak height, and the average MTR. The employment of off-resonance irradiation was first proposed by Wolff and Balaban (Wolff & Balaban, 1989), who found that use of an offresonance radio-frequency preparation pulse could generate excellent tissue contrast in images of rabbit kidney, and they referred to the technique as "magnetization transfer contrast." The initial magnetization transfer occurs between the macro-molecular protons and the transiently hound hydration layer protons. The efficiency of this interaction is directly related to the number of irradiation sites (hydrogen bonds) and their mobility. The utilization of magnetization transfer was extended to clinical imaging, including its use with gradient-echo imaging and MR angiography (Wolff et al.,1991; Pike et al., 1992). A decrease in the MTR, which reflects a reduction in the exchange of magnetization of protons that are tumbling freely and those that are bound to macromolecules, is evidence of demyelination in cerebral white matter. MTR imaging is sensitive to both microscopic and macroscopic pathology and provides quantitative data on the extent of myelin loss in MS.

By using MTI, abnormalities of the basal ganglia and SN have been reported in patients with PD, MSA and PSP. One study (Eckert et al., 2004) investigated the potential of MT imaging in the differential diagnosis of neurodegenerative parkinsonism, including 37 patients with different parkinsonian syndromes and 20 age-matched controls. The main finding in this study was a change in the MTR in the globus pallidus, putamen, caudate nucleus, SN, and white matter in PD, MSA, and PSP patients, matching the pathologic features of the underlying disorder. MTR were significantly reduced in the putamen in MSA patients compared with PD patients and healthy controls, as well as in the SN in patients with PSP, MSA, and PD. Another study (Yonca et al., 2007) determinated the role of MTR in the early period of 33 patients with PD, comparing the findings with those in 30 normal healthy volunteers. Signal intensity measurements were obtained from 15 anatomic regions: SNc, substantia nigra pars reticulate (SNPR), red nucleus, dentate nucleus, cerebellum, pons, globus pallidus, putamen, caudate nucleus, thalamus, internal capsule posterior horn, forceps major, forceps minor, genu, and splenium of corpus callosum. Results showed a significant decrease of MTR in the SNc, SNPR, red nucleus, and pons compared with normal healthy volunteers. No significant decrease in MTR were found at supratentorial paraventricular white matter and cerebellum, which may be attributed the duration of the disease.

Perhaps in the initial stages of PD, supratentorial paraventricular white matter is not influenced by the disease. The decrease of MTR at SNc, SNPR, red nucleus, and pons in PD patients can be attributed to neurodegeneration (Tambasco et al., 2003)).

These studies show that MTR analysis may be a useful technique for PD diagnosis and decrease in MTR probably begins previously than the clinical onset of the disease.

#### **3.4 Diffusion-Weighted Imaging**

14 Diagnostics and Rehabilitation of Parkinson's Disease

them after acquisition. Two forms of data analysis have been used extensively for MTR images: region of interest (ROI) and histogram analysis. ROI analysis may be useful for elucidating the degree of tissue damage within individual lesions seen on T2-weighted scans or within anatomic regions associated with particular symptoms. ROI analysis, however, can be subject to operator bias, because the placement of regions is normally done manually. This could be overcome by first registering scans to an anatomic template and using ROIs defined on the template image. With MTR histogram analysis, a histogram of pixel MTR values is formed from the whole of the brain parenchyma; thus, focal damage and more widespread diffuse tissue damage are reflected in changes to the shape of the histogram, with a general shift toward lower MTR values as the density of macromolecules is reduced with demyelination or axonal loss. Extraction of the brain parenchyma, using the same procedures that are used in atrophy measurements, is a necessary preprocessing step. After normalization (to remove any effect of the absolute brain size), the MTR histogram can be characterized by several simple statistics, such as the histogram peak position, the peak height, and the average MTR. The employment of off-resonance irradiation was first proposed by Wolff and Balaban (Wolff & Balaban, 1989), who found that use of an offresonance radio-frequency preparation pulse could generate excellent tissue contrast in images of rabbit kidney, and they referred to the technique as "magnetization transfer contrast." The initial magnetization transfer occurs between the macro-molecular protons and the transiently hound hydration layer protons. The efficiency of this interaction is directly related to the number of irradiation sites (hydrogen bonds) and their mobility. The utilization of magnetization transfer was extended to clinical imaging, including its use with gradient-echo imaging and MR angiography (Wolff et al.,1991; Pike et al., 1992). A decrease in the MTR, which reflects a reduction in the exchange of magnetization of protons that are tumbling freely and those that are bound to macromolecules, is evidence of demyelination in cerebral white matter. MTR imaging is sensitive to both microscopic and macroscopic

pathology and provides quantitative data on the extent of myelin loss in MS.

be attributed the duration of the disease.

By using MTI, abnormalities of the basal ganglia and SN have been reported in patients with PD, MSA and PSP. One study (Eckert et al., 2004) investigated the potential of MT imaging in the differential diagnosis of neurodegenerative parkinsonism, including 37 patients with different parkinsonian syndromes and 20 age-matched controls. The main finding in this study was a change in the MTR in the globus pallidus, putamen, caudate nucleus, SN, and white matter in PD, MSA, and PSP patients, matching the pathologic features of the underlying disorder. MTR were significantly reduced in the putamen in MSA patients compared with PD patients and healthy controls, as well as in the SN in patients with PSP, MSA, and PD. Another study (Yonca et al., 2007) determinated the role of MTR in the early period of 33 patients with PD, comparing the findings with those in 30 normal healthy volunteers. Signal intensity measurements were obtained from 15 anatomic regions: SNc, substantia nigra pars reticulate (SNPR), red nucleus, dentate nucleus, cerebellum, pons, globus pallidus, putamen, caudate nucleus, thalamus, internal capsule posterior horn, forceps major, forceps minor, genu, and splenium of corpus callosum. Results showed a significant decrease of MTR in the SNc, SNPR, red nucleus, and pons compared with normal healthy volunteers. No significant decrease in MTR were found at supratentorial paraventricular white matter and cerebellum, which may

Perhaps in the initial stages of PD, supratentorial paraventricular white matter is not influenced by the disease. The decrease of MTR at SNc, SNPR, red nucleus, and pons in PD

patients can be attributed to neurodegeneration (Tambasco et al., 2003)).

DWI imaging visualizes the random movement of water molecules in the tissue by applying diffusion-sensitizing gradients to assess changes in diffusion magnitude and orientation of water molecules in tissue. Quantification of the diffusivity is achieved by applying diffusion-sensitizing gradients of different degrees in 3 orthogonal directions and calculating the apparent diffusion coefficient (ADC) for each direction. The ADC is very dependent on the direction of diffusion encoding.

Fig. 4. Image-based visualization of diffusion tensor data. Top: Sphere representing directional color encoding. Second row: 2D images of image based visualization (from left: FA image, mean diffusivity image, and color-encoded image).

Early Marker for the Diagnosis of Parkinson's Disease 17

Kendi et al., 2008). These results confirm that the neurodegenerative process extends beyond

Olfactory impairment, which is common in PD and often predates clinical diagnosis, may be a useful biomarker for early PD. One study (Rolheiser et al., 2011) compared newly diagnosed PD patients with a matched control group using both olfactory testing and diffusion tensor imaging of the substantia nigra and anterior olfactory structures. Fourteen PD patients with stage 1-2 of Hoehn & Yahr were matched to a control group by age and sex. All subjects completed the University of Pennsylvania Smell Identification Test, as well as a series of MRI scans designed to examine diffusion characteristics of the olfactory tract and the substantia nigra. Olfactory testing revealed significant impairment in the patient group. Diffusion tensor imaging revealed significant group differences in both the substantia nigra and anterior olfactory region, with fractional anisotropy of the olfactory region clearly distinguishing the Parkinson's subjects from controls. This study has suggested that there may be value in combining behavioral (olfaction) and MRI testing to

identify early Parkinson's disease (Rolheiser et al., 2011; Fulton & Barret, 2008).

available on whole body MR scanners and can be acquired within a few minutes.

Concluding DWI/DTI imaging especially bears several advantages. DWI/DTI imaging may detect diffusion abnormalities in the basal ganglia and infratentorial structures in patients with PD at an early stage of disease. Furthermore, DWI/DTI imaging sequences are widely

fMRI is based on the increase in blood flow to the local vasculature that accompanies neural activity in the brain. This results in a corresponding local reduction in deoxyhemoglobin because the increase in blood flow occurs without an increase of similar magnitude in oxygen extraction (Roy & Sherrington, 1890; Fox & Raichle, 1985). Since deoxyhemoglobin is paramagnetic, it alters the T2 weighted magnetic resonance image signal (Ogawa et al, 1990). Thus, deoxyhemoglobin is sometimes referred to as an endogenous contrast enhancing agent, and serves as the source of the signal for fMRI. Using an appropriate imaging sequence, human cortical functions can be observed without the use of exogenous contrast enhancing agents on a clinical strength (1.5 T) scanner (Bandettini et al., 1992, 1993; Kwong, et al, 1992; and Turner, et al, 1993; Schneider et al, 1993). Functional activity of the brain determined from the magnetic resonance signal has confirmed known anatomically distinct processing areas in the visual cortex (Belliveau, et al, 1991; Ogawa, et al, 1992; Schneider, et al, 1993), the motor cortex, and Broca's area of speech and language-related activities (Hinke et al., 1993; Kim et al., 1995). Further, a rapidly emerging body of literature documents corresponding findings between fMRI and conventional electrophysiological techniques to localize specific functions of the human brain (Atlas et al., 1996; Puce, et al, 1995; Burgess, 1995; Detre, et al, 1995; George, et al, 1995; Ives, et al, 1993). Consequently, the number of medical and research centers with fMRI capabilities and investigational

The main advantages to fMRI as a technique to image brain activity related to a specific task

the in-plane resolution of the functional image is generally about 1.5 x 1.5 mm although

the signal does not require injections of radioactive isotopes

the total scan time required can be very short

resolutions less than 1 mm are possible.

the basal ganglia in PD (Tessa et al., 2008).

**3.5 Functional Magnetic Resonance Imaging** 

programs continues to escalate.

or sensory process include:

The random translational motion (diffusion) of water molecules in tissue is restricted by the highly organized architecture of fiber tracts in the central nervous system. Neuronal loss and gliosis disrupt this architecture, resulting in an increase of diffusivity and ADC. The complex neuronal architecture with its organization in fiber bundles that are surrounded by dense myelin sheaths leads also to a distinct anisotropy of water diffusion, which is facilitated along the direction of fiber tracts and restricted perpendicular to the fibers.

The degree of anisotropy can be quantified by applying diffusion-sensitizing gradients in at least 6 directions, which permits calculation of fractional anisotropy (FA). Decreased FA values represent tissue degeneration due to normal aging or due to pathologic reasons such as neurodegeneration. Both diffusivity and FA can be combined to form the so-called diffusion tensor, which indicates direction and extent of diffusivity with the help of a vector (Hagmann et al., 2006; Le Bihan, 2003; Schocke et al., 2004). The central nervous system (CNS) is highly organized in numerous tracts of myelinated fibre bundles, whereby the movement of the water molecules is restricted perpendicular to these fibre bundles. The resulting anisotropic diffusion is quantified by the FA, which is determined by diffusionsensitised gradients in at least six directions. Both the diffusivity and the FA form the diffusion tensor (Le Bihan, 2003).

Widespread cerebral changes are observed in advanced stages of PD, suggesting that PD is a multisystem disorder.

Recently, several studies pointed out the capability of the histogram analysis of the apparent diffusion coefficient computed from diffusion-weighted images and of the mean diffusivity and FA computed from DTI to reveal brain-tissue damage in early clinical stages of neurodegenerative diseases. A recent study including 27 patients with de novo drug-naïve PD hypothesized that global measurements of brain volume and structure, such those possible with SIENAX software (part of FSL 4.0 http://www.fmrib.ox.ac.uk/fsl/) and histogram analysis of DTI could reveal subtle tissue changes in the early clinical phase of PD (Tessa et al., 2008). Accordingly, a group of patients with drug-naive de novo PD and a group of 16 healthy controls, were investigated with SIENAX and DTI. Results showed no significant differences for total brain, GM, and WM volumes and histogram-derived mean diffusivity metrics between controls and the whole group of patients with PD or any subgroup of patients with PD. As compared with controls, patients with PD as a whole and patients with the akinetic-rigid type showed an increase of the twenty-fifth percentile of the FA histogram. In patients with the akinetic-rigid type, there also was a trend toward an increase of the mean and fiftieth and seventy-fifth percentiles, and a reduction of the skewness of the FA histogram. This finding is consistent with the hypothesis that subtle GM loss is present in patients with PD since the early clinical phases and that this feature is more pronounced in patients with akinetic-rigid type. Another recent study including only patients with newly diagnosed PD used high-resolution DTI at 3 Tesla to evaluate rostral, middle, and caudal ROIs within the SN on a single slice of the midbrain and this study found that PD patients could be completely separated from the control group based on reduced FA values in the caudal ROI of the SN, such that further confirmatory studies seem to warrant. By using statistical parametric mapping analysis of DT imaging, changes in FA were found in the frontal lobes, including the supplementary motor area, the presupplementary motor area, and the cingulum in non demented PD patients relative to controls, whereas VBM analysis in the same patients revealed no volume loss (Karagulle

The random translational motion (diffusion) of water molecules in tissue is restricted by the highly organized architecture of fiber tracts in the central nervous system. Neuronal loss and gliosis disrupt this architecture, resulting in an increase of diffusivity and ADC. The complex neuronal architecture with its organization in fiber bundles that are surrounded by dense myelin sheaths leads also to a distinct anisotropy of water diffusion, which is

The degree of anisotropy can be quantified by applying diffusion-sensitizing gradients in at least 6 directions, which permits calculation of fractional anisotropy (FA). Decreased FA values represent tissue degeneration due to normal aging or due to pathologic reasons such as neurodegeneration. Both diffusivity and FA can be combined to form the so-called diffusion tensor, which indicates direction and extent of diffusivity with the help of a vector (Hagmann et al., 2006; Le Bihan, 2003; Schocke et al., 2004). The central nervous system (CNS) is highly organized in numerous tracts of myelinated fibre bundles, whereby the movement of the water molecules is restricted perpendicular to these fibre bundles. The resulting anisotropic diffusion is quantified by the FA, which is determined by diffusionsensitised gradients in at least six directions. Both the diffusivity and the FA form the

Widespread cerebral changes are observed in advanced stages of PD, suggesting that PD is a

Recently, several studies pointed out the capability of the histogram analysis of the apparent diffusion coefficient computed from diffusion-weighted images and of the mean diffusivity and FA computed from DTI to reveal brain-tissue damage in early clinical stages of neurodegenerative diseases. A recent study including 27 patients with de novo drug-naïve PD hypothesized that global measurements of brain volume and structure, such those possible with SIENAX software (part of FSL 4.0 http://www.fmrib.ox.ac.uk/fsl/) and histogram analysis of DTI could reveal subtle tissue changes in the early clinical phase of PD (Tessa et al., 2008). Accordingly, a group of patients with drug-naive de novo PD and a group of 16 healthy controls, were investigated with SIENAX and DTI. Results showed no significant differences for total brain, GM, and WM volumes and histogram-derived mean diffusivity metrics between controls and the whole group of patients with PD or any subgroup of patients with PD. As compared with controls, patients with PD as a whole and patients with the akinetic-rigid type showed an increase of the twenty-fifth percentile of the FA histogram. In patients with the akinetic-rigid type, there also was a trend toward an increase of the mean and fiftieth and seventy-fifth percentiles, and a reduction of the skewness of the FA histogram. This finding is consistent with the hypothesis that subtle GM loss is present in patients with PD since the early clinical phases and that this feature is more pronounced in patients with akinetic-rigid type. Another recent study including only patients with newly diagnosed PD used high-resolution DTI at 3 Tesla to evaluate rostral, middle, and caudal ROIs within the SN on a single slice of the midbrain and this study found that PD patients could be completely separated from the control group based on reduced FA values in the caudal ROI of the SN, such that further confirmatory studies seem to warrant. By using statistical parametric mapping analysis of DT imaging, changes in FA were found in the frontal lobes, including the supplementary motor area, the presupplementary motor area, and the cingulum in non demented PD patients relative to controls, whereas VBM analysis in the same patients revealed no volume loss (Karagulle

facilitated along the direction of fiber tracts and restricted perpendicular to the fibers.

diffusion tensor (Le Bihan, 2003).

multisystem disorder.

Kendi et al., 2008). These results confirm that the neurodegenerative process extends beyond the basal ganglia in PD (Tessa et al., 2008).

Olfactory impairment, which is common in PD and often predates clinical diagnosis, may be a useful biomarker for early PD. One study (Rolheiser et al., 2011) compared newly diagnosed PD patients with a matched control group using both olfactory testing and diffusion tensor imaging of the substantia nigra and anterior olfactory structures. Fourteen PD patients with stage 1-2 of Hoehn & Yahr were matched to a control group by age and sex. All subjects completed the University of Pennsylvania Smell Identification Test, as well as a series of MRI scans designed to examine diffusion characteristics of the olfactory tract and the substantia nigra. Olfactory testing revealed significant impairment in the patient group. Diffusion tensor imaging revealed significant group differences in both the substantia nigra and anterior olfactory region, with fractional anisotropy of the olfactory region clearly distinguishing the Parkinson's subjects from controls. This study has suggested that there may be value in combining behavioral (olfaction) and MRI testing to identify early Parkinson's disease (Rolheiser et al., 2011; Fulton & Barret, 2008).

Concluding DWI/DTI imaging especially bears several advantages. DWI/DTI imaging may detect diffusion abnormalities in the basal ganglia and infratentorial structures in patients with PD at an early stage of disease. Furthermore, DWI/DTI imaging sequences are widely available on whole body MR scanners and can be acquired within a few minutes.

#### **3.5 Functional Magnetic Resonance Imaging**

fMRI is based on the increase in blood flow to the local vasculature that accompanies neural activity in the brain. This results in a corresponding local reduction in deoxyhemoglobin because the increase in blood flow occurs without an increase of similar magnitude in oxygen extraction (Roy & Sherrington, 1890; Fox & Raichle, 1985). Since deoxyhemoglobin is paramagnetic, it alters the T2 weighted magnetic resonance image signal (Ogawa et al, 1990). Thus, deoxyhemoglobin is sometimes referred to as an endogenous contrast enhancing agent, and serves as the source of the signal for fMRI. Using an appropriate imaging sequence, human cortical functions can be observed without the use of exogenous contrast enhancing agents on a clinical strength (1.5 T) scanner (Bandettini et al., 1992, 1993; Kwong, et al, 1992; and Turner, et al, 1993; Schneider et al, 1993). Functional activity of the brain determined from the magnetic resonance signal has confirmed known anatomically distinct processing areas in the visual cortex (Belliveau, et al, 1991; Ogawa, et al, 1992; Schneider, et al, 1993), the motor cortex, and Broca's area of speech and language-related activities (Hinke et al., 1993; Kim et al., 1995). Further, a rapidly emerging body of literature documents corresponding findings between fMRI and conventional electrophysiological techniques to localize specific functions of the human brain (Atlas et al., 1996; Puce, et al, 1995; Burgess, 1995; Detre, et al, 1995; George, et al, 1995; Ives, et al, 1993). Consequently, the number of medical and research centers with fMRI capabilities and investigational programs continues to escalate.

The main advantages to fMRI as a technique to image brain activity related to a specific task or sensory process include:


Early Marker for the Diagnosis of Parkinson's Disease 19

In a study combining fMRI and OERPs analysis in patients with PD, non-detectable OERPs patients exhibited reduced activity in the anterior cingulate gyrus and portions of the left striatum, while detectable ERP patients exhibited higher activation, especially in the amygdala, parahippocampal cortex, inferior frontal gyrus, insula, cingulate gyrus, striatum, and inferior temporal gyrus. The relationship between the expression of olfactory ERPs and cortical activation patterns seen during olfactory stimulation in fMRI in PD patients supports the idea that OERPs are a sensitive marker of neurodegeneration in olfactory regions, independent of the typically observed nigro-striatal degeneration in PD (Welge-

Olfactory dysfunction is more common in PD compared to atypical parkinsonian syndromes like PSP or MSA (Doty, 1991, 1993; Wenning et al., 1995). In a study including 37 patients with PD (Hoehn and Yahr I to IV) and 13 patients with MSA, CBD or PSP, 86 % of PD patients showed diminished sense of smell, or severe hyposmia, and 14 % were found to have moderate hyposmia, whereas 70 % of the patients with atypical parkinsonian syndromes exhibited moderate to mild hyposmia and 30 % normosmia (Muller et al., 2002). Olfactory testing may be an additive, helpful and inexpensive diagnostic instrument to support the discrimination of PD from healthy subjects and atypical parkinsonian

For the clinical assessment of olfactory function, several validated psychophysical tests exist. The best-validated olfactory tests include the University of Pennsylvania Smell Identification Test, the Connecticut Chemosensory Clinical Research Center Test and the Sniffin' Sticks Test (Cain et al., 1988; Doty et al., 1984; Hummel, 1997, 2007; Kobal et al., 2000). The Sniffin' Sticks is based on pen-like odor dispensing devices. It consists of three tests namely for odor threshold, discrimination and identification, the sum of which is defined as "TDI score". This score can give an indication of patient's olfactory performance

A useful help for the clinical diagnosis of olfactory deficits is represented by system using human electro-physiological methods such as OERPs that requests an adequate methods to

Based on the principles of air-dilution olfactometry, Kobal and Platting introduced a chemosensory stimulation with stimuli having a rectangular shape with rapid onset, precisely controlled in terms of timing, duration, intensity, not simultaneously activating other sensory systems (Kobal & Platting, 1978). This can be achieved by the olfactometer which is a complex instrument for creation of well defined, reproducible smell or pain

In conclusion the detection of early olfactory dysfunction, less frequent in other form of parkinsonism, can be used to assess risk for developing PD in asymptomatic subjects.

The defining features of PD are characterized by their insidious onset and inexorable but variable progression. Reliable and well validated early markers for PD to identify individuals "at risk" before motor and non motor symptoms, accurately diagnose individuals at the threshold of clinical PD, and monitor PD progression throughout its course would dramatically improve patient care and accelerate research into both PD cause and therapeutics. During the past two decades, much progress has been made in identifying and assessing PD markers, but as yet, no fully validated marker for PD is available.

(normosmia: TDI≥30.5, hyposmia: TDI≤30.5, functional anosmia: TDI≤16.5).

produce a selective and controlled stimulation of the olfactory system.

stimuli in the nose without tactile or thermal stimulation.

Lüssen et al., 2009).

**5. Conclusions** 

syndromes, before onset of motor symptoms.

The function or dysfunction of the several cortical regions involved in many disease, like PD, can be investigated in vivo by means of functional imaging techniques such as fMRI.

## **4. Olfactory dysfunction as a early diagnostic marker for Parkinson's Disease**

Olfactory dysfunction is a frequent non-motor symptom in PD and may be considered as an early clinical feature of the disease preceding motor symptoms by years (Ansari & Johnson, 1975). More than 96% of patients with PD present with olfactory dysfunction, compared with an established olfactory loss of at least 25% in the normal population over 52 years of age (Haehner et al., 2009). The majority of PD patients are functionally anosmic or severely hyposmic. Several studies have demonstrated an absence of correlation between olfactory loss and both duration of disease (Doty et al., 1988; Hawkes et al., 1997) and the clinical severity of PD (Ramaker et al., 2002), while other studies have found a correlation between the severity of PD and certain measures of olfactory function, such as latencies of olfactory event-related potentials (OERPs) (Hummel, 1999) and results from an odor discrimination task (Tissingh et al., 2001).

The cause of hyposmia in PD is not yet fully understood. It has been proposed that the develop of inclusion bodies, starting from the medulla oblongata and the anterior olfactory nucleus before the involvement of other central nervous structures, constitutes the reason of olfactory impairment before the motor symptoms appearance (Braak et al., 2003).

Moreover olfactory loss in PD is not a primary consequence of damage to the olfactory epithelium but rather result from distinct CNS abnormalities (Hummel et al., 2010). Studies based on biopsies from the olfactory epithelium did not reveal specific changes in the nasal mucosa of PD patients compared to patients who were hyposmic for other reasons (rhinitis, smoking or toxic agents). With regard to volumetrics of the olfactory bulb (OB) results indicated that there is little or no difference between PD patients with anosmia/hyposmia and healthy normosmic controls in terms of OB volume (Huisman et al., 2004; Hummel et al., 2010; Müller et al., 2005). Support for these results has come from a study that found an increase of (inhibitory) dopaminergic neurons in the OB in PD patients (Huisman et al., 2004). These findings have been interpreted within the context of a possible compensatory mechanism in response to the loss of dopaminergic neurons in the basal ganglia.

While cardinal motor symptoms in PD are closely related to a severe loss of dopaminergic cells in the nigro-striatal pathway, early clinical features such as olfactory impairment are more likely to be associated with extranigral pathology. Indeed atrophy in olfactory regions of the limbic and paralimbic cortex in early PD patients was found (Wattendorf et al., 2009). Moreover fMRI in PD patients indicated altered neuronal activity in the amygdaloid complex and hippocampal formation during olfactory stimulation (Takeda et al., 2010; Welge-Lüssen et al., 2009; Westermann et al., 2008). In addition, neuronal activity in components of cortico-striatal loops appears to be up-regulated indicating compensatory processes involving the dopaminergic system (Westermann et al., 2008).

Changes in olfactory function can also be observed using electrophysiological techniques such as recording OERPs (Kobal & Pattig, 1978). OERPs are the result of the sequential activation of numerous brain areas, starting with amygdala and regions of medial temporal lobe followed by the mid-orbito-frontal cortex and insular cortex, along with regions of the temporal lobe (Kettenmann et al., 1997). In PD patients OERPs are typically strongly delayed or even absent (Hawkes et al., 1999).

The function or dysfunction of the several cortical regions involved in many disease, like PD, can be investigated in vivo by means of functional imaging techniques such as fMRI.

**4. Olfactory dysfunction as a early diagnostic marker for Parkinson's Disease**  Olfactory dysfunction is a frequent non-motor symptom in PD and may be considered as an early clinical feature of the disease preceding motor symptoms by years (Ansari & Johnson, 1975). More than 96% of patients with PD present with olfactory dysfunction, compared with an established olfactory loss of at least 25% in the normal population over 52 years of age (Haehner et al., 2009). The majority of PD patients are functionally anosmic or severely hyposmic. Several studies have demonstrated an absence of correlation between olfactory loss and both duration of disease (Doty et al., 1988; Hawkes et al., 1997) and the clinical severity of PD (Ramaker et al., 2002), while other studies have found a correlation between the severity of PD and certain measures of olfactory function, such as latencies of olfactory event-related potentials (OERPs) (Hummel, 1999) and results from an odor discrimination

The cause of hyposmia in PD is not yet fully understood. It has been proposed that the develop of inclusion bodies, starting from the medulla oblongata and the anterior olfactory nucleus before the involvement of other central nervous structures, constitutes the reason of

Moreover olfactory loss in PD is not a primary consequence of damage to the olfactory epithelium but rather result from distinct CNS abnormalities (Hummel et al., 2010). Studies based on biopsies from the olfactory epithelium did not reveal specific changes in the nasal mucosa of PD patients compared to patients who were hyposmic for other reasons (rhinitis, smoking or toxic agents). With regard to volumetrics of the olfactory bulb (OB) results indicated that there is little or no difference between PD patients with anosmia/hyposmia and healthy normosmic controls in terms of OB volume (Huisman et al., 2004; Hummel et al., 2010; Müller et al., 2005). Support for these results has come from a study that found an increase of (inhibitory) dopaminergic neurons in the OB in PD patients (Huisman et al., 2004). These findings have been interpreted within the context of a possible compensatory

While cardinal motor symptoms in PD are closely related to a severe loss of dopaminergic cells in the nigro-striatal pathway, early clinical features such as olfactory impairment are more likely to be associated with extranigral pathology. Indeed atrophy in olfactory regions of the limbic and paralimbic cortex in early PD patients was found (Wattendorf et al., 2009). Moreover fMRI in PD patients indicated altered neuronal activity in the amygdaloid complex and hippocampal formation during olfactory stimulation (Takeda et al., 2010; Welge-Lüssen et al., 2009; Westermann et al., 2008). In addition, neuronal activity in components of cortico-striatal loops appears to be up-regulated indicating compensatory

Changes in olfactory function can also be observed using electrophysiological techniques such as recording OERPs (Kobal & Pattig, 1978). OERPs are the result of the sequential activation of numerous brain areas, starting with amygdala and regions of medial temporal lobe followed by the mid-orbito-frontal cortex and insular cortex, along with regions of the temporal lobe (Kettenmann et al., 1997). In PD patients OERPs are typically strongly

olfactory impairment before the motor symptoms appearance (Braak et al., 2003).

mechanism in response to the loss of dopaminergic neurons in the basal ganglia.

processes involving the dopaminergic system (Westermann et al., 2008).

delayed or even absent (Hawkes et al., 1999).

task (Tissingh et al., 2001).

In a study combining fMRI and OERPs analysis in patients with PD, non-detectable OERPs patients exhibited reduced activity in the anterior cingulate gyrus and portions of the left striatum, while detectable ERP patients exhibited higher activation, especially in the amygdala, parahippocampal cortex, inferior frontal gyrus, insula, cingulate gyrus, striatum, and inferior temporal gyrus. The relationship between the expression of olfactory ERPs and cortical activation patterns seen during olfactory stimulation in fMRI in PD patients supports the idea that OERPs are a sensitive marker of neurodegeneration in olfactory regions, independent of the typically observed nigro-striatal degeneration in PD (Welge-Lüssen et al., 2009).

Olfactory dysfunction is more common in PD compared to atypical parkinsonian syndromes like PSP or MSA (Doty, 1991, 1993; Wenning et al., 1995). In a study including 37 patients with PD (Hoehn and Yahr I to IV) and 13 patients with MSA, CBD or PSP, 86 % of PD patients showed diminished sense of smell, or severe hyposmia, and 14 % were found to have moderate hyposmia, whereas 70 % of the patients with atypical parkinsonian syndromes exhibited moderate to mild hyposmia and 30 % normosmia (Muller et al., 2002).

Olfactory testing may be an additive, helpful and inexpensive diagnostic instrument to support the discrimination of PD from healthy subjects and atypical parkinsonian syndromes, before onset of motor symptoms.

For the clinical assessment of olfactory function, several validated psychophysical tests exist. The best-validated olfactory tests include the University of Pennsylvania Smell Identification Test, the Connecticut Chemosensory Clinical Research Center Test and the Sniffin' Sticks Test (Cain et al., 1988; Doty et al., 1984; Hummel, 1997, 2007; Kobal et al., 2000). The Sniffin' Sticks is based on pen-like odor dispensing devices. It consists of three tests namely for odor threshold, discrimination and identification, the sum of which is defined as "TDI score". This score can give an indication of patient's olfactory performance (normosmia: TDI≥30.5, hyposmia: TDI≤30.5, functional anosmia: TDI≤16.5).

A useful help for the clinical diagnosis of olfactory deficits is represented by system using human electro-physiological methods such as OERPs that requests an adequate methods to produce a selective and controlled stimulation of the olfactory system.

Based on the principles of air-dilution olfactometry, Kobal and Platting introduced a chemosensory stimulation with stimuli having a rectangular shape with rapid onset, precisely controlled in terms of timing, duration, intensity, not simultaneously activating other sensory systems (Kobal & Platting, 1978). This can be achieved by the olfactometer which is a complex instrument for creation of well defined, reproducible smell or pain stimuli in the nose without tactile or thermal stimulation.

In conclusion the detection of early olfactory dysfunction, less frequent in other form of parkinsonism, can be used to assess risk for developing PD in asymptomatic subjects.

### **5. Conclusions**

The defining features of PD are characterized by their insidious onset and inexorable but variable progression. Reliable and well validated early markers for PD to identify individuals "at risk" before motor and non motor symptoms, accurately diagnose individuals at the threshold of clinical PD, and monitor PD progression throughout its course would dramatically improve patient care and accelerate research into both PD cause and therapeutics. During the past two decades, much progress has been made in identifying and assessing PD markers, but as yet, no fully validated marker for PD is available.

Early Marker for the Diagnosis of Parkinson's Disease 21

Brooks, D.J. Morphological and functional imaging studies on the diagnosis and progression

Cain, W.S.; Gent, J.F.; Goodspeed, R.B. & Leonard, G. Evaluation of olfactory dysfunction in

Clark, J.B. N-acetylaspartate: a marker for neuronal loss or mitochondrial dysfunction.

Clarke, C.E. & Lowry, M. Systematic review of proton magnetic resonance spectroscopy of

Davie, C.A.; Wenning, G.K.; Barker, G.J.; Tofts, P.S.; Kendall, B.E. et al. Differentiation of

Dexter. D.T.; Wells, F.R.; Lees, A.J.; Agid, F.; Agid, Y. et al. Increased nigral iron content and

Doty, R.L.; Deems, D. & Steller, S. Olfactory dysfunction in Parkinson's disease: a general

Doty, R.L.; Golbe, L.I.; McKeown, D.A.; Stern, M.B.; Lehrach, C.M. et al. Olfactory testing

disease. *Neurology*, Vol.43, No.5, (May 1993), pp. 962-965, ISSN 0028-3878 Doty, R.L.; Perl, D.P.; Steele, J.C.; Chen; K.M.; Pierce, J.D. Jr. et al. Olfactory dysfunction in

Doty, R.L.; Shaman, P.; Kimmelmann, C.P. & Dann, M.S. University of Pennsylvania Smell

Eckert, T.; Sailer, M.; Kaufmann, J.; Schrader, C.; Peschel, T. et al. Differentiation of

Edelman, R.R.; Ahn, S.S.; Chien, D.; Li, W.; Goldmann, A. et al. Improved time-of-flight MR

*Laryngoscope*, Vol.94, No.2, (February 1984), pp. 176-178, ISSN 1531-4995 Duguid, JR.; De La, P.R. & DeGroot, J. Magnetic resonance imaging of the midbrain in

*Neurochemistry,* Vol.52, No6, (June 1989), pp. 1830-6, ISSN 0022-3042 Di Costanzo, A.; Trojsi, F.; Tosetti, M.; Schirmer, T.; Lechner, S.M. et al. Proton MR

Vol.38, No.8, (August 1988), pp. 1237-1244, ISSN 0028-3878

Vol.21, No.1, (January 2004), pp. 229-235, ISSN 1053-8119

No.2, (August 1992), pp. 395-399, ISSN 0033-8419

Vol.98, No.1, (January 1988), pp. 83-88, ISSN 1531-4995

(November 2001), pp. 573-577, ISSN 1351-5101

1995), pp. 204-210, ISSN 0364-5134

2007), pp. 1651-1662, ISSN 0938-7994

ISSN 0016-867X

747, ISSN 0364-5134

ISSN 0340-5354

0378-5866

of Parkinson's disease. *Journal of Neurololy,* Vol. 247, No2, (April 2000), pp. 11-18,

the Connecticut Chemosensory Clinical Research Center (CCCRC). *Laryngoscope*,

*Developmental Neuroscience,* Vol.20, No.4-5, (July-October 1998), pp. 271-276, ISSN

the striatum in parkinsonian syndromes. *European Journal of Neurology,* Vol.8, No.6,

multiple system atrophy from idiopathic Parkinson's disease using proton magnetic resonance spectroscopy. *Annals of Neurology,* Vol.37, No.2, (February

alterations in other metal ions occurring in brain in Parkinson's disease. *Journal of* 

spectroscopy of the brain at 3T: an update. *European Radiology*, Vol.17, No.7, (July

deficit unrelated to neurologic signs, disease stage, or disease duration. *Neurology,*

differentiates between progressive supranuclear palsy and idiopathic Parkinson's

three neurodegenerative diseases. *Geriatrics,* Vol.46, No.1, (August 1991), pp. 47-51,

identification test: a rapid quantitative olfactory function test for the clinic.

Parkinson's disease. *Annals of Neurology,* Vol.20, No.6, (December 1986), pp. 744-

idiopathic Parkinson's disease, multiple system atrophy, progressive supranuclear palsy, and healthy controls using magnetization transfer imaging. *Neuroimage*,

angiography of the brain with magnetization transfer contrast. *Radiology*, Vol.184,

Nonetheless, there is increasing evidence that POM evaluation and advanced in vivo brain imaging will provide critical clues to assist in the early diagnosis and medical management of PD patients.

These methods are broadly defined as characteristics that are objectively measured and evaluated as indicators of normal biological processes, pathogenic processes, or pharmacological responses to a therapeutic intervention.

The lack of success of recent disease-modifying therapeutic trials coupled with the huge expense of other methods, such as the nuclear medicine, has highlighted the need for such an ambitious approach to identify and validate early markers of PD progression for future clinical studies of disease-modifying drugs.

## **6. References**


Nonetheless, there is increasing evidence that POM evaluation and advanced in vivo brain imaging will provide critical clues to assist in the early diagnosis and medical management

These methods are broadly defined as characteristics that are objectively measured and evaluated as indicators of normal biological processes, pathogenic processes, or

The lack of success of recent disease-modifying therapeutic trials coupled with the huge expense of other methods, such as the nuclear medicine, has highlighted the need for such an ambitious approach to identify and validate early markers of PD progression for future

Abe, K.; Terakawa, H.; Takanashi, M.; Watanable, Y.; Tanaka, H. et al. Proton magnetic

Allen, P.S. In vivo nuclear-magnetic-resonance spectroscopy applied to medicine. *Canadian* 

Ansari, K.A. & Johnson, A. Olfactory function in patients with Parkinson's disease. *Journal of Chronic Diseases,* Vol.28, No.9, (October 1975), pp. 493-497, ISSN 0021-9681 Atlas, S.W.; Howard, II R.S.; Maldijian, J.; Alsop, D.; Detre, J.A. et al. Functional magnetic

Avison, M.J.; Hetherington, H.P. & Shulman R.G. Application of NMR to studies of tissue

Bandettini, P.A.; Jesmanowicz, A.; Wong, E.C. & Hyde, J.S. Processing strategies for time-

Bares, M.; Bràzdil, M.; Kanovsky, P.; Jurak, P.; Daniel, P.; et al. The effect of apomorphine

Barker, G.J.; Tofts, P.S. & Gass, A. An interleaved sequence for accurate and reproducible

Bonavita, S.; Di Salle, F. & Tedeschi, G. Proton MRS in neurological disorders. *European Journal of Radiology,* Vol.30, No.2, (May 1999), pp. 125-131, ISSN 0720-048X Bowen, B.C.; Block, R.E.; Sanchez-Ramos, J.; Pattany, P.M.; Lampman, D.A. et al. Proton MR

*Neuroradiology*, Vol.16, No.1, (January 1995), pp. 61-68, ISSN 0195-6108 Braak, H.; Del Tredici, K.; Rüb, U.; De Vos, R.A.I.; Jansen Steur, E.N.H. et al. Staging of brain

*Medicine*, Vol.30, No2, (August 1993), pp. 161-173, ISSN 07403194

Vol.52, No.6, (August 2000), pp. 589-595, ISSN 0361-9230

(February 1996), pp. 329-338, ISSN 0148-396X

Vol.14, No.4, (1996), pp. 403-411, ISSN 0730-725X

No.2, (April 2003), pp. 197-211, ISSN 0197-4580

1986), pp. 377-402, ISSN 0883-9182

resonance spectroscopy of patients with parkinsonism. *Brain Research Bulletin,*

*Association of Radiologists Journal,* Vol.41, No.1, (February 1990), pp. 39-44, ISSN

resonance imaging of regional brain activity in patients with intracerebral gliomas: findings and implications for clinical management. *Neurosurgery*, Vol.38, No2,

metabolism. *Annual Review of Biophysics and Biophysical Chemistry*, Vol.15, (June

course data sets in functional MRI of the human brain. *Magnetic Resonance in* 

administration on smooth pursit ocular movements in early Parkinsonian patients. *Parkinsonism & Related Disorders*, Vol. 9, No.3, (January 2003), pp. 139–144, ISSN

clinical measurement of magnetization transfer ratio. *Magnetic Resonance Imaging,* 

Spectroscopy of the Brain in 14 Patients with Parkinson Disease. *American Journal of* 

pathology related to sporadic Parkinson's disease. *Neurobiology of Aging,* Vol.24,

pharmacological responses to a therapeutic intervention.

clinical studies of disease-modifying drugs.

of PD patients.

**6. References** 

0008-2902

1353-8020


Early Marker for the Diagnosis of Parkinson's Disease 23

Horsfield, M.A. Magnetization transfer imaging in multiple sclerosis. *Journal of* 

Huisman, E.; Uylings, H.B. & Hoogland, P.V. A 100% increase of dopaminergic cells in the

Hummel, T. (1999). Olfactory evoked potentials as a tool to measure progression of

Hummel, T.; Kobal, G.; Gudziol, H. & Mackay-Sim, A. Normative data for the

Hummel, T.; Sekinger, B.; Wolf, S.; Pauli, E. & Kobal G. "Sniffin' Sticks": olfactory

Hummel, T.; Witt, M.; Reichmann, H.; Welge-Luessen, A & Haehner A.

Inglese, M.; Horsfield, M.A. & Filippi, M. Scan-rescan variation of measures derived from

Kettenmann, B.; Hummel, C.; Stefan, H. & Kobal, G. Multiple olfactory activity in the

Kim, K.; Hirsch, J.; DeLaPaz, R.L.; Relkin, N. & Lee, K.M. Comparison of cortical areas

Kobal, G. & Plattig, K.H. Objective olfactometry: methodological annotations for recording

Kobal, G.; Klimek, L.; Wolfensberger, M.; Gudziol, H.; Temmel, A. et al. Multicenter

*Neuroradiology,* Vol.22, No.4, (April 2001), pp. 681-684, ISSN 0195-6108 Karagulle Kendi, A.T.; Lehericy, S.; Luciana, M.; Ugurbil, K. & Tuite, P. Altered diffusion in

*Neuroimaging,* Vol.15, No.4, (2005), pp. 58-67, ISSN 1051-2284

Vol.19, No.6, (June 2004), pp. 687-692, ISSN 0885-3185

No.1-2, (February 2010), pp. 119-122, ISSN 0022-510X

Vol.38, No3, (March 2008), pp. 501-5, ISSN 1936-959X

No.5, (October 1997), pp. 493-502, ISSN 0379-864X

pp. 589-594, ISSN 0740-3194

Blackwell Science, Oxford, UK

1997), pp. 39-52, ISSN 0379-864X

(1993), pp.1763, ISSN 0270-6474

7590

243, ISSN 0937-4477

multicenter pilot study. *Magnetic Resonance in Medicine,* Vol.33, No.5, (May 1995),

olfactory bulb may explain hyposmia in Parkinson's Disease. *Movement Disorders*,

Parkinson's disease. In: *Focus medicine* – *New development in the drug therapy of Parkinson's Disease,* T. Chase & P. Bedard, (Eds.), 47-53, ISBN 0-632-05174-4,

"Sniffin'Sticks" including tests of odor identification, odor discrimination, and olfactory thresholds:an upgrade based on a group of more than 3.000 subjects. *European Archives of Oto-Rhino-Laryngology,* Vol.264, No.3, (March 2007), pp. 237-

performance assessed by the combined testing of odor identification, odor discrimination and olfactory threshold. *Chemical Senses,* Vol.22, No.1, (February

Immunohistochemical, volumetric, and functional neuroimaging studies in patients with idiopathic Parkinson's disease. *Journal of the Neurological Sciences,* Vol.298,

brain magnetization transfer ratio histograms obtained in healthy volunteers by use of a semi-interleaved magnetization transfer sequence. *American Journal of* 

the frontal lobe in Parkinson disease. *AJNR American Journal of Neuroradiology*,

human neocortex identified by magnetic source imaging. *Chemical Senses,* Vol.22,

activated by primary and secondary languages in human brain using functional magnetic resonance imaging (fMRI). *Abstracts: Society of Neuroscience*, Vol.21, No3,

olfactory EEG-responses from the awake human. *EEG EMG Z Elektroenzephalogr Elektromyogr Verwandte Geb.*, Vol.9, No.3, (September 1978), pp. 135-145, ISSN 0012-

investigation of 1.036 subjects using a standardized method for the assessment of olfactory function combining tests of odor identification, odor discrimination, and


Federico, F.; Simone, I.L.; Lucivero, V.; Iliceto, G.; De Mari, M. et al. Proton magnetic

Firbank, M.J.; Harrison, R.M. & O'Brien, J.T. A comprehensive review of proton magnetic

George, JS.; Aine, CJ.; Mosher, JC.; Schmidt, MD.; Ranken, DM. et al., Mapping function in

*Neurophysiology,* Vol.12, No5, (September 1995), pp. 406-429 ISSN 1676-2649 Gupta, A.; Dawson, V.L. & Dawson, T.M. What causes cell death in Parkinson's disease? *Annals of Neurology,* Vol.64, No6, (December 2008), pp. S3-S15, ISSN 0364-5134 Haehner, A.; Boesveldt, S.; Berendse, H.W.; Mackay-Sim, A. Fleischmann, J. et al. Prevalence

Hagmann, P.; Jonasson, L.; Maeder, P.; Thiran, J.P.; Wedeen, V.J. et al. Understanding

Hattingen, E.; Magerkurth, J.; Pilatus, U.; Morez, A.; Seifried, C. et al. Phosphorus and

Hawkes, C.H.; Shephard, B.C. & Daniel, S.E. Olfactory disfunction in Parkinson's disease.

Hawkes, C.H.; Shephardm, B.C. & Daniel, S.E. Is Parkinson's disease a primary olfactory

Hinke, R.M.; Hu, X.; Stillman, A.E.; Kim, S.G.; Merkle, H. et al. Functional magnetic

Hirsch, E.C. & Hunot S. Neuroinflammation in Parkinson's disease: a target for

Holshouser, B.A.; Komu, M.; Moller, H.E.; Zijlmans, J.; Kolem, H. et al. Localized proton

*Disorders,* Vol.15, No.7, (August 2009), pp. 490-494, ISSN 1353-8020

223-229, ISSN 0392-0461

pp. S205-U219, ISSN 0271-5333

pp. 3285-3297, ISSN 0006-8950

446, ISSN 0022-3050

1474-4422

473-480, ISSN 1460-2725

(June 1993) pp. 675-678 ISSN 0959-4965

0893-133X

resonance spectroscopy in Parkinson's disease and atypical parkinsonian disorders. *Movement Disorders*, Vol.12, No.6, (November 1997), pp. 903-909, ISSN 0885-3185 Federico, F.; Simone, I.L.; Lucivero, V.; Mezzapesa, D.M.; De Mari, M. et al. Usefulness of

proton magnetic resonance spectroscopy in differentiating parkinsonian syndromes. *Italian Journal of Neurological Sciences*, Vol.20, No.4, (August 1999), pp.

resonance spectroscopy studies in dementia and Parkinson's disease. *Dementia And Geriatric Cognitive Disorders*, Vol.14, No.2, (July 2002), pp. 64-76, ISSN 1420-8008 Fox, P.T. & Raichle, M.E. Stimulus rate determines regional brain blood flow in striate cortex. *Annals of Neurology,* Vol.17, No3, (March 1985), pp.303-305, ISSN 0364-5134 Fulton, HG. & Barrett, SP. A demonstration of intravenous nicotine self-administration in

humans? *Neuropsychopharmacology,* Vol.33, No8, (Jule 2008), pp. 2042-2043, ISSN

the human brain with magneto encephalography, anatomical magnetic resonance imaging, and functional magnetic resonance imaging. *Journal of Clinical* 

of smell loss in Parkinson's Disease - a multicenter study. *Parkinsonism & Related* 

diffusion MR imaging techniques: from scalar diffusion-weighted imaging to diffusion tensor imaging and beyond. *Radiographics*, Vol.26, Sp. Iss, (October 2006),

proton magnetic resonance spectroscopy demonstrates mitochondrial dysfunction in early and advanced Parkinson's disease. *Brain,* Vol.132, No.12, (December 2009),

*Journal of Neurology Neurosurgery and Psychiatry,* Vol.62, No.5, (May 1997), pp. 436-

disorder? *Qjm-An International Journal of Medicine,* Vol.92, No.8, (August 1999), pp.

resonance imaging of Broca's area during internal speech. *NeuroReport*, Vol.4, No6,

neuroprotection? *Lancet Neurology,* Vol.8, No.4, (April 2009), pp. 382-397, ISSN:

NMR spectroscopy in the striatum of patients with idiopathic Parkinson's disease: a

multicenter pilot study. *Magnetic Resonance in Medicine,* Vol.33, No.5, (May 1995), pp. 589-594, ISSN 0740-3194


Early Marker for the Diagnosis of Parkinson's Disease 25

Schneider, W.; Noll, D.C. & Cohen, J. D. Functional topographic mapping of the cortical

Schocke, M.F.; Seppi, K.; Esterhammer, R.; Kremser, C.; Mair, K.J. et al. Trace of diffusion

Sofic, E.; Riederer, P.; Heinsen, H.; Beckmann, H.; Reynolds, G.P. et al. Increased iron (III)

Srinivasan, R.; Vigneron, D.; Sailasuta, N.; Hurd, R. & Nelson, S. A comparative study of

Takeda, A.; Saito, N.; Baba, T.; Kikuchi, A.; Sugeno, N. et al. Functional imaging studies of

Tambasco, N.; Pelliccioli, G.P.; Chiarini, P.; Montanari, G.E.; Leone, F. et al. Magnetization

Tedeschi, G.; Litvan, I.; Bonavita, S.; Bertolino, A.; Lundbom, N. et al. Proton magnetic

Tessa, C.; Giannelli, M.; Della Nave, R.; Lucetti, C.; Berti, C.A. et al. A whole-brain analysis

Tissingh, G.; Berendse, H.W.; Bergmans, P.; DeWaard, R.; Drukarch, B. et al. Loss of

Watanabe, H.; Fukatsu, H.; Katsuno, M.; Sugiura, M.; Hamada, K. et al. Multiple regional

*Psychiatry,* Vol.75, No.1, (January 2004), pp. 103-109, ISSN 0022-3050 Wattendorf, E.; Welge-Lüssen, A.; Fiedler K.; Bilecen, D.; Wolfensberger, M. et al. Olfactory

(September 1993), pp. 150-152, ISSN 0028-0836

No.4, (May 2004), pp. 523-528, ISSN 0730-725X

Vol.45, No.4, (April 2003), pp. 224-230, ISSN 0028-3940

(February 2010), pp. 36-39, ISSN 0022-510X

(April 2008), pp. 674-680, ISSN 01956108

1541-1552, ISSN 0006-8950

537-543, ISSN 0306-4522

No.4, (April 1995), pp. 247-250, ISSN 0001-6314

9564

3185

ribbon in human vision with conventional MRI scanners. *Nature*, Vol365, No6442,

tensor differentiates the Parkinson variant of multiple system atrophy and Parkinson's disease. *Neuroimage*, Vol.19, Suppl.9, (2004), pp. S53-S53, ISSN 08853185

and total iron content in post mortem substantia nigra of parkinsonian brain. *Journal of Neural Transmission,* Vol.74, No3, (August 1988), pp.199-205, ISSN 0300-

myo-inositol quantification using LCmodel at 1.5 T and 3.0 T with 3D 1H proton spectroscopic imaging of the human brain. *Magnetic Resonance Imaging*, Vol.22,

hyposmia in Parkinson's disease. *Journal of the Neurological Sciences*, Vol.298, No.1-2,

Transfer changes of gray and white matter in Parkinson 's disease. *Neuroradiology*,

resonance spectroscopic imaging in progressive supranuclear palsy, Parkinson's disease and corticobasal degeneration. *Brain,* Vol.120, No.9, (September 1997), pp.

in de novo Parkinson disease. *AJNR American Journal of Neuroradiology*, Vol.29, No4,

olfaction in de novo and treated Parkinson's disease: possible implications for early diagnosis. *Movement Disorders*, Vol.16, No.1, (January 2001), pp. 41-46, ISSN 0885-

1H-MR spectroscopy in multiple system atrophy: NAA/Cr reduction in pontine base as a valuable diagnostic marker. *Journal of Neurology, Neurosurgery, and* 

Impairment Predicts Brain Atrophy in Parkinson's Disease. *The Journal of Neuroscience*, Vol.29, No.49, (December 2009), pp. 15410-15413, ISSN 0270-6474 Welge-Lüssen, A.; Wattendorf, E.; Schwerdtfeger, U.; Fuhr, P.; Bilecen, D. et al. Olfactory

induced brain activity in Parkinson's disease relates to the expression of eventrelated potentials—an fMRI study. *Neuroscience,* Vol.162, No.2, (August 2009), pp.

function in atypical parkinsonian syndromes. *Acta Neurologica Scandinavica*, Vol.91,

Wenning, G.K.; Shephard, B.; Hawkes, C.; Petruckevitch, A.; Lees, A. et al. Olfactory

olfactory thresholds. *European Archives of Oto-Rhino-Laryngology,* Vol.257, No.4, (April 2000), pp. 205-211, ISSN 0937-4477


Le Bihan, D. Looking into the functional architecture of the brain with diffusion MRI. *Nature Reviews Neuroscience*, Vol.4, No.6, (June 2003), pp. 469-480, ISSN 1471-0048 Lekwuwa, G.U.; Barnes, G.R.; Collins, C.J.S.; Limousin, P. Progressive bradykinesia and

Lin, A.; Ross, B.D.; Harris, K. & Wong, W. Efficacy of proton magnetic resonance

Marino, S.; Sessa, E.; Di Lorenzo, G.; Lanzafame, P.; Scullica, G.; et al. Quantitative analysis

Marino, S.; Lanzafame, P.; Sessa, E.; Bramanti, A.; Bramanti, P. The effect of L-Dopa

*Neurological Sciences*, Vol. 31,No. 3 (June 2010), pp. 381-385, ISSN 0022-510X Muller, A.; Reichmann, H.; Livermore, A. & Hummel T. Olfactory function in idiopathic

Ogawa, S.; Lee, T.M.; Nayak, A. S.; & Glynn, P. Oxygenation-sensitive contrast in magnetic

Pike, G.B.; Hu, B.S.; Glover, G.H. & Enzmann, D.R. Magnetization transfer time-of-flight

Ramaker, C.; Marinus, J.; Stiggelbout, A.M. & Van Hilten, B.J. Systematic evaluation of

Ross, B. & Bluml, S. Magnetic resonace spectroscopy of the human brain. *The Anatomical* 

Roy, C. & Sherrington, C. On the regulation of the blood-supply of the brain. *The Journal of* 

Rutledge, J.N.; Hilal, S.K.; Silver, A.J.; Defendini, R. & Fahn S. Study of movement disorders

Savoiardo, M.; Girotti, F.; Strada, L. & Ciceri, E. Magnetic resonance imaging in progressive

*Disorders*, Vol.17, No.5, (September 2002), pp. 867-876, ISSN 0885-3185 Rolheiser, T.M.; Fulton, H.G.; Good, K.P.; Fisk, J.D.; McKelvey, J.R. et al. Diffusion tensor

(April 2000), pp. 205-211, ISSN 0937-4477

No.2, (April 2005), pp. 197-214, ISSN 1545-5343

Vol.109, No.5-6, (May 2002), pp. 805-811, ISSN 0300-9564

*Medicine*, Vol.14, No1, (April 1990) pp.68-78, ISSN 07403194

*Journal of Neurology,* (February 2011), ISSN 0340-5354

*Record,* Vol.265, No.2, (April 2001), pp. 54-84, ISSN 0003-276X

*Physiology,* Vol.11*,* No1-2, (January 1890), pp.85-108, ISSN 1469-7793

(June 1992), pp. 372-379, ISSN 0740-3194

1987), pp.365-379, ISSN 0361-803X

Vol.42, (1994), pp. 93-110, ISSN 0303-6995

ISSN 0014-3022

olfactory thresholds. *European Archives of Oto-Rhino-Laryngology,* Vol.257, No.4,

hypokinesia of ocular pursuit in Parkinson's disease. *Journal of Neurology Neurosurgery and Psychiatry,* Vol.66, No.6, (June 1999), pp. 746–753, ISSN 0022-3050

spectroscopy inneurological diagnosis and neurotherapic decision making. *NeuroRx: the journal of the American Society for Experimental NeuroTherapeutics,* Vol.2,

of pursuit ocular movements in Parkinson's disease by using a video-based eye tracking system. *European Neurology*, Vol. 58, No. 4, (September 2007), pp. 193-7,

administration on pursuit ocular movements in suspected Parkinson's disease.

Parkinson's disease (IPD): results from cross-sectional studies in IPD patients and long-term follow-up of de-novo IPD patients. *Journal of Neural Transmission,* 

resonance image of rodent brain at high magnetic fields. *Magnetic Resonance in* 

magnetic resonance angiography. *Magnetic Resonance in Medicine,* Vol.25, No.2,

rating scales for impairment and disability in Parkinson's disease. *Movement* 

imaging and olfactory identification testing in early-stage Parkinson's disease.

and brain iron by MR. *AJR American Journal of Roentgenology,* Vol.149, No2, (August

supranuclear palsy and other parkinsonian disorders. *Journal of Neural Transmission,* 


**2** 

*Kiev, Ukraine* 

**Diagnosis of Parkinson's Disease** 

**by Electrophysiological Methods** 

Elena Lukhanina, Irina Karaban and Natalia Berezetskaya *Department of Brain Physiology, A.A. Bogomoletz Institute of Physiology,* 

Effective differential diagnosis of Parkinson's disease (PD) needs in informative indices that objectively reflect the functional state of the extrapyramidal system. And also, when evaluating an efficacy of antiparkinsonian therapy, it is essential to have both, qualitative and quantitative characteristics, permitting to correct treatment and predict a disease course. One of informative diagnostic method in PD is surface (interference) electromyography. As know, a nigrostriatal dopamine deficit results in disturbances of the central supraspinal control over the muscle tonic activity and voluntary movements (Valls-Solé & Valldeoriola, 2002). Electromyographically, the extrapyradimal insufficiency shows itself by a high level of bioelectrical activity of muscles at rest, changes of motor unit conduction velocity and synchronization (Farina et al., 2004; Semmler & Nordstrom, 1999). The traditional methods to evaluate surface electromyograms (EMGs) are based on amplitude and spectral analysis. However, myoelectric signals are nonlinear by its nature (Nieminen & Takala, 1996). A surface EMG is formed by the summation of a number of single muscle fiber action potentials. Therefore different world clinics have been searching for new relevant methods based on nonlinear time-series analyses of EMG to quantify the motor features of the disorder in PD (Del Santo et al., 2007; Meigal et al., 2009). Some other novel EMG characteristics, such as dimensionality based on fractal analysis or higher order statistics of EMG distribution have also proved to be sensitive to neuromuscular status (Swie et al.,

Although the cardinal symptoms of the disease are movement disorders the manifestations of PD also comprise a variety of diverse abnormalities including disturbance of sensory gating and cognitive decline (Lewis & Byblow, 2002). Several authors suggested that movement disorders in PD might be also developed because of dysregulation of sensory processing that affects sensorimotor integration (Abbruzzese & Berardelli, 2003). This is an important issue because one of the proposed key functions of basal ganglia is the gating of sensory input for motor control (Kaji, 2001). Numerous studies have demonstrated marked changes in the somatosensory (Rossini et al., 1998), acoustic (Teo et al., 1997;) and visual (Sadekov, 1997) evoked potentials in PD patients. Evaluation of brain evoked potentials may have potential in the assessment of the severity of PD. In contemporary neurophysiology,

**1. Introduction** 

2005).

*Parkinson's Disease Treatment Centre, Institute of Gerontology,* 


## **Diagnosis of Parkinson's Disease by Electrophysiological Methods**

Elena Lukhanina, Irina Karaban and Natalia Berezetskaya *Department of Brain Physiology, A.A. Bogomoletz Institute of Physiology, Parkinson's Disease Treatment Centre, Institute of Gerontology, Kiev, Ukraine* 

## **1. Introduction**

26 Diagnostics and Rehabilitation of Parkinson's Disease

Westermann, B.; Wattendorf, E.; Schwerdtfeger, U.; Husner, A.; Fuhr, P. et al. Functional

Wilms, H.; Zecca, L.; Rosenstiel, P.; Sievers, J.; Deuschl, G. et al. Inflammation in Parkinson's

Wolff, S.D. & Balaban, R.S. Magnetization transfer contrast (MTC) and tissue water proton

Wolff, S.D.; Eng, J. & Balaban, R.S. Magnetization transfer contrast: method for improving

Yonca, A.; Pervin, I.; Ali, D.; Sezer, K. & Naighan, I. Magnetization Transfer Ratio in Early

Youdim, M.B.H.; Ben-Schachar, D.; Yehuda, S. & Riederer, P. The role of iron in the basal ganglia. *Advances in Neurology,* Vol.53, (1990) pp.155-162, ISSN: 0091-3952

19-24, ISSN 0022-3050

135-144, ISSN 0740-3194

133-137, ISSN 0033-8419

189-192, ISSN 1076-6332

6128

imaging of the cerebral olfactory system in patients with Parkinson's disease. *Journal of Neurology, Neurosurgery, and Psychiatry,* Vol.79, No.1, (January 2008), pp.

diseases and other neurodegenerative diseases: cause and therapeutic implications. *Current Pharmaceutical Design,* Vol.13*,* No18, (June 2007), pp. 1925-1928, ISSN: 1381-

relaxation in vivo. *Magnetic Resonance in Medicine,* Vol.10, No.1, (April 1989), pp.

contrast in gradient-recalled-echo images*. Radiology*, Vol.179, No.1, (April 1991), pp.

Period of Parkinson Disease. *Academic Radiology*, Vol.14, No.2, (February 2007), pp.

Effective differential diagnosis of Parkinson's disease (PD) needs in informative indices that objectively reflect the functional state of the extrapyramidal system. And also, when evaluating an efficacy of antiparkinsonian therapy, it is essential to have both, qualitative and quantitative characteristics, permitting to correct treatment and predict a disease course. One of informative diagnostic method in PD is surface (interference) electromyography. As know, a nigrostriatal dopamine deficit results in disturbances of the central supraspinal control over the muscle tonic activity and voluntary movements (Valls-Solé & Valldeoriola, 2002). Electromyographically, the extrapyradimal insufficiency shows itself by a high level of bioelectrical activity of muscles at rest, changes of motor unit conduction velocity and synchronization (Farina et al., 2004; Semmler & Nordstrom, 1999). The traditional methods to evaluate surface electromyograms (EMGs) are based on amplitude and spectral analysis. However, myoelectric signals are nonlinear by its nature (Nieminen & Takala, 1996). A surface EMG is formed by the summation of a number of single muscle fiber action potentials. Therefore different world clinics have been searching for new relevant methods based on nonlinear time-series analyses of EMG to quantify the motor features of the disorder in PD (Del Santo et al., 2007; Meigal et al., 2009). Some other novel EMG characteristics, such as dimensionality based on fractal analysis or higher order statistics of EMG distribution have also proved to be sensitive to neuromuscular status (Swie et al., 2005).

Although the cardinal symptoms of the disease are movement disorders the manifestations of PD also comprise a variety of diverse abnormalities including disturbance of sensory gating and cognitive decline (Lewis & Byblow, 2002). Several authors suggested that movement disorders in PD might be also developed because of dysregulation of sensory processing that affects sensorimotor integration (Abbruzzese & Berardelli, 2003). This is an important issue because one of the proposed key functions of basal ganglia is the gating of sensory input for motor control (Kaji, 2001). Numerous studies have demonstrated marked changes in the somatosensory (Rossini et al., 1998), acoustic (Teo et al., 1997;) and visual (Sadekov, 1997) evoked potentials in PD patients. Evaluation of brain evoked potentials may have potential in the assessment of the severity of PD. In contemporary neurophysiology,

Diagnosis of Parkinson's Disease by Electrophysiological Methods 29

Studies were performed in two groups: 48 patients with PD, 1.5-3.5 Hoehn-Yahr scale (23 men and 25 women, mean ± SE age 62.2 1.6, range 49-75 years) and 42 age-matched healthy controls (20 men and 22 women, mean ± SE age 65.8 ± 1.43, range 58-74 years). All of them were right-handed persons. The study patients, who regularly underwent treatment at the Parkinson's Disease Centre of Institute of Gerontology, gave their written informed consent to participate in this investigation. They had 4-13-year history of an idiopathic PD and received an antiparkinsonian medication (an individual dose 0.250-12g of levodopa/carbidopa, daily). All patients were studied in the OFF state. For the quantitative evaluation of levodopa therapy 20 patients (in which the clinical "ON-OFF" phenomenon was verified) were studied also in the ON state, one hour after an intake of the single individual dose of levodopa/carbidopa. The motor activity of PD patients was evaluated in

For each subject, we recorded the surface EMG from the flexors and extensors (mm. biceps and triceps brachii) of the right and left arms. The subject lay on his back, with the arms lying on the horizontal surface. The EMGs were recorded using four bipolar skin electrodes (0.5x1.0 cm) with an interelectrode distance constant of 1.5 cm. Bioelectrical potentials were amplified with a band pass of 10 Hz - 10 kHz. The amplified analogue signals were fed to a computer, which digitized them at a sampling rate of 1000 Hz and then stored the data for further measurements. The time of each recording was 10 sec. EMG recordings were made:

2. During a voluntary m. biceps brachii contraction started after a command to bend the arm at the elbow, fingers touching the shoulder. The sound signal served as a command to initiate arm bending, and it was simultaneously registered on EMG. An electrical contact enclosure marked the start of arm lifting, being recorded

3. During a tonic m. biceps brachii strain under weight holding (2 kg) in the hand, with

In resting EMG recordings, the average and maximal EMG amplitudes were calculated to evaluate the muscle ability for relaxation. An artefact-free section on the EMG record was selected by the experimenter. The single EMG wave amplitude was defined as the difference between the values of upper and lower peaks. The oscillations with no less than 2 V amplitude were considered. The resting average EMG amplitude was calculated based on minimum 100 measurements. The maximum amplitude value was calculated on the same

The resting EMG recordings were also used for assessing the burst muscle discharges For this purpose, the Fourier spectrum diagrams for the low frequency area were constructed. In doing this, part of the data with negative EMG amplitude meanings were discarded. Data with the positive EMG amplitude meanings underwent a Butterworth digital sinus low pass filtering with a band pass of 0-20 Hz As a result, the envelope of EMG amplitude was formed, which then served as the data array for fast Fourier transformation. The data for each lead were divided into several successive sections, each containing 512 points, which underwent the fast Fourier transformation. The obtained spectra were averaged, and the envelope of EMG amplitude frequency with a maximal power was determined (Fig 1). The statistical significance of the frequency peak was determined by means of constructing the

**2.1.1 Methods** 

ON state, according to sections II-III of the UPDRS.

1. At a resting state, being cautious that the subject is relaxed.

simultaneously with the EMG on a free channel.

95% confidential intervals (M 2 S.D).

arm lifted upward and stretched forward, for 5 sec.

EMG section. The amplitude distribution histograms were constructed.

studies of the central mechanisms underlying the organization of motor function and its impairments increasingly involve analysis of endogenous cortical event-related potentials, a set of potentials which includes contingent negative variation (CNV). The CNV extent depends on the level of attention, motivation, and volitional effort (Deecke, 2001). The magnitude of this potential is known to decrease in diseases accompanied by motor disorders, including PD (Aotsuka et al., 1996; Pulvenmuller et al., 1996). CNV has been shown to display significant increases after administration of levodopa in patients with Parkinson's disease, suggesting a role for the central dopaminergic system in its generation (Oishi et al., 1995).

Although electrophysiological methods objectively reflect motor and sensory dysfunctions, they are still used rather rarely in the clinical evaluation of PD. We carried out systematic and detailed research of surface EMG characteristics in PD patients in comparison with agematched healthy subjects, paying the special attention on the correlation associations between EMG parameters and subitems of the Unified Parkinson's Disease Rating Scale (UPDRS). Amplitude and spectral features, statistics of distribution, fractal dynamics of the EMG signals were investigated. Separate research was dedicated to the study of EMG characteristics of clinically healthy kinsmen of the patients suffering from PD in order to detect latent symptoms of extrapyramidal insufficiency that can be considered genetic determinants of the risk of development of the above disease. Since the question of the relationship the early and late phases of CNV with the mechanisms controlling motor functions in PD has received inadequate study we conducted such research in patients with this disease. With the purpose of evaluation of the brain inhibitory processes in PD patients the study of cortical evoked potentials upon paired-click auditory stimulation was performed. The results of these investigations are presented below.

## **2. Surface electromyography**

Surface EMG is a simple and noninvasive method that permits to estimate the severity of symptomatology in patients and also may help to exposure of the hidden manifestations of the disturbed muscles activity on the presymptomatic stage of the neurodegenerative process (Kryzhanovsky et al., 2002; Lukhanina et al., 2010). In PD patients the EMG characteristics of the tonic and phasic shoulder muscle activities at rest, during voluntary contraction and under tonic muscle strain were studied. In kinsmen of the patients, suffering from PD, EMGs in the resting state and under conditions of two functional tests (retention of load and retention of arms in the elevated and outstretched state) were recorded.

#### **2.1 Amplitude and spectral analysis of EMGs in patients with Parkinson's disease**

One of the informative EMG sign of extrapyramidal insufficiency appears to be the resting EMG amplitude values that reflect the muscle ability for relaxation. Spectral analysis of resting EMGs is used for assessing the burst muscle discharges with a frequency of 4-8 Hz reflecting parkinsonian tremor. Amplitude values of the EMGs recorded during the voluntary muscle contraction serve to calculate the phasic activation coefficient. This coefficient clearly reflects the competitive relationships between the tonic and phasic processes. Study of the reflex agonist/antagonist muscle involvement under tonic strain is valid for establishing coordinating muscle relationships.

#### **2.1.1 Methods**

28 Diagnostics and Rehabilitation of Parkinson's Disease

studies of the central mechanisms underlying the organization of motor function and its impairments increasingly involve analysis of endogenous cortical event-related potentials, a set of potentials which includes contingent negative variation (CNV). The CNV extent depends on the level of attention, motivation, and volitional effort (Deecke, 2001). The magnitude of this potential is known to decrease in diseases accompanied by motor disorders, including PD (Aotsuka et al., 1996; Pulvenmuller et al., 1996). CNV has been shown to display significant increases after administration of levodopa in patients with Parkinson's disease, suggesting a role for the central dopaminergic system in its generation

Although electrophysiological methods objectively reflect motor and sensory dysfunctions, they are still used rather rarely in the clinical evaluation of PD. We carried out systematic and detailed research of surface EMG characteristics in PD patients in comparison with agematched healthy subjects, paying the special attention on the correlation associations between EMG parameters and subitems of the Unified Parkinson's Disease Rating Scale (UPDRS). Amplitude and spectral features, statistics of distribution, fractal dynamics of the EMG signals were investigated. Separate research was dedicated to the study of EMG characteristics of clinically healthy kinsmen of the patients suffering from PD in order to detect latent symptoms of extrapyramidal insufficiency that can be considered genetic determinants of the risk of development of the above disease. Since the question of the relationship the early and late phases of CNV with the mechanisms controlling motor functions in PD has received inadequate study we conducted such research in patients with this disease. With the purpose of evaluation of the brain inhibitory processes in PD patients the study of cortical evoked potentials upon paired-click auditory stimulation was

Surface EMG is a simple and noninvasive method that permits to estimate the severity of symptomatology in patients and also may help to exposure of the hidden manifestations of the disturbed muscles activity on the presymptomatic stage of the neurodegenerative process (Kryzhanovsky et al., 2002; Lukhanina et al., 2010). In PD patients the EMG characteristics of the tonic and phasic shoulder muscle activities at rest, during voluntary contraction and under tonic muscle strain were studied. In kinsmen of the patients, suffering from PD, EMGs in the resting state and under conditions of two functional tests (retention of load and retention of arms in the elevated and outstretched state) were

**2.1 Amplitude and spectral analysis of EMGs in patients with Parkinson's disease**  One of the informative EMG sign of extrapyramidal insufficiency appears to be the resting EMG amplitude values that reflect the muscle ability for relaxation. Spectral analysis of resting EMGs is used for assessing the burst muscle discharges with a frequency of 4-8 Hz reflecting parkinsonian tremor. Amplitude values of the EMGs recorded during the voluntary muscle contraction serve to calculate the phasic activation coefficient. This coefficient clearly reflects the competitive relationships between the tonic and phasic processes. Study of the reflex agonist/antagonist muscle involvement under tonic strain is

performed. The results of these investigations are presented below.

valid for establishing coordinating muscle relationships.

**2. Surface electromyography** 

recorded.

(Oishi et al., 1995).

Studies were performed in two groups: 48 patients with PD, 1.5-3.5 Hoehn-Yahr scale (23 men and 25 women, mean ± SE age 62.2 1.6, range 49-75 years) and 42 age-matched healthy controls (20 men and 22 women, mean ± SE age 65.8 ± 1.43, range 58-74 years). All of them were right-handed persons. The study patients, who regularly underwent treatment at the Parkinson's Disease Centre of Institute of Gerontology, gave their written informed consent to participate in this investigation. They had 4-13-year history of an idiopathic PD and received an antiparkinsonian medication (an individual dose 0.250-12g of levodopa/carbidopa, daily). All patients were studied in the OFF state. For the quantitative evaluation of levodopa therapy 20 patients (in which the clinical "ON-OFF" phenomenon was verified) were studied also in the ON state, one hour after an intake of the single individual dose of levodopa/carbidopa. The motor activity of PD patients was evaluated in ON state, according to sections II-III of the UPDRS.

For each subject, we recorded the surface EMG from the flexors and extensors (mm. biceps and triceps brachii) of the right and left arms. The subject lay on his back, with the arms lying on the horizontal surface. The EMGs were recorded using four bipolar skin electrodes (0.5x1.0 cm) with an interelectrode distance constant of 1.5 cm. Bioelectrical potentials were amplified with a band pass of 10 Hz - 10 kHz. The amplified analogue signals were fed to a computer, which digitized them at a sampling rate of 1000 Hz and then stored the data for further measurements. The time of each recording was 10 sec. EMG recordings were made:


In resting EMG recordings, the average and maximal EMG amplitudes were calculated to evaluate the muscle ability for relaxation. An artefact-free section on the EMG record was selected by the experimenter. The single EMG wave amplitude was defined as the difference between the values of upper and lower peaks. The oscillations with no less than 2 V amplitude were considered. The resting average EMG amplitude was calculated based on minimum 100 measurements. The maximum amplitude value was calculated on the same EMG section. The amplitude distribution histograms were constructed.

The resting EMG recordings were also used for assessing the burst muscle discharges For this purpose, the Fourier spectrum diagrams for the low frequency area were constructed. In doing this, part of the data with negative EMG amplitude meanings were discarded. Data with the positive EMG amplitude meanings underwent a Butterworth digital sinus low pass filtering with a band pass of 0-20 Hz As a result, the envelope of EMG amplitude was formed, which then served as the data array for fast Fourier transformation. The data for each lead were divided into several successive sections, each containing 512 points, which underwent the fast Fourier transformation. The obtained spectra were averaged, and the envelope of EMG amplitude frequency with a maximal power was determined (Fig 1). The statistical significance of the frequency peak was determined by means of constructing the 95% confidential intervals (M 2 S.D).

Diagnosis of Parkinson's Disease by Electrophysiological Methods 31

average EMG amplitude values for mm. biceps and triceps brachii varied across the subjects within 3-12 V and the maximal amplitude values within 4-34 V. For the whole group of healthy subjects, mean ± SE average amplitude did not exceed 5.9 ± 0.2 V and mean ± SE

Table 1. Resting EMG amplitudes of shoulder muscles in healthy controls and patients with Parkinson's disease. Notes: values are Mean Standard Error; n - number of subjects in each

There were no statistical differences in EMG amplitude values between men and women. A significant positive correlation (p<0.05) was found for the resting activities of the antagonist muscles of the upper extremities in the age-matched healthy subjects (Table 2). The use of the envelope EMG construction technique demonstrated the presence of rhythmic burst muscle discharges in those cases where they were badly visualized on the EMG. Low amplitude burst discharges were occasionally identified in healthy subjects with the help of this technique (Fig 2, Control 2)**.** Of EMG recordings taken from flexors and extensors on both sides in 42 persons of the control group, 11 recordings (6.5%) made in eight subjects displayed the burst discharges, with a mean ± SE frequency of 6.1 ± 0.3 Hz. The maximal amplitude of burst

Healthy controls PD patients, OFF state PD patients, ON state

0.42\*\* 0.40\*\* 0.27 0.33\* - 0.26 0.32\*

In the group of PD patients, we observed a significant increase in the resting EMG amplitudes, which was ascribed to muscle relaxation disturbances. The mean average amplitude values for various study muscles, estimated for a whole PD group, were 2-3 times

Table 2. Correlation coefficients between average EMG amplitudes of the antagonist shoulder muscles (mm. biceps and triceps brachii) at rest in healthy controls and PD

less impaired side

more impaired side

less impaired side

group; \*, \*\*\* significant difference compared to healthy controls according to Mann-

discharges in control subjects did not exceed 11-18 V (mean ± SE = 14.3 ± 1.1 V).

more impaired side

patients. \* significant correlation, p< 0.05; \*\* is . p< 0.01.

5.30.8 5.50.7 5.90.2 5.40.8

12.52.1 12,82.3 11.91.3 12.32.1

Healthy controls (n = 42) PD patients (n = 48)

19.53.8 \*\*\* 21.47.4 \*\*\* 13.11.7 \* 12.53.1

74.017.9 \*\*\* 73.815.6 \*\*\* 60.17.4 61.912.0

maximal amplitude was no more 12.8 ± 2.3 V (Table 1).

m. biceps dexter m. biceps sinister m. triceps dexter m. triceps sinister

m. biceps dexter m. biceps sinister m. triceps dexter m. triceps sinister

Whitney test, p< 0.05 and p< 0.001, respectively.

right side left side

Average EMG amplitude (V)

Maximal EMG amplitude (V)

The EMG recordings made during the voluntary m. biceps brachii contraction served to calculate the phasic activation coefficient (PhAC), from a formula :

$$\text{PhAC} = \text{(AVCa-ARa)} / \text{AVCa} \tag{1}$$

where AVCa is the average EMG amplitude on a section with most marked changes in the muscle activity under voluntary contraction, and ARa is the average amplitude of resting EMG. In the conditions of low tonic muscle activity at rest the phasic muscle activation during voluntary movement is facilitated and the PhAC value is close to 1. On the contrary, when the resting tonic activity increases, the PhAC falls.

Fig. 1. The spectrogram of the power of EMG envelope frequency with 95% confidence intervals (A) and the corresponding resting EMG pattern of m. biceps brachii in a patient with Parkinson's disease (B). Peak deviation on the spectrogram reflects the frequency of the rhythmic burst muscle discharges.

The functional test with weight holding served for study of the reflex agonist and antagonist involvement during tonic muscle strain. We calculated the coefficients of reflex involvement of the muscles of the opposite arm, which characterized "distant" synergies. The coefficients of reflex involvement, respectively for the m. biceps brachii and m. triceps brachii of the opposite upper limb were obtained by calculation of the ratio of the mean amplitude recorded from the m. biceps (or triceps) on the resting side and mean amplitude recorded from the m. biceps on the side of retention of a load; the latter value was taken as 100%. In the norm, the value should not exceed 15%. An increase in this index is indicative of abnormal intensification of muscle coordinative interactions; if the coefficient of reflex involvement value exceeds 50%, such a disorder is qualified as gross.

Statistical analysis of the obtained data was performed using Statistic 8 software. Dispersion analysis ANOVA and a non-parametric two-tailed Mann-Whitney criterion were used in the course of comparison of the values observed in the different groups of the tested persons. Data obtained from the same patients before and after Levodopa treatment were compared using two-tailed paired t-test. The nonparametric Spearman test was used to evaluate possible correlation between above EMG parameters and subitems of UPDRS. Differences were considered to be significant at *Р* < 0.05.

#### **2.1.2 Results and discussion**

In the group of age-matched healthy subjects, the EMG amplitude of the shoulder flexors and extensors during muscle relaxation showed low values (Fig 2, Controls 1, 2). The

The EMG recordings made during the voluntary m. biceps brachii contraction served to

where AVCa is the average EMG amplitude on a section with most marked changes in the muscle activity under voluntary contraction, and ARa is the average amplitude of resting EMG. In the conditions of low tonic muscle activity at rest the phasic muscle activation during voluntary movement is facilitated and the PhAC value is close to 1. On the contrary,

Fig. 1. The spectrogram of the power of EMG envelope frequency with 95% confidence intervals (A) and the corresponding resting EMG pattern of m. biceps brachii in a patient with Parkinson's disease (B). Peak deviation on the spectrogram reflects the frequency of the

involvement value exceeds 50%, such a disorder is qualified as gross.

The functional test with weight holding served for study of the reflex agonist and antagonist involvement during tonic muscle strain. We calculated the coefficients of reflex involvement of the muscles of the opposite arm, which characterized "distant" synergies. The coefficients of reflex involvement, respectively for the m. biceps brachii and m. triceps brachii of the opposite upper limb were obtained by calculation of the ratio of the mean amplitude recorded from the m. biceps (or triceps) on the resting side and mean amplitude recorded from the m. biceps on the side of retention of a load; the latter value was taken as 100%. In the norm, the value should not exceed 15%. An increase in this index is indicative of abnormal intensification of muscle coordinative interactions; if the coefficient of reflex

Statistical analysis of the obtained data was performed using Statistic 8 software. Dispersion analysis ANOVA and a non-parametric two-tailed Mann-Whitney criterion were used in the course of comparison of the values observed in the different groups of the tested persons. Data obtained from the same patients before and after Levodopa treatment were compared using two-tailed paired t-test. The nonparametric Spearman test was used to evaluate possible correlation between above EMG parameters and subitems of UPDRS. Differences

In the group of age-matched healthy subjects, the EMG amplitude of the shoulder flexors and extensors during muscle relaxation showed low values (Fig 2, Controls 1, 2). The

PhAC = (AVCa-ARa)/AVCa (1)

calculate the phasic activation coefficient (PhAC), from a formula :

when the resting tonic activity increases, the PhAC falls.

rhythmic burst muscle discharges.

were considered to be significant at *Р* < 0.05.

**2.1.2 Results and discussion** 

average EMG amplitude values for mm. biceps and triceps brachii varied across the subjects within 3-12 V and the maximal amplitude values within 4-34 V. For the whole group of healthy subjects, mean ± SE average amplitude did not exceed 5.9 ± 0.2 V and mean ± SE maximal amplitude was no more 12.8 ± 2.3 V (Table 1).


Table 1. Resting EMG amplitudes of shoulder muscles in healthy controls and patients with Parkinson's disease. Notes: values are Mean Standard Error; n - number of subjects in each group; \*, \*\*\* significant difference compared to healthy controls according to Mann-Whitney test, p< 0.05 and p< 0.001, respectively.

There were no statistical differences in EMG amplitude values between men and women. A significant positive correlation (p<0.05) was found for the resting activities of the antagonist muscles of the upper extremities in the age-matched healthy subjects (Table 2). The use of the envelope EMG construction technique demonstrated the presence of rhythmic burst muscle discharges in those cases where they were badly visualized on the EMG. Low amplitude burst discharges were occasionally identified in healthy subjects with the help of this technique (Fig 2, Control 2)**.** Of EMG recordings taken from flexors and extensors on both sides in 42 persons of the control group, 11 recordings (6.5%) made in eight subjects displayed the burst discharges, with a mean ± SE frequency of 6.1 ± 0.3 Hz. The maximal amplitude of burst discharges in control subjects did not exceed 11-18 V (mean ± SE = 14.3 ± 1.1 V).


Table 2. Correlation coefficients between average EMG amplitudes of the antagonist shoulder muscles (mm. biceps and triceps brachii) at rest in healthy controls and PD patients. \* significant correlation, p< 0.05; \*\* is . p< 0.01.

In the group of PD patients, we observed a significant increase in the resting EMG amplitudes, which was ascribed to muscle relaxation disturbances. The mean average amplitude values for various study muscles, estimated for a whole PD group, were 2-3 times

Diagnosis of Parkinson's Disease by Electrophysiological Methods 33

Of interest is the fact that PD patients did not show a significant correlation between resting activities of the antagonist mm. biceps and triceps brachii on the dominant side, as has been true in the cases of the right and left sides in healthy subjects (Table 2). On the EMGs taken from PD patients with the akinetic-rigid-trembling form of the disease burst muscle discharges occurred as a rule with a frequency of 4-8 Hz. The mean ± SE frequency of the burst discharges was 5.2 ± 0.2 Hz. They had high amplitude (Fig.3). The maximal amplitude of burst discharges varied from 30 to 508 V (mean ± SE = 89.8 ± 15.6 V). Of special note, no significant correlation was found between resting EMG amplitude values and the

When comparing the data in OFF-state and ON-state, we observed a noticeable decrease in amplitude values. Mean ± SE average EMG amplitude of different muscles decreased to 8.2 ± 1.9 – 12.3 ± 5.1 V and mean ± SE maximal amplitude decreased to 20.1± 5.3 – 32.1± 7.1 V In this respect, the amplitude histograms made for the same patient during both states were

Fig. 4. The decrease of resting EMG amplitude values (calculated from peak to peak) of m. biceps brachii in a PD patient after intake of single dose of levodopa/carbidopa. OFF: the histograms of distribution of EMG amplitude values during the off-medication state; ON :

But the treatment with levodopa did not result in the normalization of correlations between resting activities of the antagonist shoulder muscles (Table 2). Following a single dose of levodopa/carbidopa the number of cases displaying burst discharges with a frequency of 4- 8 Hz decreased. In some patients the rhythmic discharges disappeared, as is shown in Figure 5, top records. In the other patients who displayed discharges after a dose of

EMGs recorded during the performance of voluntary arm bending were found to differ considerably in healthy subjects and PD patients. In the healthy subjects, we clearly distinguished an onset of muscle phasic activation on the EMG (Fig. 6, Control). Means ± SE of average EMG amplitude of the mm. biceps brachii during their voluntary contraction was 49 ± 8 V from the right and 44 ± 2 V from the left, the maximal amplitude – 189 ± 31 V and 137 ± 30 V, respectively. In view of the low resting EMG amplitude value in healthy subjects, the coefficient of phasic activation in most cases was equal to 0.7-0.9; mean ± SE

In contrast, in the PD patients it was often difficult to locate a site on the EMG at which the muscle phasic activation started because of increased resting tonic muscle activity and a very delayed rise in EMG amplitude after the delivery of a command to move (Fig. 6, PD). During peak voluntary flexor contraction, some patients showed a noticeable reduction of

levodopa, an increase occurred in the discharges frequency (Fig. 5, lower records).

one hour after drug intake. The bin of histograms is 4 V.

occurrence of burst muscle discharges.

very illustrative (Fig.4).

was 0.77 ± 0.04.

greater compared to control values, and the mean maximal amplitude values were approximately 5-6 times greater (Table 1). In some patients, average amplitude from a more impaired side reached 44-123 V and maximal amplitude - 210-508 V

Fig. 2. The spectrograms of the power of EMG envelope frequency with 95% confidence intervals (A) and the corresponding resting EMG patterns of m. biceps brachii (B) in healthy controls (Control 1 and 2) Peak deviation on the spectrogram reflects the frequency of the rhythmic burst muscle discharges. Control 1: the absence of the burst muscle activity. Control 2: low amplitude burst muscle discharges with a frequency of 4 Hz in a healthy control subject.

Fig. 3. Typical examples of EMGs registered in the resting state from three patients with the akinetic-rigid trembling form of Parkinson's disease (three upper records) and a patient with the akinetic-rigid form of this disease (low record).

greater compared to control values, and the mean maximal amplitude values were approximately 5-6 times greater (Table 1). In some patients, average amplitude from a more

Fig. 2. The spectrograms of the power of EMG envelope frequency with 95% confidence intervals (A) and the corresponding resting EMG patterns of m. biceps brachii (B) in healthy controls (Control 1 and 2) Peak deviation on the spectrogram reflects the frequency of the rhythmic burst muscle discharges. Control 1: the absence of the burst muscle activity. Control 2: low amplitude burst muscle discharges with a frequency of 4 Hz in a healthy

Fig. 3. Typical examples of EMGs registered in the resting state from three patients with the akinetic-rigid trembling form of Parkinson's disease (three upper records) and a patient

with the akinetic-rigid form of this disease (low record).

control subject.

impaired side reached 44-123 V and maximal amplitude - 210-508 V

Of interest is the fact that PD patients did not show a significant correlation between resting activities of the antagonist mm. biceps and triceps brachii on the dominant side, as has been true in the cases of the right and left sides in healthy subjects (Table 2). On the EMGs taken from PD patients with the akinetic-rigid-trembling form of the disease burst muscle discharges occurred as a rule with a frequency of 4-8 Hz. The mean ± SE frequency of the burst discharges was 5.2 ± 0.2 Hz. They had high amplitude (Fig.3). The maximal amplitude of burst discharges varied from 30 to 508 V (mean ± SE = 89.8 ± 15.6 V). Of special note, no significant correlation was found between resting EMG amplitude values and the occurrence of burst muscle discharges.

When comparing the data in OFF-state and ON-state, we observed a noticeable decrease in amplitude values. Mean ± SE average EMG amplitude of different muscles decreased to 8.2 ± 1.9 – 12.3 ± 5.1 V and mean ± SE maximal amplitude decreased to 20.1± 5.3 – 32.1± 7.1 V In this respect, the amplitude histograms made for the same patient during both states were very illustrative (Fig.4).

Fig. 4. The decrease of resting EMG amplitude values (calculated from peak to peak) of m. biceps brachii in a PD patient after intake of single dose of levodopa/carbidopa. OFF: the histograms of distribution of EMG amplitude values during the off-medication state; ON : one hour after drug intake. The bin of histograms is 4 V.

But the treatment with levodopa did not result in the normalization of correlations between resting activities of the antagonist shoulder muscles (Table 2). Following a single dose of levodopa/carbidopa the number of cases displaying burst discharges with a frequency of 4- 8 Hz decreased. In some patients the rhythmic discharges disappeared, as is shown in Figure 5, top records. In the other patients who displayed discharges after a dose of levodopa, an increase occurred in the discharges frequency (Fig. 5, lower records).

EMGs recorded during the performance of voluntary arm bending were found to differ considerably in healthy subjects and PD patients. In the healthy subjects, we clearly distinguished an onset of muscle phasic activation on the EMG (Fig. 6, Control). Means ± SE of average EMG amplitude of the mm. biceps brachii during their voluntary contraction was 49 ± 8 V from the right and 44 ± 2 V from the left, the maximal amplitude – 189 ± 31 V and 137 ± 30 V, respectively. In view of the low resting EMG amplitude value in healthy subjects, the coefficient of phasic activation in most cases was equal to 0.7-0.9; mean ± SE was 0.77 ± 0.04.

In contrast, in the PD patients it was often difficult to locate a site on the EMG at which the muscle phasic activation started because of increased resting tonic muscle activity and a very delayed rise in EMG amplitude after the delivery of a command to move (Fig. 6, PD). During peak voluntary flexor contraction, some patients showed a noticeable reduction of

Diagnosis of Parkinson's Disease by Electrophysiological Methods 35

Fig. 6. The phasic activation of m. biceps brachii dexter during voluntary movement in healthy control and off-medication patient with Parkinson's disease. Two sections of native EMGs in a healthy 64-year old subject (Control) and in a 53-year old patient (PD). The vertical thick lines designate the moment of experimenter's command. The horizontal lines over EMGs designate the periods, at which the amplitude measurements of maximal muscular activity during voluntary contractions (AVC) were taken. The resting EMG

The described quantitative EMG parameters added fundamentally to the clinical motor characteristics of PD patients and correlated selectively with definite UPDRS subitems. Thus, the most significant increases in the resting average and maximal EMG amplitude and decreases in the coefficient of phasic activation were observed in patients whose part III UPDRS scores exceeded 50. At the same time, the greatest involvement of the reciprocal muscles during the functional test was noted in patients with high part II UPDRS scores on the upper extremity daily activity. These patients also, had the greatest dyskinesia (disability) scores. We established the following statistically significant correlations (see Table 3). The resting muscle amplitudes from the more affected side correlated positively with the upper extremity rigidity and general motor scores. We found negative correlations between the phasic activation coefficient of m. biceps brachii from the more affected side and the upper extremity rigidity and general motor scores. The antagonist muscle involvement during the tonic strain holding a weight correlated positively with the handwriting, cutting food, dressing score and dyskinesia (disability) score. Levodopa intake did not influence essentially on the correlations between the reflex muscle involvement and

We found the present computer EMG analysis to be a sensitive tool for the objective evaluation of PD symptoms and to quantify the efficacy of levodopa therapy. One of the most useful EMG signs of extrapyramidal insufficiency appears to be the resting EMG amplitude value. In PD, the resting EMGs in the OFF state were characterized by the splashes of high muscle activity in contrast to the flat EMGs seen in age-matched healthy subjects. The average resting EMG amplitude was 2-3 times and the maximal amplitude was 5-6 times greater in PD patients than in the control group. Of interest are our data indicating the statistically significant correlation between levels of resting activity in the antagonist muscles (m. biceps and triceps brachii) in healthy subjects and the lack of such a correlation in PD patients on the most affected side. These findings appear to be an objective EMG

amplitudes (AR) were measured during 5 s prior to command presentation.

UPDRS scores.

Fig. 5. Changes in the rhythmic (4-7 Hz) burst muscle discharges in patients with Parkinson's disease after the intake of an individual single dose of levodopa/carbidopa. The spectrograms of the power of EMG envelope frequency in two patients in an off-medication state (OFF) and one hour after drug intake (ON). After the drug intake, burst muscle discharges disappeared in one patient, while their frequency somewhat increased in another patient. For other notes see Fig. 1.

EMG amplitude, while other patients had the same meanings as healthy subjects. The coefficient of phasic activation of a voluntarily contracting mm biceps brachii in PD patients was generally decreased to 0.1-0.6 and, sometimes (when amplitude values at rest exceeded those that were observed during phasic activation), even had negative values; mean ± SE was 0.42 ± 0.08. The magnitude of the phasic activation decrease showed a distinct dependence on the side which was most affected. Reduction of the coefficient of phasic activation in the OFF patient group compared to healthy individuals was statistically significant (p<0.01).

Disturbance of coordinative muscle interactions was one more typical feature of the EMG recorded in Parkinson's patients. This was manifested in increased reflex involvement of the muscles of the opposite arm (distant synergy) at tonic tension of the m. biceps brachii at one of the sides within the period of retention of a load. In this group, the mean values of the coefficients of reflex involvement for the m. biceps brachii and m. triceps brachii of the opposite side exceeded 50%. As was already mentioned, these phenomena should be considered a gross disturbance of coordinative interactions. In control group the mean values of the coefficients of reflex involvement were 20-26%.

The dose of levodopa/carbidopa in PD patients produced a decrease in resting muscle activity parallel to an increase, more often, to normal values (0.8-0.9) of the coefficient of phasic activation during voluntary contraction of flexors. With regard to the reflex activation of the agonist and antagonist muscles during weight holding, a noticeable reduction of the coefficients of reflex involvement was only observed in some of the patients who took levodopa/carbidopa treatment. In seven patients, after medication even more marked enhancement developed in the agonist or antagonist muscles, and was registered during the performance of the functional test.

Fig. 5. Changes in the rhythmic (4-7 Hz) burst muscle discharges in patients with

patient. For other notes see Fig. 1.

significant (p<0.01).

Parkinson's disease after the intake of an individual single dose of levodopa/carbidopa. The spectrograms of the power of EMG envelope frequency in two patients in an off-medication state (OFF) and one hour after drug intake (ON). After the drug intake, burst muscle

discharges disappeared in one patient, while their frequency somewhat increased in another

EMG amplitude, while other patients had the same meanings as healthy subjects. The coefficient of phasic activation of a voluntarily contracting mm biceps brachii in PD patients was generally decreased to 0.1-0.6 and, sometimes (when amplitude values at rest exceeded those that were observed during phasic activation), even had negative values; mean ± SE was 0.42 ± 0.08. The magnitude of the phasic activation decrease showed a distinct dependence on the side which was most affected. Reduction of the coefficient of phasic activation in the OFF patient group compared to healthy individuals was statistically

Disturbance of coordinative muscle interactions was one more typical feature of the EMG recorded in Parkinson's patients. This was manifested in increased reflex involvement of the muscles of the opposite arm (distant synergy) at tonic tension of the m. biceps brachii at one of the sides within the period of retention of a load. In this group, the mean values of the coefficients of reflex involvement for the m. biceps brachii and m. triceps brachii of the opposite side exceeded 50%. As was already mentioned, these phenomena should be considered a gross disturbance of coordinative interactions. In control group the mean

The dose of levodopa/carbidopa in PD patients produced a decrease in resting muscle activity parallel to an increase, more often, to normal values (0.8-0.9) of the coefficient of phasic activation during voluntary contraction of flexors. With regard to the reflex activation of the agonist and antagonist muscles during weight holding, a noticeable reduction of the coefficients of reflex involvement was only observed in some of the patients who took levodopa/carbidopa treatment. In seven patients, after medication even more marked enhancement developed in the agonist or antagonist muscles, and was registered

values of the coefficients of reflex involvement were 20-26%.

during the performance of the functional test.

Fig. 6. The phasic activation of m. biceps brachii dexter during voluntary movement in healthy control and off-medication patient with Parkinson's disease. Two sections of native EMGs in a healthy 64-year old subject (Control) and in a 53-year old patient (PD). The vertical thick lines designate the moment of experimenter's command. The horizontal lines over EMGs designate the periods, at which the amplitude measurements of maximal muscular activity during voluntary contractions (AVC) were taken. The resting EMG amplitudes (AR) were measured during 5 s prior to command presentation.

The described quantitative EMG parameters added fundamentally to the clinical motor characteristics of PD patients and correlated selectively with definite UPDRS subitems. Thus, the most significant increases in the resting average and maximal EMG amplitude and decreases in the coefficient of phasic activation were observed in patients whose part III UPDRS scores exceeded 50. At the same time, the greatest involvement of the reciprocal muscles during the functional test was noted in patients with high part II UPDRS scores on the upper extremity daily activity. These patients also, had the greatest dyskinesia (disability) scores. We established the following statistically significant correlations (see Table 3). The resting muscle amplitudes from the more affected side correlated positively with the upper extremity rigidity and general motor scores. We found negative correlations between the phasic activation coefficient of m. biceps brachii from the more affected side and the upper extremity rigidity and general motor scores. The antagonist muscle involvement during the tonic strain holding a weight correlated positively with the handwriting, cutting food, dressing score and dyskinesia (disability) score. Levodopa intake did not influence essentially on the correlations between the reflex muscle involvement and UPDRS scores.

We found the present computer EMG analysis to be a sensitive tool for the objective evaluation of PD symptoms and to quantify the efficacy of levodopa therapy. One of the most useful EMG signs of extrapyramidal insufficiency appears to be the resting EMG amplitude value. In PD, the resting EMGs in the OFF state were characterized by the splashes of high muscle activity in contrast to the flat EMGs seen in age-matched healthy subjects. The average resting EMG amplitude was 2-3 times and the maximal amplitude was 5-6 times greater in PD patients than in the control group. Of interest are our data indicating the statistically significant correlation between levels of resting activity in the antagonist muscles (m. biceps and triceps brachii) in healthy subjects and the lack of such a correlation in PD patients on the most affected side. These findings appear to be an objective EMG

Diagnosis of Parkinson's Disease by Electrophysiological Methods 37

norm, and, in addition, a decrease in the active contraction amplitudes occurred in some of the patients. As a consequence, the phasic activation coefficient of the voluntarily contracting m. biceps brachii of the patients was generally reduced to 0.1-0.6 and even had negative values, while in healthy subjects the value of the phasic activation coefficient was in most cases 0.7-0.9. Our study data suggest that phasic activation coefficients represent a sufficiently informative index that may be used to quantify phasic muscle activity in PD. Involvement of the agonist and antagonist muscles appeared to be useful for establishing coorditating muscle relationships. We demonstrated significantly (p<0.05) greater activation of the agonist and antagonist muscles during m. biceps brachii tonic strain in PD patients compared to age-matched healthy subjects. The present findings confirm the results of other investigators who consider this fact to be a consequence of increased excitation in the motor

When examining the action of an individual dose of levodopa/carbidopa in PD patients with ON-OFF phenomenon, we observed distinct positive drug effects in the following EMG parameters: average and maximal EMG amplitudes at rest; the number of cases with registered rhythmic burst muscle discharges of 4-8 Hz; value of the phasic activation coefficient during voluntary muscle contraction. However, levodopa therapy didn't appear to be effective in terms of the normalization of coordinating agonist - antagonist muscle relationships, either at rest and during holding a weight. In some of patients on levodopa/carbidopa, we even observed an enhancement of the activation of agonist and antagonist muscles during above functional test. We believe that such an increased coactivation of the agonist and antagonist muscles is an objective indicator of risk for developing levodopa-induced dyskinesia in PD patients. According to the literature data, the latter is the result of hypersensitivity of the dopamine receptors in the nigrostriatal system or of a disturbed balance between the degrees of activation of D1 and D2 dopamine

The histograms of distribution of EMG amplitude values at rest are informative characteristic of muscle activity (Meigal, 2009). The histogram sharpness and statistical EMG parameters, such as range, variance and kurtosis reflect the magnitude of bioelectrical

Fig. 7. Example of the histogram of resting EMG amplitudes distribution in healthy subject. An artifact-free EMG recordings (no less than 10 seconds in duration) from the flexors and extensors (mm. biceps and triceps brachii), registered in 33 patients with PD and 24 age-

centers, caused by dopaminergic control failure (Kryzhanovsky et al, 2002).

**2.2 Statistics of EMG distribution in patients with Parkinson's disease** 

muscle signals and the level of motor unit synchronization.

receptors (Jenner, 1994).


manifestation of the disorganisation of the brain's neuronal excitatory-inhibitory processes and the loss of functional balance between the structures which regulate muscle tone, all of which are due to a neostriatal dopamine deficit (DeLong, 1990).

Table 3. Significant correlations of EMG indices with the UPDRS scores studied in 20 PD patients in which the clinical "ON-OFF" phenomenon was verified. Data on the EMG indices and the upper extremity rigidity score for a more impaired side are presented. \*- p<0.05; \*\* p<0.01; "ns "- not statistically significant (p0.05).

A distinguishing feature of the bioelectrical muscle activity at rest in PD patients was the presence of burst muscle discharges with a rhythm of 4-8 Hz. It should be noted that no significant correlation was found between resting EMG amplitude value and the occurrence of burst muscle discharges. This fact is consistent with the viewpoint that muscle rigidity and tremor at PD do not constitute symptoms which influence one another, and that different pathophysiological mechanisms underlie their origin (Furukawa et al., 1991; Otsuka et al., 1996).

The brain systems, regulating the tonic and phasic muscle activities, were shown to be antagonistically interrelated: the activation of the phasic processes is accompanied by an inhibition of the tonic impulses and, vice versa, the enhancement of tonic impulsation hampers the phasic activity (Houk, 1979). In PD, due to an increased tonic muscle activity at rest, increment in EMG amplitude during phasic activation was reduced relative to the

manifestation of the disorganisation of the brain's neuronal excitatory-inhibitory processes and the loss of functional balance between the structures which regulate muscle tone, all of

> Motor score (points 18-31)

Handwriting, cutting food, dressing score (points 8-10)

0.51 \* 0.50 \* ns ns

ns 0.67 \*\* ns ns


ns -0.64 \*\* ns ns

ns ns 0.55 \* 0.52\*

ns ns 0.52 \* ns

ns ns 0.57 \*\* 0.48\*

ns ns 0.53 \* 0.51\*

Table 3. Significant correlations of EMG indices with the UPDRS scores studied in 20 PD patients in which the clinical "ON-OFF" phenomenon was verified. Data on the EMG indices and the upper extremity rigidity score for a more impaired side are presented. \*- p<0.05; \*\* -

A distinguishing feature of the bioelectrical muscle activity at rest in PD patients was the presence of burst muscle discharges with a rhythm of 4-8 Hz. It should be noted that no significant correlation was found between resting EMG amplitude value and the occurrence of burst muscle discharges. This fact is consistent with the viewpoint that muscle rigidity and tremor at PD do not constitute symptoms which influence one another, and that different pathophysiological mechanisms underlie their origin (Furukawa et al., 1991;

The brain systems, regulating the tonic and phasic muscle activities, were shown to be antagonistically interrelated: the activation of the phasic processes is accompanied by an inhibition of the tonic impulses and, vice versa, the enhancement of tonic impulsation hampers the phasic activity (Houk, 1979). In PD, due to an increased tonic muscle activity at rest, increment in EMG amplitude during phasic activation was reduced relative to the

Dyskinesia (disability) score (point 33)

which are due to a neostriatal dopamine deficit (DeLong, 1990).

ON

OFF

ON

OFF

ON

OFF

ON

OFF

p<0.01; "ns "- not statistically significant (p0.05).

Upper extremity rigidity score (point 22)

EMG indices

of m. biceps at rest

Phasic activation coefficient of m .biceps during voluntary contraction

involvement under tonic strain of m. biceps of the opposite arm

Otsuka et al., 1996).

Average amplitude

M. biceps

M. triceps involvement under tonic strain of m. biceps of the opposite arm

norm, and, in addition, a decrease in the active contraction amplitudes occurred in some of the patients. As a consequence, the phasic activation coefficient of the voluntarily contracting m. biceps brachii of the patients was generally reduced to 0.1-0.6 and even had negative values, while in healthy subjects the value of the phasic activation coefficient was in most cases 0.7-0.9. Our study data suggest that phasic activation coefficients represent a sufficiently informative index that may be used to quantify phasic muscle activity in PD.

Involvement of the agonist and antagonist muscles appeared to be useful for establishing coorditating muscle relationships. We demonstrated significantly (p<0.05) greater activation of the agonist and antagonist muscles during m. biceps brachii tonic strain in PD patients compared to age-matched healthy subjects. The present findings confirm the results of other investigators who consider this fact to be a consequence of increased excitation in the motor centers, caused by dopaminergic control failure (Kryzhanovsky et al, 2002).

When examining the action of an individual dose of levodopa/carbidopa in PD patients with ON-OFF phenomenon, we observed distinct positive drug effects in the following EMG parameters: average and maximal EMG amplitudes at rest; the number of cases with registered rhythmic burst muscle discharges of 4-8 Hz; value of the phasic activation coefficient during voluntary muscle contraction. However, levodopa therapy didn't appear to be effective in terms of the normalization of coordinating agonist - antagonist muscle relationships, either at rest and during holding a weight. In some of patients on levodopa/carbidopa, we even observed an enhancement of the activation of agonist and antagonist muscles during above functional test. We believe that such an increased coactivation of the agonist and antagonist muscles is an objective indicator of risk for developing levodopa-induced dyskinesia in PD patients. According to the literature data, the latter is the result of hypersensitivity of the dopamine receptors in the nigrostriatal system or of a disturbed balance between the degrees of activation of D1 and D2 dopamine receptors (Jenner, 1994).

#### **2.2 Statistics of EMG distribution in patients with Parkinson's disease**

The histograms of distribution of EMG amplitude values at rest are informative characteristic of muscle activity (Meigal, 2009). The histogram sharpness and statistical EMG parameters, such as range, variance and kurtosis reflect the magnitude of bioelectrical muscle signals and the level of motor unit synchronization.

Fig. 7. Example of the histogram of resting EMG amplitudes distribution in healthy subject.

An artifact-free EMG recordings (no less than 10 seconds in duration) from the flexors and extensors (mm. biceps and triceps brachii), registered in 33 patients with PD and 24 age-

Diagnosis of Parkinson's Disease by Electrophysiological Methods 39

Patients with the akinetic-rigid-trembling form of PD had the highest values of the EMG statistical parameters. The histograms of EMG amplitude distribution had a sharp peak (fig. 8, B). In some patients of this PD group range reached 382 V, variance – 951 and kurtosis – 13 (table 4). Correlation analyses revealed statistically significant connection of kurtosis with scores of the point 20 of UPDRS, estimating intensity of tremor of the hand, on which EMG was registered. A coefficient of nonparametric Spearman rank-order correlation between these indices was 0.46 (p<0.01). This fact is in accordance with the point of view (Meigal, 2009) that kurtosis well reflects synchronization of motor units responsible for the origin of

Fractal analysis is a new method for biomedical signal processing. Nonlinear analysis techniques are necessary to understand the complexity of the EMG. Study of fractal dynamics of EMG data is based on detrended fluctuation analysis and calculation of Hurst exponent. The Hurst exponent is used as a measure of the long term memory of time series,

Fractal dynamics of EMG signals was studied in 33 patients with akinetic-rigid-trembling form of PD (mean ± SE age 62.1 2.6, range 48-77 years), 30 age-matched healthy subjects ( mean ± SE age 65.4 ± 1.9, range 57-78 years) and 20 persons of middle age ( mean ± SE age 48.3 ± 1.49, range 45-58 years). EMGs were recorded from m. biceps brachii at the side, where morbid affection was more expressed, at rest no less than 10 seconds in duration. The rescaled range was calculated for time series. The first step was calculation of the mean. Then mean-adjusted series were created and the cumulative deviate series were calculated

> 1 () ( ) *k i*

Where *z* is the mean and *z(i)* is the value from time series. Then the row of values *y (k), k = 1,…N* was divided into the segments of length *n,* and within the limits of each segment the equalization of stright, approximating the sequence of *y(k),* was defined by least squares method. It is considered that approximation of *yn (k )* is the local trend. Further a standard

Dependence *lg F(lg n)* was further built, the angle of slope of approximating line was

We identified three different patterns of surface EMG signals according to fractal dimension (Fig 9, 10): with one, two and three scaling regions, every of which is characteristic by own local exponent. In healthy subjects, the fractal dimension with two exponents was most frequently observed, in 60% among persons of middle age and in 50% among elderly individuals. One exponent was observed in 20% in both groups of healthy subjects and three exponents – in 20% and 30% in middle age and elderly, respectively. In patients with akineticrigid-trembling form of PD the fractal dimension of surface EMG signals with three exponents

determined and the value of Hurst index was estimated (Stanley et al., 1999).

was most characteristic (64%). One exponent did not occur in PD patients (Fig. 11).

(2)

<sup>1</sup> <sup>2</sup> ( ) [( ( ) ( )] *Fn yk y k <sup>n</sup> <sup>N</sup>* (3)

*i y k zz* 

**2.3 Fractal dynamics of EMGs in patients with Parkinson's disease** 

i.e., the autocorrelation of the time series. (Talebinejad et al., 2010).

burst muscle discharges.

from the formula:

deviation was created from the formula:

matched healthy subjects at rest, were analyzed by the computer programs "Origin 8" and "Statistic 8". EMG in healthy subjects was characterized by low amplitude, flat symmetric histogram (fig. 7) and small values of range, variance and kurtosis (table 4). Range did not exceed 20 V, variance – 7 and kurtosis – 0.4. In patients with the akinetic-rigid form of the disease, the amplitude of EMG signals was considerably increased because of impossibility of entire muscle relaxation. The mean values of EMG statistical parameters, estimated for this PD group, were significantly (p<0.001) augmented as compared to control. In some patients range amounted to 66 V, variance – 56 and kurtosis – 1.4 (table 4).

Fig. 8. Examples of the histograms of resting EMG amplitudes distribution in a patient with akinetic-rigid form of Parkinson's disease (A) and a patient with akinetic-rigid- trembling form of this disease (B).


Table 4. Statistical characteristics of EMG in patients with Parkinson's disease and agematched healthy subjects. EMG characteristics in patients were taken at the side where morbid affection was more expressed; in healthy persons such characteristics were taken at the side where higher values were observed. \*\* p <0.01, \*\*\* p <0.001 compared to control group. n is number of subjects in each group. In brackets the range of indices in different tested persons is presented.

matched healthy subjects at rest, were analyzed by the computer programs "Origin 8" and "Statistic 8". EMG in healthy subjects was characterized by low amplitude, flat symmetric histogram (fig. 7) and small values of range, variance and kurtosis (table 4). Range did not exceed 20 V, variance – 7 and kurtosis – 0.4. In patients with the akinetic-rigid form of the disease, the amplitude of EMG signals was considerably increased because of impossibility of entire muscle relaxation. The mean values of EMG statistical parameters, estimated for this PD group, were significantly (p<0.001) augmented as compared to control. In some

Fig. 8. Examples of the histograms of resting EMG amplitudes distribution in a patient with akinetic-rigid form of Parkinson's disease (A) and a patient with akinetic-rigid- trembling

> 112.75 ± 16.80 \*\*\* (24.48 – 381.62)

36.81 ± 3.96 \*\*\* (21.16 – 66.18)

11.18 ± 0.71 (7.08 – 16.46)

Table 4. Statistical characteristics of EMG in patients with Parkinson's disease and agematched healthy subjects. EMG characteristics in patients were taken at the side where morbid affection was more expressed; in healthy persons such characteristics were taken at the side where higher values were observed. \*\* p <0.01, \*\*\* p <0.001 compared to control group. n is number of subjects in each group. In brackets the range of indices in different

Statistical parameters Range (V) Variance Kurtosis

> 147.20 ± 44.38 \*\* (6.91 – 950.63)

20.26 ± 4.26 \*\*\* (7.89 – 56.16)

2.30 ± 0.31 (0.70 – 5.14) 4.32 ± 0.51 \*\*\* (1.04 – 12.98)

0.58 ± 0.14 \*\*\* (0.07 – 1.40)


patients range amounted to 66 V, variance – 56 and kurtosis – 1.4 (table 4).

form of this disease (B).

Groups of the tested

Patients with akineticrigid-trembling form of the

form of the disease

Control group of agematched healthy subjects

tested persons is presented.

Patients with akinetic-rigid

persons

disease n=20

n=13

n=24

Patients with the akinetic-rigid-trembling form of PD had the highest values of the EMG statistical parameters. The histograms of EMG amplitude distribution had a sharp peak (fig. 8, B). In some patients of this PD group range reached 382 V, variance – 951 and kurtosis – 13 (table 4). Correlation analyses revealed statistically significant connection of kurtosis with scores of the point 20 of UPDRS, estimating intensity of tremor of the hand, on which EMG was registered. A coefficient of nonparametric Spearman rank-order correlation between these indices was 0.46 (p<0.01). This fact is in accordance with the point of view (Meigal, 2009) that kurtosis well reflects synchronization of motor units responsible for the origin of burst muscle discharges.

#### **2.3 Fractal dynamics of EMGs in patients with Parkinson's disease**

Fractal analysis is a new method for biomedical signal processing. Nonlinear analysis techniques are necessary to understand the complexity of the EMG. Study of fractal dynamics of EMG data is based on detrended fluctuation analysis and calculation of Hurst exponent. The Hurst exponent is used as a measure of the long term memory of time series, i.e., the autocorrelation of the time series. (Talebinejad et al., 2010).

Fractal dynamics of EMG signals was studied in 33 patients with akinetic-rigid-trembling form of PD (mean ± SE age 62.1 2.6, range 48-77 years), 30 age-matched healthy subjects ( mean ± SE age 65.4 ± 1.9, range 57-78 years) and 20 persons of middle age ( mean ± SE age 48.3 ± 1.49, range 45-58 years). EMGs were recorded from m. biceps brachii at the side, where morbid affection was more expressed, at rest no less than 10 seconds in duration.

The rescaled range was calculated for time series. The first step was calculation of the mean. Then mean-adjusted series were created and the cumulative deviate series were calculated from the formula:

$$y(k) = \sum\_{i=1}^{k} (z\_i - \overline{z}) \tag{2}$$

Where *z* is the mean and *z(i)* is the value from time series. Then the row of values *y (k), k = 1,…N* was divided into the segments of length *n,* and within the limits of each segment the equalization of stright, approximating the sequence of *y(k),* was defined by least squares method. It is considered that approximation of *yn (k )* is the local trend. Further a standard deviation was created from the formula:

$$F(n) = \sqrt{\frac{1}{N} [(y(k) - y\_n(k))^2]} \tag{3}$$

Dependence *lg F(lg n)* was further built, the angle of slope of approximating line was determined and the value of Hurst index was estimated (Stanley et al., 1999).

We identified three different patterns of surface EMG signals according to fractal dimension (Fig 9, 10): with one, two and three scaling regions, every of which is characteristic by own local exponent. In healthy subjects, the fractal dimension with two exponents was most frequently observed, in 60% among persons of middle age and in 50% among elderly individuals. One exponent was observed in 20% in both groups of healthy subjects and three exponents – in 20% and 30% in middle age and elderly, respectively. In patients with akineticrigid-trembling form of PD the fractal dimension of surface EMG signals with three exponents was most characteristic (64%). One exponent did not occur in PD patients (Fig. 11).

Diagnosis of Parkinson's Disease by Electrophysiological Methods 41

significantly (p<0.001) differed from same value in elderly subjects (2.7 ± 0.5). Negative

Fig. 11. Comparison of the incidence of different patterns of surface EMG signals fractal dimension (with one, two or three Hurst exponents) in persons of middle age (control 1),

Overall, the present investigation has demonstrated the following distinctive features of surface EMG signals fractal dimension in patients with akinetic-rigid-trembling form of PD: 1) correlation behavior of the resting EMG time series in patients was more complex compared to healthy subjects and often suggested three scaling regions; 2) the value of Hurst exponent was significantly lower in patients, its value may descend to 0.1 - 0.2 that indicates a time series with negative autocorrelation (e.g. a decrease between values will probably be followed by an increase); 3) considerable degradation of short and longer range correlation properties, that, perhaps, is associated with the loss of integrated physiological

**2.4 EMG characteristics of clinically healthy kinsmen of the patients with Parkinson's** 

According to modern concepts, the genetic factor plays a considerable role in the development of Parkinson's disease. Modifications in several genetic loci responsible for the development of this disease have been identified. The considerable role of the genetic factor for the propensity to Parkinson's disease has been confirmed by the data of epidemiological studies. The frequency of development of this disease in kinsmen of Parkinsonian patients is two to seven times higher than that in persons of the control groups (Elbaz et al., 1999). Symptoms of functional insufficiency of the extrapyramidal system can be identified very early, namely several decades prior to possible onset of the development of the clinical form of Parkinson's disease (Berg et al., 2002). Prevention or deceleration of the development of this disease can be provided by the detection of the early, presymptomatic stage of the neurodegenerative process and identification of informative "biomarkers" of PD (Illarioshkin, 2008). We studied surface EMG in clinically healthy kinsmen of the patients

elderly subjects (control 2) and patients with Parkinson's disease (PD).

responsiveness at this disease (Goldberger et al., 2002).

**disease** 

correlation between H1 and H3 (r = -0.67, p<0.01) was revealed in PD patients.

Fig. 9. Patterns of fractal dimension of the surface EMG signals with one and two exponents. Thick line is general exponent, thin lines are local exponents. H is general Hurst index; H1, H2, H3 are values of local Hurst indices.

Fig. 10. Pattern of fractal dimension of the surface EMG signals with three exponents. Thick line is general exponent, thin lines are local exponents. H is general Hurst index; H1, H2, H3 are values of local Hurst indices.

Another difference concerned the value of general Hurst index (H). In persons of middle age mean value of H was 0.47 ± 0.02 (range 0.32 – 0.71) and in elderly – 0.44 ± 0.02 (range 0.33 – 0.57). In PD patients mean value of H was significantly (p<0.01) lower as compared to elderly subjects – 0.31 ± 0.03 (range 0.09 – 0.49). In PD patients, the value of Hurst index of the third scaling region (H3) in patterns with three exponents also significantly differed from H3 in healthy subjects. H3 was 0.30 ± 0.06 in middle-aged persons, 0.39 ± 0.05 in elderly and 0.14 ± 0.05 in patients with PD (p<0.05). It is of interest that the tendency to negative correlation between H and motor scores of part III UPDRS was observed in patients with PD (r = -0.35, p =0.05).

Our data showed essential alterations in short and more long-range EMG correlation properties in patients with akinetic-rigid-trembling form of PD. The mean value of H1/H3 in patterns with three exponents in the group of PD patients came up to 27.2 ± 9.5 that

Fig. 9. Patterns of fractal dimension of the surface EMG signals with one and two exponents. Thick line is general exponent, thin lines are local exponents. H is general Hurst index; H1,

Fig. 10. Pattern of fractal dimension of the surface EMG signals with three exponents. Thick line is general exponent, thin lines are local exponents. H is general Hurst index; H1, H2, H3

Another difference concerned the value of general Hurst index (H). In persons of middle age mean value of H was 0.47 ± 0.02 (range 0.32 – 0.71) and in elderly – 0.44 ± 0.02 (range 0.33 – 0.57). In PD patients mean value of H was significantly (p<0.01) lower as compared to elderly subjects – 0.31 ± 0.03 (range 0.09 – 0.49). In PD patients, the value of Hurst index of the third scaling region (H3) in patterns with three exponents also significantly differed from H3 in healthy subjects. H3 was 0.30 ± 0.06 in middle-aged persons, 0.39 ± 0.05 in elderly and 0.14 ± 0.05 in patients with PD (p<0.05). It is of interest that the tendency to negative correlation between H and motor scores of part III UPDRS was observed in

Our data showed essential alterations in short and more long-range EMG correlation properties in patients with akinetic-rigid-trembling form of PD. The mean value of H1/H3 in patterns with three exponents in the group of PD patients came up to 27.2 ± 9.5 that

H2, H3 are values of local Hurst indices.

are values of local Hurst indices.

patients with PD (r = -0.35, p =0.05).

significantly (p<0.001) differed from same value in elderly subjects (2.7 ± 0.5). Negative correlation between H1 and H3 (r = -0.67, p<0.01) was revealed in PD patients.

Fig. 11. Comparison of the incidence of different patterns of surface EMG signals fractal dimension (with one, two or three Hurst exponents) in persons of middle age (control 1), elderly subjects (control 2) and patients with Parkinson's disease (PD).

Overall, the present investigation has demonstrated the following distinctive features of surface EMG signals fractal dimension in patients with akinetic-rigid-trembling form of PD: 1) correlation behavior of the resting EMG time series in patients was more complex compared to healthy subjects and often suggested three scaling regions; 2) the value of Hurst exponent was significantly lower in patients, its value may descend to 0.1 - 0.2 that indicates a time series with negative autocorrelation (e.g. a decrease between values will probably be followed by an increase); 3) considerable degradation of short and longer range correlation properties, that, perhaps, is associated with the loss of integrated physiological responsiveness at this disease (Goldberger et al., 2002).

#### **2.4 EMG characteristics of clinically healthy kinsmen of the patients with Parkinson's disease**

According to modern concepts, the genetic factor plays a considerable role in the development of Parkinson's disease. Modifications in several genetic loci responsible for the development of this disease have been identified. The considerable role of the genetic factor for the propensity to Parkinson's disease has been confirmed by the data of epidemiological studies. The frequency of development of this disease in kinsmen of Parkinsonian patients is two to seven times higher than that in persons of the control groups (Elbaz et al., 1999). Symptoms of functional insufficiency of the extrapyramidal system can be identified very early, namely several decades prior to possible onset of the development of the clinical form of Parkinson's disease (Berg et al., 2002). Prevention or deceleration of the development of this disease can be provided by the detection of the early, presymptomatic stage of the neurodegenerative process and identification of informative "biomarkers" of PD (Illarioshkin, 2008). We studied surface EMG in clinically healthy kinsmen of the patients

Diagnosis of Parkinson's Disease by Electrophysiological Methods 43

Fig. 12. Types of burst-like muscle discharges with frequency 6, 8 and 10 Hz recorded in

Statistical parameters Number of tested persons

Variance [8.08 – 13.55] 5 (14 %)

Kurtosis [0.67; 0.70] 2 (5 %)

Total 16 (43 %)

Table 5. Incidence of resting EMG statistical parameters, going out outside a norm, in kinsmen of patients with Parkinson's disease. In square brackets the range of indices in

The data obtained in our work agree with findings of other authors who emphasized that data obtained using EMG techniques are of a high informative value in the diagnostics of subclinical manifestations of weakening of the supraspinal control (Robichaud et al., 2009), and that the genetic factor responsible for the propensity for development of the extrapyramidal insufficiency is rather important (Elbaz et al., 1999; Illarioshkin, 2002). A single common pathogenetic factor, namely conformational modifications of some cellular proteins at a post-translational stage of their synthesis, underlies most neurodegenerative diseases, including PD. Due to the existence of powerful compensatory and detoxication

9 (24 %)

three kinsmen of patients suffering from Parkinson's disease.

Range [22.46 – 113.61 V] Variance [8.23 – 112.32] Kurtosis [0.43 – 16.30]

Range [24.22 – 28.33 V]

different persons is presented.

suffering from PD in order to detect latent symptoms of extrapyramidal insufficiency that can be considered genetic determinants of the risk of development of the above disease. The task of our study included estimation of the frequency of occurrence of muscle activity disorders in kinsmen of the patients, characterization of correlations between the appearance symptoms of extrapyramidal insufficiency and age of the tested persons, and formulation of recommendations for individuals belonging, from the aspect of risk of development of PD, to a risk group.

We examined two groups of persons. The first group consisted of 37 clinically healthy kinsmen/kinswomen of patients suffering from PD (children, brothers, and sisters; 22 women and 15 men aged 30 to 56; mean age 45.6 ± 1.5 years). The second group (control) included 30 healthy young and middle-aged persons (19 women and 11 men; age nearly corresponding to that of persons of the first group , i.e., from 34 to 58 years, mean age 46.9 ± 2.2 years. All examined persons gave informed consent to be involved in the study. We recorded surface EMG at rest using superficial bipolar electrodes fixed on the flexor and extensor of the elbow joint (m. biceps brachii and m. triceps brachii, respectively); an electroneuromyograph NeuroMPF (Russia) was used. The detailed description of method is presented in section 2.1.1.

In 9 (24%) clinically healthy kinsmen of PD patients symptoms of functional insufficiency of the extrapyramidal system were evident. They demonstrated the mean amplitude value of 5.4-12.4 V, maximal amplitude value of 25-93 V, and the mean power of EMG oscillations reached 0.85- 1.8 mV/sec. In control group the mean amplitude value varied from 3.4 to 5.0 V, maximal amplitude varied from 5.6 to 20.6 V and the mean power of EMG oscillations did not exceed 0.02-0.71 mV/sec. Higher values of the intensity of electrical muscle activity in kinsmen of PD patients positively correlated with their age; it should be noted that, in this respect, age older than 45 years can be considered to be critical. The number of elder (older than 45 years) subjects with values of the mean power of EMG oscillations higher than the mean value of this parameter in persons of control group exceeded significantly the number of elder persons with low values of the mean power of EMG oscillations (p < 0.05, χ<sup>2</sup> criterion). The correlation coefficient between the age of the tested persons and the value of mean power of EMG oscillations was 0.40 (p < 0.05).

In 6 (16%) kinsmen of PD patients short burst-like discharges consisting of two to three oscillations generated with a frequency of 5-10 Hz were observed within the resting EMG (Fig. 12). As a rule, the amplitude of these potentials did not exceed 52 V.

For more detailed investigation we used statistics of EMG distribution, namely such parameters as range, variance and kurtosis. Range and variance reflect the extent of bioelectrical muscle signals. Kurtosis characterizes motor unit synchronization. We supposed that statistical methods might appear effective for exposure of pathological signs of muscle activity. In control group of healthy middle-aged persons the extreme value of resting EMG amplitude range was 20 V, variance – 7 and kurtosis – 0.4. The parameters of range, variance and kurtosis were considered going out outside a norm, if they exceeded the extreme values of these indices in the control group. We found 16 (43 %) kinsmen of patients with PD, who had high statistical parameters of EMG signals. In 14 kinsmen (38 %) range and variance were augmented compared to the extreme values of these indices in the control group. In 11 kinsmen (29 %) kurtosis had higher values than normal, presumably, reflecting enhanced synchronization in activity of motor units (Table 5).

suffering from PD in order to detect latent symptoms of extrapyramidal insufficiency that can be considered genetic determinants of the risk of development of the above disease. The task of our study included estimation of the frequency of occurrence of muscle activity disorders in kinsmen of the patients, characterization of correlations between the appearance symptoms of extrapyramidal insufficiency and age of the tested persons, and formulation of recommendations for individuals belonging, from the aspect of risk of

We examined two groups of persons. The first group consisted of 37 clinically healthy kinsmen/kinswomen of patients suffering from PD (children, brothers, and sisters; 22 women and 15 men aged 30 to 56; mean age 45.6 ± 1.5 years). The second group (control) included 30 healthy young and middle-aged persons (19 women and 11 men; age nearly corresponding to that of persons of the first group , i.e., from 34 to 58 years, mean age 46.9 ± 2.2 years. All examined persons gave informed consent to be involved in the study. We recorded surface EMG at rest using superficial bipolar electrodes fixed on the flexor and extensor of the elbow joint (m. biceps brachii and m. triceps brachii, respectively); an electroneuromyograph NeuroMPF (Russia) was used. The detailed description of method is

In 9 (24%) clinically healthy kinsmen of PD patients symptoms of functional insufficiency of the extrapyramidal system were evident. They demonstrated the mean amplitude value of 5.4-12.4 V, maximal amplitude value of 25-93 V, and the mean power of EMG oscillations reached 0.85- 1.8 mV/sec. In control group the mean amplitude value varied from 3.4 to 5.0 V, maximal amplitude varied from 5.6 to 20.6 V and the mean power of EMG oscillations did not exceed 0.02-0.71 mV/sec. Higher values of the intensity of electrical muscle activity in kinsmen of PD patients positively correlated with their age; it should be noted that, in this respect, age older than 45 years can be considered to be critical. The number of elder (older than 45 years) subjects with values of the mean power of EMG oscillations higher than the mean value of this parameter in persons of control group exceeded significantly the number of elder persons with low values of the mean power of EMG oscillations (p < 0.05, χ<sup>2</sup> criterion). The correlation coefficient between the age of the tested persons and the value of

In 6 (16%) kinsmen of PD patients short burst-like discharges consisting of two to three oscillations generated with a frequency of 5-10 Hz were observed within the resting EMG

For more detailed investigation we used statistics of EMG distribution, namely such parameters as range, variance and kurtosis. Range and variance reflect the extent of bioelectrical muscle signals. Kurtosis characterizes motor unit synchronization. We supposed that statistical methods might appear effective for exposure of pathological signs of muscle activity. In control group of healthy middle-aged persons the extreme value of resting EMG amplitude range was 20 V, variance – 7 and kurtosis – 0.4. The parameters of range, variance and kurtosis were considered going out outside a norm, if they exceeded the extreme values of these indices in the control group. We found 16 (43 %) kinsmen of patients with PD, who had high statistical parameters of EMG signals. In 14 kinsmen (38 %) range and variance were augmented compared to the extreme values of these indices in the control group. In 11 kinsmen (29 %) kurtosis had higher values than normal, presumably,

(Fig. 12). As a rule, the amplitude of these potentials did not exceed 52 V.

reflecting enhanced synchronization in activity of motor units (Table 5).

development of PD, to a risk group.

presented in section 2.1.1.

mean power of EMG oscillations was 0.40 (p < 0.05).

Fig. 12. Types of burst-like muscle discharges with frequency 6, 8 and 10 Hz recorded in three kinsmen of patients suffering from Parkinson's disease.


Table 5. Incidence of resting EMG statistical parameters, going out outside a norm, in kinsmen of patients with Parkinson's disease. In square brackets the range of indices in different persons is presented.

The data obtained in our work agree with findings of other authors who emphasized that data obtained using EMG techniques are of a high informative value in the diagnostics of subclinical manifestations of weakening of the supraspinal control (Robichaud et al., 2009), and that the genetic factor responsible for the propensity for development of the extrapyramidal insufficiency is rather important (Elbaz et al., 1999; Illarioshkin, 2002). A single common pathogenetic factor, namely conformational modifications of some cellular

proteins at a post-translational stage of their synthesis, underlies most neurodegenerative diseases, including PD. Due to the existence of powerful compensatory and detoxication

Diagnosis of Parkinson's Disease by Electrophysiological Methods 45

Monopolar recordings were made of CNV from intermediate leads: frontal (Fz), central (Cz), and parietal (Pz). The indifferent electrode was located on the earlobe. The ground electrode was located on the left forearm. During studies, subjects were in a relaxed, calm state with the eyes closed. Bioelectrical signals were passed to an amplifier with a bandpass of 0.08–15 Hz and then to a computer hard disk. CNV was recorded using two sound stimuli of different intensities with a 1-sec interval: the ready signal was at 50 dB HL and the trigger signal was at 80 dB HL. The subject pressed a key in response to the trigger signal. Analysis was performed using computer programs. The sampling frequency was 200 Hz. The analysis time was 3.1 sec, the first 400 ms being a record of the baseline electroencephalogram. Mean initial activity was determined from an artefact-free part of the electroencephalogram trace. Averaging of 30 trials yielded: 1) the duration of CNV measured as the time interval between the start of the negative deviation from the baseline after the ready stimulus and the moment of presentation of the trigger stimulus (ms); 2) the areas of Bereitschaftpotential and negative slope, between the baseline and the negativity curve of the corresponding region *S* = (Σ*Ai*) × Δ*t* (mV·ms), where *Ai* is the amplitude of the negative deviation from the initial level at a sampling frequency of 200 Hz and Δ*t* is a time interval of duration 5 ms; 3) the mean amplitudes of Bereitschaftpotential and negative slope defined by A = ΣAi/n (V). The program also allowed calculation of the simple sensorimotor reaction time (the mean latent period of

Motor symptomatology was assessed quantitatively in patients with PD in points using the unified scale UPDRS. The total score was measured in each of three dimension scales: I (impairments in thought, mood), II (decreased daily activity, impairment of hygiene activities); and III (disturbances of motor function, including bradykinesia, rigidity, and tremor) using four-point subscales for each symptom. General cognitive status of the PD patients was characterized using the standard quantitative scale Mini Mental State Examination (MMSE). The overall assessment of mental functions in normal patients yielded 30 points. Decreases in

The state of coordinatory muscle interactions was studied in 29 patients with PD (aged 47– 72 years) by assessing the level of reciprocal involvement of the triceps brachii antagonist muscle (the extensor muscle of the shoulder) on functional loading of the biceps brachii muscle (the flexor of the shoulder) on the right side. EMGs were recorded using bipolar skin electrodes (0.5 × 1.0 cm2) with a constant interelectrode distance of 1.5 cm. Bioelectrical signals were passed to the amplifiers of a Medikor MG440 (Budapest) electromyography with a bandpass of 10 Hz to 10 kHz. Functional loading on the biceps brachii muscle was applied by holding a load of 2 kg on the elevated and forward extended arm for 5 sec. With the patient in the calm, relaxed state and during holding of the load, at least 100 measurements were processed using the computer program to determine the mean EMG amplitude in the triceps brachii at rest (*A*r) and loading (*Al*), the coefficient of reciprocal involvement of the antagonist muscle was calculated as *Al*/(*A*r + *Al*). This coefficient had a value of 0.5 when the amplitude on loading showed no change. If the amplitude decreased, then coefficient of reciprocal involvement had values of less than 0.5, while increases

The effects of cerebrolysin on measures of CNV were studied in 21 patients with PD that were taking antiparkinsonian therapy, which was not changed during one month before cerebrolysin treatment and under the whole cerebrolysin course (intravenously 10 ml, during 10 days). Before and after cerebrolysin treatment we studied clinical scores of

pressing the key in response to the trigger signal).

the total score to less than 25 were regarded as a sign of early dementia.

yielded coefficient of reciprocal involvement values of greater than 0.5.

UPDRS and CNV.

systems in cells, such units are capable of successfully "overcoming" abnormal protein substrates for many years (Sherman & Goldberg, 2001). Delayed manifestation of clinical symptoms of the above disease is a feature of "conformational" pathologies of the brain. Latent pathological process can run course up to 30 years (Kryzhanovskii et al., 1995). The rate of pathological modifications in nerve cells within the presymptomatic period of PD is relatively low, but neuronal death is intensified significantly with transition to the stage of manifestation of this disease (Antonini et al., 1998). In relation to the above data, it is quite obvious that early diagnostics of the existence of latent extrapyramidal insufficiency is of exceptional importance. To prevent manifestations of PD in persons belonging to the risk group with respect to the development of parkinsonism, certain basic recommendations should be taken into account. They included a description of the rational daily routine, a recommended dietary intake with an increased content of vitamin В6 (pyridoxine, which is the main catalyzer in the synthesis of dopamine), and also a list of the drugs whose longterm administration should be avoided. Among the latter drugs are therapeutic agents whose administration leads either to depletion of the regulatory function of the dopaminergic system or to an increase in the functional activity of this system. Among such agents are haloperidol, the indole reserpine, fluoxetine (Prozac), metoclopramide (Cerucal), clozapine, and Cordarone, as well as derivatives of phenothiazine and butyrophenol, and also lithium preparations.

### **3. Investigation of contingent negative variation**

Endogenous cortical movement-related reaction contingent negative variation (CNV) is a sensitive indicator for the objective evaluation of the severity of PD and quantifying the efficacy of antiparkinsonian therapy. Many authors identify two phases in CNV: an early phase, the Bereitschaftpotential, and a late phase, the negative slope (Filipovic et al., 1997). It has been suggested that these are generated by different brain structures: the cerebellar efferent system is more involved in generating the Bereitschaftpotential, while the basal ganglia are more involved in generating the negative slope (Ikeda et al., 1997). The question of the relationships between each of these phases and higher integrative processes and the mechanisms of direct motor control have received insufficient study. The aims of the present work were to study the extent to which the early and late phases of CNV depend on motor and mental functions and to determine what effect have neurotrophic agents on CNV. Tasks to be addressed were: 1) identification of the individual characteristics of the early and late phases of CNV in patients with PD as compared with subjects of similar age; 2) identification of correlation the measures of the two phases of CNV with clinical characteristics of the PD patients; 3) investigation the effect of the brain-derived peptide drug cerebrolysin on the amplitude characteristics of CNV in PD.

#### **3.1 Methods**

Studies were performed using 28 healthy subjects (13 male, 15 female, age 48–73 years, mean age 60.9 ± 1.2 years) and 56 patients with idiopathic PD (23 male, 33 female, age 45–74 years, mean age 61.3 ± 1.1 years). Patients were in stages 1.5–3.0 (2.2 ± 0.1) H-Y (international classification of Hoehn and Yaht, 1967). Patients received basic antiparkinsonism treatment with levodopa-containing agents (levodopa/carbidopa). Individual daily doses of levodopa were 250–750 mg. All subjects were right-handed.

systems in cells, such units are capable of successfully "overcoming" abnormal protein substrates for many years (Sherman & Goldberg, 2001). Delayed manifestation of clinical symptoms of the above disease is a feature of "conformational" pathologies of the brain. Latent pathological process can run course up to 30 years (Kryzhanovskii et al., 1995). The rate of pathological modifications in nerve cells within the presymptomatic period of PD is relatively low, but neuronal death is intensified significantly with transition to the stage of manifestation of this disease (Antonini et al., 1998). In relation to the above data, it is quite obvious that early diagnostics of the existence of latent extrapyramidal insufficiency is of exceptional importance. To prevent manifestations of PD in persons belonging to the risk group with respect to the development of parkinsonism, certain basic recommendations should be taken into account. They included a description of the rational daily routine, a recommended dietary intake with an increased content of vitamin В6 (pyridoxine, which is the main catalyzer in the synthesis of dopamine), and also a list of the drugs whose longterm administration should be avoided. Among the latter drugs are therapeutic agents whose administration leads either to depletion of the regulatory function of the dopaminergic system or to an increase in the functional activity of this system. Among such agents are haloperidol, the indole reserpine, fluoxetine (Prozac), metoclopramide (Cerucal), clozapine, and Cordarone, as well as derivatives of phenothiazine and butyrophenol, and

Endogenous cortical movement-related reaction contingent negative variation (CNV) is a sensitive indicator for the objective evaluation of the severity of PD and quantifying the efficacy of antiparkinsonian therapy. Many authors identify two phases in CNV: an early phase, the Bereitschaftpotential, and a late phase, the negative slope (Filipovic et al., 1997). It has been suggested that these are generated by different brain structures: the cerebellar efferent system is more involved in generating the Bereitschaftpotential, while the basal ganglia are more involved in generating the negative slope (Ikeda et al., 1997). The question of the relationships between each of these phases and higher integrative processes and the mechanisms of direct motor control have received insufficient study. The aims of the present work were to study the extent to which the early and late phases of CNV depend on motor and mental functions and to determine what effect have neurotrophic agents on CNV. Tasks to be addressed were: 1) identification of the individual characteristics of the early and late phases of CNV in patients with PD as compared with subjects of similar age; 2) identification of correlation the measures of the two phases of CNV with clinical characteristics of the PD patients; 3) investigation the effect of the brain-derived peptide

Studies were performed using 28 healthy subjects (13 male, 15 female, age 48–73 years, mean age 60.9 ± 1.2 years) and 56 patients with idiopathic PD (23 male, 33 female, age 45–74 years, mean age 61.3 ± 1.1 years). Patients were in stages 1.5–3.0 (2.2 ± 0.1) H-Y (international classification of Hoehn and Yaht, 1967). Patients received basic antiparkinsonism treatment with levodopa-containing agents (levodopa/carbidopa). Individual daily doses of levodopa

also lithium preparations.

**3.1 Methods** 

**3. Investigation of contingent negative variation** 

drug cerebrolysin on the amplitude characteristics of CNV in PD.

were 250–750 mg. All subjects were right-handed.

Monopolar recordings were made of CNV from intermediate leads: frontal (Fz), central (Cz), and parietal (Pz). The indifferent electrode was located on the earlobe. The ground electrode was located on the left forearm. During studies, subjects were in a relaxed, calm state with the eyes closed. Bioelectrical signals were passed to an amplifier with a bandpass of 0.08–15 Hz and then to a computer hard disk. CNV was recorded using two sound stimuli of different intensities with a 1-sec interval: the ready signal was at 50 dB HL and the trigger signal was at 80 dB HL. The subject pressed a key in response to the trigger signal. Analysis was performed using computer programs. The sampling frequency was 200 Hz. The analysis time was 3.1 sec, the first 400 ms being a record of the baseline electroencephalogram. Mean initial activity was determined from an artefact-free part of the electroencephalogram trace. Averaging of 30 trials yielded: 1) the duration of CNV measured as the time interval between the start of the negative deviation from the baseline after the ready stimulus and the moment of presentation of the trigger stimulus (ms); 2) the areas of Bereitschaftpotential and negative slope, between the baseline and the negativity curve of the corresponding region *S* = (Σ*Ai*) × Δ*t* (mV·ms), where *Ai* is the amplitude of the negative deviation from the initial level at a sampling frequency of 200 Hz and Δ*t* is a time interval of duration 5 ms; 3) the mean amplitudes of Bereitschaftpotential and negative slope defined by A = ΣAi/n (V). The program also allowed calculation of the simple sensorimotor reaction time (the mean latent period of pressing the key in response to the trigger signal).

Motor symptomatology was assessed quantitatively in patients with PD in points using the unified scale UPDRS. The total score was measured in each of three dimension scales: I (impairments in thought, mood), II (decreased daily activity, impairment of hygiene activities); and III (disturbances of motor function, including bradykinesia, rigidity, and tremor) using four-point subscales for each symptom. General cognitive status of the PD patients was characterized using the standard quantitative scale Mini Mental State Examination (MMSE). The overall assessment of mental functions in normal patients yielded 30 points. Decreases in the total score to less than 25 were regarded as a sign of early dementia.

The state of coordinatory muscle interactions was studied in 29 patients with PD (aged 47– 72 years) by assessing the level of reciprocal involvement of the triceps brachii antagonist muscle (the extensor muscle of the shoulder) on functional loading of the biceps brachii muscle (the flexor of the shoulder) on the right side. EMGs were recorded using bipolar skin electrodes (0.5 × 1.0 cm2) with a constant interelectrode distance of 1.5 cm. Bioelectrical signals were passed to the amplifiers of a Medikor MG440 (Budapest) electromyography with a bandpass of 10 Hz to 10 kHz. Functional loading on the biceps brachii muscle was applied by holding a load of 2 kg on the elevated and forward extended arm for 5 sec. With the patient in the calm, relaxed state and during holding of the load, at least 100 measurements were processed using the computer program to determine the mean EMG amplitude in the triceps brachii at rest (*A*r) and loading (*Al*), the coefficient of reciprocal involvement of the antagonist muscle was calculated as *Al*/(*A*r + *Al*). This coefficient had a value of 0.5 when the amplitude on loading showed no change. If the amplitude decreased, then coefficient of reciprocal involvement had values of less than 0.5, while increases yielded coefficient of reciprocal involvement values of greater than 0.5.

The effects of cerebrolysin on measures of CNV were studied in 21 patients with PD that were taking antiparkinsonian therapy, which was not changed during one month before cerebrolysin treatment and under the whole cerebrolysin course (intravenously 10 ml, during 10 days). Before and after cerebrolysin treatment we studied clinical scores of UPDRS and CNV.

Diagnosis of Parkinson's Disease by Electrophysiological Methods 47

Fig. 14. Example traces of contingent negative variation (CNV) in two patients with

Group Area of early

phase, mV**·**ms

healthy subjects (non-parametric Mann–Whitney test).

scores for rigidity, tremor, or bradykinesia.

**characteristics of the patients with Parkinson's disease** 

Parkinson's disease aged 65 and 54, years. CNV in the first patient was poorly expressed, and the second patient showed no negative deviation. For further details see caption to Fig. 13.

> Mean amplitude of early phase, V

Healthy, n = 28 2,8 0,4 9,0 1,1 3,2 0,2 10,6 1,0 Patients, n = 53 1,8 0,2 \* 5,7 0,4 \* 1,7 0,1 \* 6,5 0,5 \* p < 0,01 < 0,01 < 0,001 < 0,01 Table 6. Differences in measures of the two phases of contingent negative variation in the median central lead (Cz) in healthy subjects and patients with Parkinson's disease. n is the number of subjects in the group.\* is significant difference between patients with PD and

**3.2.2 Correlation between the amplitudes of the early and late phases of CNV and** 

Correlation analysis revealed moderately significant relationships between the amplitudes of the two phases of CNV and point scores for individual subscales on the UPDRS. Table 7 shows that the mean amplitudes of Bereitschaftpotential and negative slope were negatively related (rS = –0.31 and rS = –0.3 respectively, *p* < 0.05) to the total point score on UPDRS subscale II, which reflects decreases in the activities of daily living (impairments of hygiene habits, cutting and holding food, difficulty dressing and walking). It was interesting to note that there was a selective negative correlation (rS = –0.32, *p* < 0.05) between the magnitude of negative slope and symptoms on subscale II such as gait freezing, while Bereitschaftpotential showed no significant relationships of this type. There were no significant correlational relationships between measures of the two phases of CNV and the total score on UPDRS subscale III, reflecting intrinsic motor functions, or with clinical point

Area of late phase, mV**·**ms Mean amplitude of late phase, V

Data obtained in healthy subjects and patients with PD were compared using the nonparametric Mann–Whitney test. Data obtained in individual patients before and after administration of cerebrolysin were analyzed using the t- test for pairwise dependent variables. Correlations between the amplitudes of the two phases of CNV, the UPDRS and MMSE scales, and the levels of reciprocal involvement of antagonist muscles were identified by calculating the correlation coefficient by the non-parametric Spearman method (rS). Relationships were regarded as moderate at 0.3 ≤ rS ≤ 0.5 and considerable at rS > 0.5. Differences were taken as significant at *p* < 0.05.

#### **3.2 Results**

#### **3.2.1 Characteristics of the early and late phases of CNV in healthy subjects and patients with Parkinson's disease**

Repeat studies in individual subjects showed that CNV had the most stable characteristics in the central medial lead (Cz), so data obtained from this lead were analyzed in detail. CNV in healthy subjects could usually be discriminated into two phases: an early phase (Bereitschaftpotential) 505–728 (596.3 ± 12.1) ms before the trigger signal and a late phase (negative slope) apparent as an additional negative deviation 170–365 (230.2 ± 15.4) ms before the trigger signal (Fig. 13). However, the second phase was not always clearly evident; in this situation CNV consisted of an initial drop-off followed by a uniform negative deviation from baseline lasting to the trigger signal. In these cases, the second half of CNV was analyzed as the second phase. In healthy subjects, the mean amplitudes of Bereitschaftpotential and negative slope were 9.0 ± 1.1 and 10.6 ± 1.0 V respectively, with areas of 2.8 ± 0.4 and 3.2 ± 0.2 mV**·**ms (Table 6). Unlike healthy subjects, CNV in many patients was poorly evident, in some, no negativity at all developed between the ready and trigger signals (Fig. 14). Statistical analysis of the data revealed significant decreases in the mean amplitudes and areas of both phases of CNV in patients with PD as compared with healthy subjects (Table 6). In addition, patients showed an increase in the simple sensorimotor reaction time for pressing the key in response to the trigger signal, from 240.9 ± 13.7 ms in healthy subjects to 299.6 ± 17.3 ms (*p* < 0.05).

Fig. 13. Characteristic record of contingent negative variation (CNV) in healthy subject aged 64 years. A is the amplitude, V. BP is the early phase and NS′ is the late phase of CNV. Vertical lines show the moments of presentation of the warning and trigger signals, with an interval of 1 sec. Positivity is shown by upward deviations from the baseline and negativity by downward deviations. Low trace is simple sensorimotor reaction times for pressing the key after presentation of the trigger signal. Units show the number of keypresses.

Data obtained in healthy subjects and patients with PD were compared using the nonparametric Mann–Whitney test. Data obtained in individual patients before and after administration of cerebrolysin were analyzed using the t- test for pairwise dependent variables. Correlations between the amplitudes of the two phases of CNV, the UPDRS and MMSE scales, and the levels of reciprocal involvement of antagonist muscles were identified by calculating the correlation coefficient by the non-parametric Spearman method (rS). Relationships were regarded as moderate at 0.3 ≤ rS ≤ 0.5 and considerable at rS > 0.5.

**3.2.1 Characteristics of the early and late phases of CNV in healthy subjects and** 

Repeat studies in individual subjects showed that CNV had the most stable characteristics in the central medial lead (Cz), so data obtained from this lead were analyzed in detail. CNV in healthy subjects could usually be discriminated into two phases: an early phase (Bereitschaftpotential) 505–728 (596.3 ± 12.1) ms before the trigger signal and a late phase (negative slope) apparent as an additional negative deviation 170–365 (230.2 ± 15.4) ms before the trigger signal (Fig. 13). However, the second phase was not always clearly evident; in this situation CNV consisted of an initial drop-off followed by a uniform negative deviation from baseline lasting to the trigger signal. In these cases, the second half of CNV was analyzed as the second phase. In healthy subjects, the mean amplitudes of Bereitschaftpotential and negative slope were 9.0 ± 1.1 and 10.6 ± 1.0 V respectively, with areas of 2.8 ± 0.4 and 3.2 ± 0.2 mV**·**ms (Table 6). Unlike healthy subjects, CNV in many patients was poorly evident, in some, no negativity at all developed between the ready and trigger signals (Fig. 14). Statistical analysis of the data revealed significant decreases in the mean amplitudes and areas of both phases of CNV in patients with PD as compared with healthy subjects (Table 6). In addition, patients showed an increase in the simple sensorimotor reaction time for pressing the key in response to the trigger signal, from 240.9

Fig. 13. Characteristic record of contingent negative variation (CNV) in healthy subject aged 64 years. A is the amplitude, V. BP is the early phase and NS′ is the late phase of CNV. Vertical lines show the moments of presentation of the warning and trigger signals, with an interval of 1 sec. Positivity is shown by upward deviations from the baseline and negativity by downward deviations. Low trace is simple sensorimotor reaction times for pressing the

key after presentation of the trigger signal. Units show the number of keypresses.

Differences were taken as significant at *p* < 0.05.

± 13.7 ms in healthy subjects to 299.6 ± 17.3 ms (*p* < 0.05).

**patients with Parkinson's disease** 

**3.2 Results** 

Fig. 14. Example traces of contingent negative variation (CNV) in two patients with Parkinson's disease aged 65 and 54, years. CNV in the first patient was poorly expressed, and the second patient showed no negative deviation. For further details see caption to Fig. 13.


Table 6. Differences in measures of the two phases of contingent negative variation in the median central lead (Cz) in healthy subjects and patients with Parkinson's disease. n is the number of subjects in the group.\* is significant difference between patients with PD and healthy subjects (non-parametric Mann–Whitney test).

#### **3.2.2 Correlation between the amplitudes of the early and late phases of CNV and characteristics of the patients with Parkinson's disease**

Correlation analysis revealed moderately significant relationships between the amplitudes of the two phases of CNV and point scores for individual subscales on the UPDRS. Table 7 shows that the mean amplitudes of Bereitschaftpotential and negative slope were negatively related (rS = –0.31 and rS = –0.3 respectively, *p* < 0.05) to the total point score on UPDRS subscale II, which reflects decreases in the activities of daily living (impairments of hygiene habits, cutting and holding food, difficulty dressing and walking). It was interesting to note that there was a selective negative correlation (rS = –0.32, *p* < 0.05) between the magnitude of negative slope and symptoms on subscale II such as gait freezing, while Bereitschaftpotential showed no significant relationships of this type. There were no significant correlational relationships between measures of the two phases of CNV and the total score on UPDRS subscale III, reflecting intrinsic motor functions, or with clinical point scores for rigidity, tremor, or bradykinesia.

Diagnosis of Parkinson's Disease by Electrophysiological Methods 49

Before CER 423.1 43.3 3.1 0.9 5.6 0.7 14.2 1.3 40.4 3.4

After CER 600.6 38.5\* 6.8 1.4\*\*\* 3.5 0.7\*\*\* 11.3 1.4\*\*\* 32.9 3.2\*

Table 8. Change of contingent negative variation (CNV) and UPDRS scores in Parkinson's disease patients after cerebrolysin (CER) treatment. Footnotes: \* - the significant change after

The results showed that patients with PD, as compared with healthy subjects, had significant decreases in the amplitudes and areas of both the early and late phases of CNV. We established that one significant factor decreasing both phases of CNV in patients with parkinsonism is impairment of coordinatory muscle interactions. Thus, the more significant the coordinatory impairment, apparent as an increase in the reciprocal involvement of the antagonist muscle during functional tests, the smaller the values of Bereitschaftpotential and negative slope in patients (rS = –0.58 and rS = –0.51 respectively; *p* < 0.01). As shown by the present data, a further significant factor affecting both phases of CNV was the state of mental functions. The positive correlation between Bereitschaftpotential amplitude and point 4 on the MMSE scale, characterizing memory (rS = 0.56, *p* < 0.01), was the most marked. This suggests that CNV can be regarded not only as a correlate of the initiation and preparation of motor structures for performing an action, but also as a neurophysiological component of mental functions. This point of view is in good agreement with published data showing sharp reductions in CNV in Alzheimer's-type dementia (Zappoli et al.,1991). The suggestion (Ikeda et al., 1997), that the nigrostriatal dopaminergic system has a greater role in generating the late phase than the early phase of CNV is supported by our finding of the existence of a selective negative correlation (*p* < 0.05) between the magnitude of negative slope and the severity of symptoms such as gait freezing; there was no such correlation for Bereitschaftpotential. The symptom of "gait freezing" does not correlate with rigidity or bradykinesia (Bartels et al., 2003), is significantly decreased by levodopa (Schaarsma et al., 2003) and depends on the functional state of the globus pallidus: stimulation of its internal zone (the main source of the efferent output of the whole of the striopallidal complex)

effectively eliminates the phenomenon of "gait freezing" (Katayama et al., 2000).

can also reduce the glutamate induced excitotoxicity (Hutter-Pair, al., 1998).

The results of the present study enlarge the perspectives in application of cerebrolysin and are in agreement with literature data on the efficacy of cerebrolysin in neurological practice. Thus, it was shown that cerebrolysin might be useful in patients with senile dementia of the Alzheimer type (Ruther, al., 2002). The positive therapeutic effect of the brain-derived peptide drug cerebrolysin can be connected with its ability to increase the expression of BBB-GLUT1 and MAP2 genes, that improves the transport of the glucose through bloodbrain barrier and keeps the cytoskeleton wholeness accordingly (Boado, 2001). Cerebrolysin

UPDRS part I scores

UPDRS part II scores

UPDRS part III scores

Mean amplitude of CNV (V)

Time of investigation

**3.3 Discussion** 

Duration of CNV (ms)

cerebrolysin treatment, p < 0.05; \*\*\* - p < 0.001 (paired t-test).


Table 7. Relationships between the amplitudes of the two phases of contingent negative variation in the median central lead (Cz) and clinical point scores in patients with Parkinson's disease. n is number of investigated persons. rS is the Spearman correlation coefficient.. \* is p<0.05; \*\* is p<0.01.

The existence of a link between measures of the two phases of CNV and the state of coordinatory muscle interactions was addressed by studying the relationship between the amplitude characteristics of CNV and the extent of reciprocal impairments between antagonist muscles in patients with PD by calculating the coefficient of reciprocal involvement. Bereitschaftpotential and negative slope were completely absent in those patients in whom the coefficient of reciprocal involvement was high (0.67–0.8), which is evidence for an abnormal increase in the reciprocal involvement of the extensor muscles in the operation of the flexor muscles. Conversely, low coefficient of reciprocal involvement was associated with maximal amplitudes for both phases of CNV. The negative correlations between the extent of reciprocal muscle involvement and the amplitudes of Bereitschaftpotential and negative slope were significant (rS = –0.58 and rS = –0.51 respectively, *p* < 0.01). Comparison of the parameters of CNV and quantitative measures on the MMSE scale revealed an identical moderate positive relationship (rS = 0.47, *p* < 0.05) with the magnitudes of both phases and the state of mental functions in patients with PD (Table 7). The strongest relationship was between Bereitschaftpotential and point 4 of the MMSE scale, which reflects memory (rS = 0.56, *p* < 0.01).

#### **3.2.3 Effects of cerebrolysin on measures of CNV in patients with Parkinson's disease**

The results of the present study showed that the course of cerebrolysin treatment combined with levodopa has the positive therapeutic effect, such as: a significant decrease of the part I, II and III UPDRS scores. A decrease of the UPDRS part I scores (improvement in thought, mood) and part II scores (that is an increase of daily activity and ability of more full value selfattendance) was the most expressed. Significant increase of the CNV amplitude value and duration well reflected the enhancement of the brain activity (Table 8).



Table 8. Change of contingent negative variation (CNV) and UPDRS scores in Parkinson's disease patients after cerebrolysin (CER) treatment. Footnotes: \* - the significant change after cerebrolysin treatment, p < 0.05; \*\*\* - p < 0.001 (paired t-test).

#### **3.3 Discussion**

48 Diagnostics and Rehabilitation of Parkinson's Disease

Mean amplitude of late

phase

phase

n=56 rS = -0,31 \* rS = -0,30 \*

(gait freezing); n=56 rS = -0,24 rS = -0,32 \*

between antagonist muscles; n=29 rS = -0,58 \*\* rS = -0,51 \*\*

functions); n=28 rS = 0,47 \* rS = 0,47 \*

n=28 rS = 0,56 \*\* rS = 0,46 \*

Table 7. Relationships between the amplitudes of the two phases of contingent negative variation in the median central lead (Cz) and clinical point scores in patients with Parkinson's disease. n is number of investigated persons. rS is the Spearman correlation

The existence of a link between measures of the two phases of CNV and the state of coordinatory muscle interactions was addressed by studying the relationship between the amplitude characteristics of CNV and the extent of reciprocal impairments between antagonist muscles in patients with PD by calculating the coefficient of reciprocal involvement. Bereitschaftpotential and negative slope were completely absent in those patients in whom the coefficient of reciprocal involvement was high (0.67–0.8), which is evidence for an abnormal increase in the reciprocal involvement of the extensor muscles in the operation of the flexor muscles. Conversely, low coefficient of reciprocal involvement was associated with maximal amplitudes for both phases of CNV. The negative correlations between the extent of reciprocal muscle involvement and the amplitudes of Bereitschaftpotential and negative slope were significant (rS = –0.58 and rS = –0.51 respectively, *p* < 0.01). Comparison of the parameters of CNV and quantitative measures on the MMSE scale revealed an identical moderate positive relationship (rS = 0.47, *p* < 0.05) with the magnitudes of both phases and the state of mental functions in patients with PD (Table 7). The strongest relationship was between Bereitschaftpotential and point 4 of the

**3.2.3 Effects of cerebrolysin on measures of CNV in patients with Parkinson's disease**  The results of the present study showed that the course of cerebrolysin treatment combined with levodopa has the positive therapeutic effect, such as: a significant decrease of the part I, II and III UPDRS scores. A decrease of the UPDRS part I scores (improvement in thought, mood) and part II scores (that is an increase of daily activity and ability of more full value selfattendance) was the most expressed. Significant increase of the CNV amplitude value

and duration well reflected the enhancement of the brain activity (Table 8).

Clinical scale points Mean amplitude of early

Total points on UPDRS subscale II;

Points on UPDRS subscale II; item 14

Coefficient of reciprocal involvement

Total score on MMSE (mental

Points on MMSE item 4 (memory);

coefficient.. \* is p<0.05; \*\* is p<0.01.

MMSE scale, which reflects memory (rS = 0.56, *p* < 0.01).

The results showed that patients with PD, as compared with healthy subjects, had significant decreases in the amplitudes and areas of both the early and late phases of CNV. We established that one significant factor decreasing both phases of CNV in patients with parkinsonism is impairment of coordinatory muscle interactions. Thus, the more significant the coordinatory impairment, apparent as an increase in the reciprocal involvement of the antagonist muscle during functional tests, the smaller the values of Bereitschaftpotential and negative slope in patients (rS = –0.58 and rS = –0.51 respectively; *p* < 0.01). As shown by the present data, a further significant factor affecting both phases of CNV was the state of mental functions. The positive correlation between Bereitschaftpotential amplitude and point 4 on the MMSE scale, characterizing memory (rS = 0.56, *p* < 0.01), was the most marked. This suggests that CNV can be regarded not only as a correlate of the initiation and preparation of motor structures for performing an action, but also as a neurophysiological component of mental functions. This point of view is in good agreement with published data showing sharp reductions in CNV in Alzheimer's-type dementia (Zappoli et al.,1991).

The suggestion (Ikeda et al., 1997), that the nigrostriatal dopaminergic system has a greater role in generating the late phase than the early phase of CNV is supported by our finding of the existence of a selective negative correlation (*p* < 0.05) between the magnitude of negative slope and the severity of symptoms such as gait freezing; there was no such correlation for Bereitschaftpotential. The symptom of "gait freezing" does not correlate with rigidity or bradykinesia (Bartels et al., 2003), is significantly decreased by levodopa (Schaarsma et al., 2003) and depends on the functional state of the globus pallidus: stimulation of its internal zone (the main source of the efferent output of the whole of the striopallidal complex) effectively eliminates the phenomenon of "gait freezing" (Katayama et al., 2000).

The results of the present study enlarge the perspectives in application of cerebrolysin and are in agreement with literature data on the efficacy of cerebrolysin in neurological practice. Thus, it was shown that cerebrolysin might be useful in patients with senile dementia of the Alzheimer type (Ruther, al., 2002). The positive therapeutic effect of the brain-derived peptide drug cerebrolysin can be connected with its ability to increase the expression of BBB-GLUT1 and MAP2 genes, that improves the transport of the glucose through bloodbrain barrier and keeps the cytoskeleton wholeness accordingly (Boado, 2001). Cerebrolysin can also reduce the glutamate induced excitotoxicity (Hutter-Pair, al., 1998).

Diagnosis of Parkinson's Disease by Electrophysiological Methods 51

auditory evoked potentials were recorded at the vertex (Cz) referenced to a linked-ear electrode. The ground electrode was placed at the left wrist. The impedance of the electrodes was less than 10 k. The electrode signal was amplified using a bandpass filter (0.53 - 30

The pattern for double stimulation consisted of paired auditory clicks with 500, 700, 800, 900, 1100 and 2000 ms interstimulus intervals. The identical parameters (duration of 0.15 ms and intensity of 80 dB HL - hearing level) were used for the preceding conditioning click and following test click. Pairs of clicks were delivered once every 7 s for each interstimulus interval. Previous studies have shown that stimulation at faster frequencies can lead to a decrement in the cortical evoked potentials. A 2000 - 3000 ms electroencephalography epoch was recorded for each trial, including a 300 ms prestimulus baseline. The recording time depended on interstimulus intervals. The epochs contaminated with blinks or other artefacts were excluded from the data and twenty acceptable artefact-free trials were averaged for each interstimulus interval and used for further analysis. In electroencephalography recordings upon paired stimulation, amplitudes of N1-P2 complex (peak to peak) in the first (A1) and the second (A2) responses were measured. The amplitudes of the components N1 and P2 were estimated in the 60 – 150 ms and 120 - 220 ms ranges of time, respectively. The percent of pairedpulse inhibition of the N1-P2 complex was calculated using the formula: (A1-A2)/A1×100. The effects of cerebrolysin on the postexcitatory inhibition of the N1/P2 complex of the cortical evoked potentials on auditory paired-click stimulation were studied in 21 patients with PD that were taking antiparkinsonian therapy, which was not changed during one month before cerebrolysin treatment and under the whole cerebrolysin course

The results were analyzed statistically. Comparisons between PD patients and control groups were made using a non-parametric two-tailed Mann-Whitney criterion. Data obtained from the same patients before and after cerebrolysin treatment were compared

The postexcitatory cortical inhibition in response to auditory stimulation studied with a paired-pulse paradigm was significantly reduced in patients with PD compared to control subjects. Amplitudes of N1-P2 complexes following the second stimulus of a pair at interstimulus intervals of 500, 700 and 900 ms were greater in PD patients. The mean values of paired-pulse inhibition in the group of PD patients were decreased to 29.8 4.8 % (p<0.01), 25.4 3.2 % (p<0.001) and 15.1 2.6 % (p<0.001) for intervals 500, 700 and 900 ms respectively as compared to these values (54.1 4.2 %; 49.8 2.3 % and 42.9 2.7 %) in the

The mean amplitude of N1-P2 complex elicited by a single (first) auditory stimulus in the group of PD patients was 16.2 0.8 V which was less than in age-matched subjects (18.5

A distinct positive effect of the course of cerebrolysin treatment on the postexcitatory cortical inhibition at paired-click stimulation was observed in the group of 21 PD patients. A noticeable shift of the paired-pulse inhibition value for 700, 800 and 900 ms intervals

**4.2.2 The influence of cerebrolysin treatment on the postexcitatory inhibition** 

1.6 V) but this difference was not statistically significant (p>0.05).

towards the values of the healthy control was found (Table 10, Fig 15).

**4.2.1 Investigation of the postexcitatory inhibition following paired stimulation** 

Hz*),* digitised with 200 Hz sampling rate and stored for further analysis.

(intravenously 10 ml, during 10 days).

group of age-matched controls (Table 9).

using two-tailed paired t-test.

**4.2 Results** 

Obtained data proof that CNV appears to be a good tool for the evaluation of the medication efficiency. The parameters of CNV well reflected the improvement of the functional state of the patients after the course of cerebrolysin treatment.

## **4. Cortical evoked potentials upon paired-click auditory stimulation**

It has been previously reported in clinical and experimental studies that movement disorders in PD largely occur due to the imbalance of inhibitory and excitatory processes in motor cortical and subcortical neuronal circuits following a nigrostriatal dopamine deficit (Ridding et al., 1995). A paired-pulse paradigm is usually used to study postexcitatory inhibition effect related to sensory gating mechanisms and synaptic processes in neurotransmitters release (Chu et al., 2009). There are two mechanisms that might explain paired-pulse inhibition phenomena. The first mechanism is the decrease in release probability of excitatory neurotransmitters from terminals of afferent axons (Szabo et al., 2000). Another possible mechanism of the decrement of the second response on paired stimulation is connected with synaptically released GABA from terminals of inhibitory interneurons (Chu & Hablitz, 2003). As the paired-pulse facilitation, paired-pulse inhibition is considered to be a form of a short-term synaptic plasticity. The investigation of cortical evoked potentials to paired-pulse sensory stimulation may provide additional information about mechanisms of neurological disturbances in PD.

The aim of this study was to investigate the postexcitatory inhibition of the N1/P2 complex of the cortical evoked potentials on auditory paired-click stimulation in patients with PD in comparison with age-matched healthy subjects. Our second goal was to evaluate the influence of neurotrophic drug cerebrolysin on postexcitatory cortical inhibition.

## **4.1 Methods**

Studies were performed in two groups. The first group included 58 PD patients, with the severity of the disease corresponding to 1.5 - 3.0 of Hoehn M.M. and Yahr M.D. (1967) scale (28 men and 30 women, mean SE age 61.5 1.1, range 45 - 74 years). The second group was control and consisted of 22 age-matched healthy subjects (10 men and 12 women, mean SE age 61.4 1.1, range 48 - 73 years).

The study was approved in advance by the Ethical Committee of the Institute of Gerontology and was in accordance with the Declaration of Helsinki. The patients regularly underwent treatment at the Parkinson's Disease Centre of the Institute of Gerontology and gave written informed consent to participate in this study. The diagnosis of Parkinson's disease was determined according to the UK Bank Criteria (Hughes A. et al., 1992). The patients had from 2 to 22 year individual histories of idiopathic PD and were taking antiparkinsonian therapy at individual dose of 187.5 - 750 mg of levodopa / carbidopa daily. Besides levodopa / carbidopa, the patients were using other antiparkinsonian medication: selegiline, pramipexol, amantadine. The neurological status of PD patients was evaluated with Unified Parkinson's Disease Rating Scale (UPDRS; Fahn S. and R. Elton., 1987; Holloway R.G. et al., 2004) in the "ON" state 1 hour after levodopa / carbidopa intake. Mini Mental State Examination (MMSE) was used to study general cognitive status of the PD patients.

Auditory evoked potentials were recorded in the PD patients in their "OFF" state in the morning, after they were free from levodopa treatment and other antiparkinsonian medications for at least 12 hours. During registration of evoked potentials the subjects were sitting comfortably in a semi-reclined armchair in a quiet room with closed eyes. Cortical auditory evoked potentials were recorded at the vertex (Cz) referenced to a linked-ear electrode. The ground electrode was placed at the left wrist. The impedance of the electrodes was less than 10 k. The electrode signal was amplified using a bandpass filter (0.53 - 30 Hz*),* digitised with 200 Hz sampling rate and stored for further analysis.

The pattern for double stimulation consisted of paired auditory clicks with 500, 700, 800, 900, 1100 and 2000 ms interstimulus intervals. The identical parameters (duration of 0.15 ms and intensity of 80 dB HL - hearing level) were used for the preceding conditioning click and following test click. Pairs of clicks were delivered once every 7 s for each interstimulus interval. Previous studies have shown that stimulation at faster frequencies can lead to a decrement in the cortical evoked potentials. A 2000 - 3000 ms electroencephalography epoch was recorded for each trial, including a 300 ms prestimulus baseline. The recording time depended on interstimulus intervals. The epochs contaminated with blinks or other artefacts were excluded from the data and twenty acceptable artefact-free trials were averaged for each interstimulus interval and used for further analysis. In electroencephalography recordings upon paired stimulation, amplitudes of N1-P2 complex (peak to peak) in the first (A1) and the second (A2) responses were measured. The amplitudes of the components N1 and P2 were estimated in the 60 – 150 ms and 120 - 220 ms ranges of time, respectively. The percent of pairedpulse inhibition of the N1-P2 complex was calculated using the formula: (A1-A2)/A1×100. The effects of cerebrolysin on the postexcitatory inhibition of the N1/P2 complex of the cortical evoked potentials on auditory paired-click stimulation were studied in 21 patients with PD that were taking antiparkinsonian therapy, which was not changed during one month before cerebrolysin treatment and under the whole cerebrolysin course (intravenously 10 ml, during 10 days).

The results were analyzed statistically. Comparisons between PD patients and control groups were made using a non-parametric two-tailed Mann-Whitney criterion. Data obtained from the same patients before and after cerebrolysin treatment were compared using two-tailed paired t-test.

#### **4.2 Results**

50 Diagnostics and Rehabilitation of Parkinson's Disease

Obtained data proof that CNV appears to be a good tool for the evaluation of the medication efficiency. The parameters of CNV well reflected the improvement of the functional state of

It has been previously reported in clinical and experimental studies that movement disorders in PD largely occur due to the imbalance of inhibitory and excitatory processes in motor cortical and subcortical neuronal circuits following a nigrostriatal dopamine deficit (Ridding et al., 1995). A paired-pulse paradigm is usually used to study postexcitatory inhibition effect related to sensory gating mechanisms and synaptic processes in neurotransmitters release (Chu et al., 2009). There are two mechanisms that might explain paired-pulse inhibition phenomena. The first mechanism is the decrease in release probability of excitatory neurotransmitters from terminals of afferent axons (Szabo et al., 2000). Another possible mechanism of the decrement of the second response on paired stimulation is connected with synaptically released GABA from terminals of inhibitory interneurons (Chu & Hablitz, 2003). As the paired-pulse facilitation, paired-pulse inhibition is considered to be a form of a short-term synaptic plasticity. The investigation of cortical evoked potentials to paired-pulse sensory stimulation may provide additional information

The aim of this study was to investigate the postexcitatory inhibition of the N1/P2 complex of the cortical evoked potentials on auditory paired-click stimulation in patients with PD in comparison with age-matched healthy subjects. Our second goal was to evaluate the

Studies were performed in two groups. The first group included 58 PD patients, with the severity of the disease corresponding to 1.5 - 3.0 of Hoehn M.M. and Yahr M.D. (1967) scale (28 men and 30 women, mean SE age 61.5 1.1, range 45 - 74 years). The second group was control and consisted of 22 age-matched healthy subjects (10 men and 12 women, mean

The study was approved in advance by the Ethical Committee of the Institute of Gerontology and was in accordance with the Declaration of Helsinki. The patients regularly underwent treatment at the Parkinson's Disease Centre of the Institute of Gerontology and gave written informed consent to participate in this study. The diagnosis of Parkinson's disease was determined according to the UK Bank Criteria (Hughes A. et al., 1992). The patients had from 2 to 22 year individual histories of idiopathic PD and were taking antiparkinsonian therapy at individual dose of 187.5 - 750 mg of levodopa / carbidopa daily. Besides levodopa / carbidopa, the patients were using other antiparkinsonian medication: selegiline, pramipexol, amantadine. The neurological status of PD patients was evaluated with Unified Parkinson's Disease Rating Scale (UPDRS; Fahn S. and R. Elton., 1987; Holloway R.G. et al., 2004) in the "ON" state 1 hour after levodopa / carbidopa intake. Mini Mental State Examination (MMSE)

Auditory evoked potentials were recorded in the PD patients in their "OFF" state in the morning, after they were free from levodopa treatment and other antiparkinsonian medications for at least 12 hours. During registration of evoked potentials the subjects were sitting comfortably in a semi-reclined armchair in a quiet room with closed eyes. Cortical

influence of neurotrophic drug cerebrolysin on postexcitatory cortical inhibition.

**4. Cortical evoked potentials upon paired-click auditory stimulation** 

the patients after the course of cerebrolysin treatment.

about mechanisms of neurological disturbances in PD.

was used to study general cognitive status of the PD patients.

SE age 61.4 1.1, range 48 - 73 years).

**4.1 Methods** 

#### **4.2.1 Investigation of the postexcitatory inhibition following paired stimulation**

The postexcitatory cortical inhibition in response to auditory stimulation studied with a paired-pulse paradigm was significantly reduced in patients with PD compared to control subjects. Amplitudes of N1-P2 complexes following the second stimulus of a pair at interstimulus intervals of 500, 700 and 900 ms were greater in PD patients. The mean values of paired-pulse inhibition in the group of PD patients were decreased to 29.8 4.8 % (p<0.01), 25.4 3.2 % (p<0.001) and 15.1 2.6 % (p<0.001) for intervals 500, 700 and 900 ms respectively as compared to these values (54.1 4.2 %; 49.8 2.3 % and 42.9 2.7 %) in the group of age-matched controls (Table 9).

The mean amplitude of N1-P2 complex elicited by a single (first) auditory stimulus in the group of PD patients was 16.2 0.8 V which was less than in age-matched subjects (18.5 1.6 V) but this difference was not statistically significant (p>0.05).

#### **4.2.2 The influence of cerebrolysin treatment on the postexcitatory inhibition**

A distinct positive effect of the course of cerebrolysin treatment on the postexcitatory cortical inhibition at paired-click stimulation was observed in the group of 21 PD patients. A noticeable shift of the paired-pulse inhibition value for 700, 800 and 900 ms intervals towards the values of the healthy control was found (Table 10, Fig 15).

Diagnosis of Parkinson's Disease by Electrophysiological Methods 53

The main result of this study showed that PD patients had significantly reduced pairedpulse inhibition of the N1/P2 component of evoked potentials in the auditory cortex for interstimulus intervals of 500, 700 and 900 ms compared to the healthy age-matched subjects. Possible explanation of the reduced cortical inhibition in PD is the functional deficiency of inhibitory interneurons caused by depletion of dopaminergic innervation in the cerebral cortex (Gaspar et al., 1991). As already established (Krnjevic et al., 1966), afferent volleys after initial excitatory postsynaptic potentials (EPSPs) result in inhibitory postsynaptic potentials (IPSPs). A system of GABAergic interneurons, which can be activated by direct and indirect stimulation, may play the major role in the genesis of these IPSPs (Hanajima & Ugawa, 2000). The synaptic release of GABA is regulated by presynaptic GABA receptors of the B-type (Chu & Hablitz, 2003). There is also strong evidence that dopamine regulates inhibitory transmission at the synapses between pyramidal cells and interneurons by activating D1-like receptors located on the presynaptic terminals of GABAergic axons (Gonzalez-Islas & Hablitz, 2001). Dysfunction of cortical interneurons in PD also might be a result of noradrenergic denervation and monoamine terminal loss (Marie et al., 1995), as some investigations showed that cortical GABAergic interneurons can be

Another possible explanation of the reduced inhibition in the auditory cortex in patients with PD may be the loss of dopaminergic transmission in the basal ganglia and the dysfunction of the caudal pallidum that sends its direct projections to the inferior colliculus, medial geniculate nucleus and temporal cerebral cortex (Shammah-Lagnado et al., 1996). The basal ganglia appear to "gate" sensory inputs at various levels and activation of basal ganglia outputs (entopeduncular nucleus and substantia nigra pars reticulate) is able to

Our findings allow to suppose that drugs, which are able to activate cerebral inhibitory GABAergic system, can be useful in medication of PD. Phenibut (noofen) belongs to such drugs (Marshall & Foord, 2010). Application of noofen in complex therapy of PD appeared effective for the improvement of cognitive functions, enhancement of emotional state and

This study demonstrated that course of cerebrolysin treatment promotes normalization of the inhibitory brain processes. The positive effect of cerebrolysin indicates that neurotrophic drugs can also be useful in complex antiparkinsonian therapy for advance of the ability of

The present investigation has shown that the surface EMG data add essential information to the clinical characteristics of PD patients. We found that separate EMG indices correlated, in a specific manner, with certain UPDRS sub-items, which could result in a better understanding of the pathogenesis of clinical PD symptoms. Motor disorders in PD (part III UPDRS scores) were found to be predominantly associated with disturbances in regulation of the tonic and phasic muscle activities. At the same time, disorders of the upper extremity daily activity (points 8-10 of UPDRS) and the dyskinesia (disability) (point 33 of UPDRS) are largely conditioned by the disturbance of reflex coordinating relationships between the muscles in PD. EMG analysis seems to be a useful tool for levodopa therapy adjustment and

excited via alpha-adrenoreceptors (Kawaguchi & Shindou, 1998).

increase of social adaptation of the PD patients (Karaban et al., 2006).

inhibit sensory responses (Boecker et al., 1999).

the brain to provide normal inhibition.

for predicting the course of disease.

**5. Conclusion** 

**4.3 Discussion** 


Table 9. Inhibition of the second N1-P2 complex of cortical auditory evoked potentials at paired-click stimulation in age-matched control group and patients with Parkinson's disease (Mean ± SE).

\* - P<0.01; \*\* - P<0.001 compared to control subjects (nonparametric Mann-Whitney test).


Table 10. The influence of the course of cerebrolysin treatment on the postexcitatory inhibition following paired-click auditory stimulation in patients with Parkinson's disease (Mean SE).

Fig. 15. Cortical auditory evoked potentials at paired auditory stimulation with interstimulus intervals of 800 and 900 ms in healthy control and patient with Parkinson's disease (PD) before and after the course of cerebrolysin (CER). N1(I), P2(I) – the components of cortical evoked potentials on the first conditional stimulus and N1(II), P2(II) – on the second test stimulus. Vertical solid bars on the records correspond to the onset of auditory signals.

### **4.3 Discussion**

52 Diagnostics and Rehabilitation of Parkinson's Disease

Age-matched control 54.1 ± 4.2 49.8± 2.3 42.9 ± 2.7 48.0 ± 2,1

Table 9. Inhibition of the second N1-P2 complex of cortical auditory evoked potentials at paired-click stimulation in age-matched control group and patients with Parkinson's disease

Time of investigation **Averaged value of paired-pulse inhibition** (%) **at interstimulus** 

Before cerebrolysin 29.9 3.9 26.7 3.4 17.1 3.1 24.6 2.3 After cerebrolysin 38.1 3.2 37.1 3.3 27.5 4.1 34.2 2.9 P (paired t-test) <0.01 <0.001 <0.05 <0.001 Table 10. The influence of the course of cerebrolysin treatment on the postexcitatory inhibition following paired-click auditory stimulation in patients with Parkinson's disease

Fig. 15. Cortical auditory evoked potentials at paired auditory stimulation with interstimulus intervals of 800 and 900 ms in healthy control and patient with Parkinson's disease (PD) before and after the course of cerebrolysin (CER). N1(I), P2(I) – the components of cortical evoked potentials on the first conditional stimulus and N1(II), P2(II) – on the second test stimulus. Vertical solid bars on the records correspond to the onset of auditory signals.

29.8 ± 4.8

\* - P<0.01; \*\* - P<0.001 compared to control subjects (nonparametric Mann-Whitney test).

Inhibition in % of the second N1-P2

500 ms 700 ms 900 ms

\*\* 25.4 ± 3.2

complex at interstimulus intervals Averaged

**intervals (ms) 700 800** 900 **Averaged data** 

\*\* 15.1± 2.6

\*\* 21,4 ±2,4

Investigated groups

(Mean ± SE).

(Mean SE).

PD patients \*

The main result of this study showed that PD patients had significantly reduced pairedpulse inhibition of the N1/P2 component of evoked potentials in the auditory cortex for interstimulus intervals of 500, 700 and 900 ms compared to the healthy age-matched subjects. Possible explanation of the reduced cortical inhibition in PD is the functional deficiency of inhibitory interneurons caused by depletion of dopaminergic innervation in the cerebral cortex (Gaspar et al., 1991). As already established (Krnjevic et al., 1966), afferent volleys after initial excitatory postsynaptic potentials (EPSPs) result in inhibitory postsynaptic potentials (IPSPs). A system of GABAergic interneurons, which can be activated by direct and indirect stimulation, may play the major role in the genesis of these IPSPs (Hanajima & Ugawa, 2000). The synaptic release of GABA is regulated by presynaptic GABA receptors of the B-type (Chu & Hablitz, 2003). There is also strong evidence that dopamine regulates inhibitory transmission at the synapses between pyramidal cells and interneurons by activating D1-like receptors located on the presynaptic terminals of GABAergic axons (Gonzalez-Islas & Hablitz, 2001). Dysfunction of cortical interneurons in PD also might be a result of noradrenergic denervation and monoamine terminal loss (Marie et al., 1995), as some investigations showed that cortical GABAergic interneurons can be excited via alpha-adrenoreceptors (Kawaguchi & Shindou, 1998).

Another possible explanation of the reduced inhibition in the auditory cortex in patients with PD may be the loss of dopaminergic transmission in the basal ganglia and the dysfunction of the caudal pallidum that sends its direct projections to the inferior colliculus, medial geniculate nucleus and temporal cerebral cortex (Shammah-Lagnado et al., 1996). The basal ganglia appear to "gate" sensory inputs at various levels and activation of basal ganglia outputs (entopeduncular nucleus and substantia nigra pars reticulate) is able to inhibit sensory responses (Boecker et al., 1999).

Our findings allow to suppose that drugs, which are able to activate cerebral inhibitory GABAergic system, can be useful in medication of PD. Phenibut (noofen) belongs to such drugs (Marshall & Foord, 2010). Application of noofen in complex therapy of PD appeared effective for the improvement of cognitive functions, enhancement of emotional state and increase of social adaptation of the PD patients (Karaban et al., 2006).

This study demonstrated that course of cerebrolysin treatment promotes normalization of the inhibitory brain processes. The positive effect of cerebrolysin indicates that neurotrophic drugs can also be useful in complex antiparkinsonian therapy for advance of the ability of the brain to provide normal inhibition.

## **5. Conclusion**

The present investigation has shown that the surface EMG data add essential information to the clinical characteristics of PD patients. We found that separate EMG indices correlated, in a specific manner, with certain UPDRS sub-items, which could result in a better understanding of the pathogenesis of clinical PD symptoms. Motor disorders in PD (part III UPDRS scores) were found to be predominantly associated with disturbances in regulation of the tonic and phasic muscle activities. At the same time, disorders of the upper extremity daily activity (points 8-10 of UPDRS) and the dyskinesia (disability) (point 33 of UPDRS) are largely conditioned by the disturbance of reflex coordinating relationships between the muscles in PD. EMG analysis seems to be a useful tool for levodopa therapy adjustment and for predicting the course of disease.

Diagnosis of Parkinson's Disease by Electrophysiological Methods 55

Berg, D., Roggendorf, W., Schroder, U. et al.. 2002. Echogenicity of the substantia nigra:

Boado, R.J. 2001. Amplification of blood-brain barrier GLUTI glucose transporter gene expression by brain-derived peptides. Neurosci. Res, Vol. 40, No. 4, pp. 337-342. Boecker, H., Ceballos-Baumann, A., Bartenstein, P., Weindl, A., Siebner, H.R., Fassbender,

Chu, Z., Hablitz, J.J. 2003. GABA (B) receptor-mediated heterosynaptic depression of

Chu, J., Wagle-Shukla, A., Gunraj, C., Lang, A. E. & Chen R. 2009. Impaired presynaptic

Deecke, L. 2001. Clinical neurophysiology of Parkinson's disease. Bereitschaftpotential and

DeLong, M.R. 1990. Primate models of movement disorders of basal ganglia origin. Trends

Del Santo, F, Gelli, F.,Mazzocchio, R. & Rossi, A. 2007. Recurrence quantification analysis of

Elbaz, A., Grigoletto, F., Baldereschi, M. et al. 1999. Familial aggregation of Parkinson's

Fahn, S., Elton, R. & members of the UPDRS Development Committee. 1987.Unified

Farina, D., Merletti, R. & Enoka, R.M. 2004. The extraction of neural strategies from the

Filipovic, S. R., Covickovic-Sternic, N., Radovic, V. M., Dragasevic, N., Stoyanovic-Svete, M.

between akineto-rigid and tremor types. J Neurol Sci,, Vol. 103, pp. 232-237 Gaspar, P., Duyckaerts, C., Alvarez, C., Javoy-Agid, F., & Berger B. 1991 Alterations of

Goldberger, A. L., Amaral, L.A.N., Hausdorff, J.V., Ivanov, P.Ch., Peng, C.-K. & Stanley,

contingent negative variation. Adv. Neurol., Vol. 86, pp. 257–271.

nigrostriatal injury. Arch. Neurol., Vol. 59, No. 6, pp. 999-1005.


39-49.

842–849.

(Pt 9), pp. 1651-1665.

Neurosci., Vol. 13, pp. 281-286.

inhibition. Exp. Brain Res, Vol. 178, pp. 308-315.

Study Group. Neurology, Vol. 52, No. 9, pp. 1876-1882.

surface EMG. J. Appl. Physiol, Vol. 96, pp. 1486-1495.

disease. Ann. Neurol., Vol. 30, pp. 365-374.

Natl. Acad. Sci. USA, Vol. 99, Suppl. 1, pp. 2466-2472.

293-304, NJ Macmillan Health Care Information, Florham Park.

FOG is not correlated with bradykinesia. J. Clin. Neurosci., Vol. 10, No. 5, pp. 584

association with increased iron content and marker for susceptibility to

T., Munz, F., Schwaige,r M. & Conrad, B. 1999. Sensory processing in Parkinson's and Huntington's disease: investigations with 3D H(2) (15) O-PET. Brain, Vol. 122

excitatory synaptic transmission in rat frontal neocortex. Brain Res., Vol. 959, pp.

inhibition in the motor cortex in Parkinson disease. Neurology, vol. 72, No. 9, pp.

surface EMG detects changes in motor unit Synchronization induced by recurrent

disease: a population-based case-control study in Europe. EUROPARKINSON

Parkinson's disease rating scale. In: Recent developments in Parkinson's disease, Fahn, S., Marsden, C.D., Calne, D.B., & Goldstein, M., editors, Vol. 2, pp. 153-163,

& Kostic, V. S.. 1997. Correlation between Bereitschaftspotential and reaction time measurements in patients with Parkinson's disease. Measuring the impaired supplementary motor area function? J. Neurol. Sci., Vol. 147, No. 2, pp. 177 - 183. Furukawa, Y., Kondo, T., Nishi, K., Yokochi, F. & Narabayashi, H. 1991. Total biopterin

levels in the ventricular CSF of patients with Parkinson's disease: a comparison

dopaminergic and noradrenergic innervations in motor cortex in Parkinson's

H.E. 2002. Fractal dynamics in physiology: alterations with disease and aging. Proc.

In this study critical values of normal statistics of surface EMG distribution at rest were defined. Evaluation of statistical parameters of the EMG signals, to our opinion, appeared to be effective for the detection of signs of the disturbed muscle activity. Range and variance reflect the extent of bioelectrical muscle signals. Kurtosis characterizes motor unit synchronization. These EMG characteristics assist to detect latent symptoms of extrapyramidal insufficiency in clinically healthy kinsmen of the patients suffering from PD that can be considered genetic determinants of the risk of development of the above disease. Formulation of recommendations for individuals belonging to a risk group is of exceptional importance to prevent manifestations of PD.

Novel EMG characteristic is fractal dynamics of EMG data based on detrended fluctuation analysis and calculation of Hurst exponent. Fractal dimension studies the non-linear properties of EMG. The present investigation has demonstrated distinctive features of surface EMG signals fractal dimension at rest in patients with akinetic-rigid-trembling form of PD: 1) fractal dimension in PD patients is more complex compared to healthy subjects; 2) the value of Hurst exponent is significantly less in patients; 3) there is the considerable degradation of short and longer range correlation properties of EMG signals in PD. Fractal analysis has proved to be sensitive to neuromuscular status and may have potential in the assessment of the severity of PD

Evaluation of brain evoked potentials provides additional information about the mechanisms of neurological disturbances in PD. The results obtained in the present study produce evidence for significant relationships between both the early and late phases of movement-related potential CNV and the neurophysiological mechanisms supporting coordinatory muscle interactions and mental functions, including the simultaneous activity of numerous specific and non-specific brain structures (motor cortex, supplementary motor cortex, prefrontal cortex, cerebellar and thalamic projections). The existence of a selective negative correlation between the magnitude of the late CNV phase and the severity of symptoms such as "gait freezing" suggests a great role of efferent system of the basal ganglia in generating this phase of CNV. The investigation of cortical evoked potentials at paired-pulse sensory stimulation shows that inhibitory processes are deficient in PD patients. The findings may suggest that drugs, being the derivates of GABA, can be useful in treatment of PD. The parameters of CNV and the value of postexcitatory cortical inhibition at paired-click sensory stimulation well characterize the state of brain activity. Together with other neurophysiological parameters the brain evoked potentials might be a good tool for quantifying the efficacy of medication of PD patients.

## **6. References**


In this study critical values of normal statistics of surface EMG distribution at rest were defined. Evaluation of statistical parameters of the EMG signals, to our opinion, appeared to be effective for the detection of signs of the disturbed muscle activity. Range and variance reflect the extent of bioelectrical muscle signals. Kurtosis characterizes motor unit synchronization. These EMG characteristics assist to detect latent symptoms of extrapyramidal insufficiency in clinically healthy kinsmen of the patients suffering from PD that can be considered genetic determinants of the risk of development of the above disease. Formulation of recommendations for individuals belonging to a risk group is of exceptional

Novel EMG characteristic is fractal dynamics of EMG data based on detrended fluctuation analysis and calculation of Hurst exponent. Fractal dimension studies the non-linear properties of EMG. The present investigation has demonstrated distinctive features of surface EMG signals fractal dimension at rest in patients with akinetic-rigid-trembling form of PD: 1) fractal dimension in PD patients is more complex compared to healthy subjects; 2) the value of Hurst exponent is significantly less in patients; 3) there is the considerable degradation of short and longer range correlation properties of EMG signals in PD. Fractal analysis has proved to be sensitive to neuromuscular status and may have potential in the

Evaluation of brain evoked potentials provides additional information about the mechanisms of neurological disturbances in PD. The results obtained in the present study produce evidence for significant relationships between both the early and late phases of movement-related potential CNV and the neurophysiological mechanisms supporting coordinatory muscle interactions and mental functions, including the simultaneous activity of numerous specific and non-specific brain structures (motor cortex, supplementary motor cortex, prefrontal cortex, cerebellar and thalamic projections). The existence of a selective negative correlation between the magnitude of the late CNV phase and the severity of symptoms such as "gait freezing" suggests a great role of efferent system of the basal ganglia in generating this phase of CNV. The investigation of cortical evoked potentials at paired-pulse sensory stimulation shows that inhibitory processes are deficient in PD patients. The findings may suggest that drugs, being the derivates of GABA, can be useful in treatment of PD. The parameters of CNV and the value of postexcitatory cortical inhibition at paired-click sensory stimulation well characterize the state of brain activity. Together with other neurophysiological parameters the brain evoked potentials might be a good tool

Abbruzzese, G. & Berardelli, A. 2003. Sensorimotor integration in movement disorders.

Antonini, A., Leenders, K. L. & Eidelberg, D. 1998. [11C]raclopride-PET studies of the

Aotsuka, A., Wheate, S. J., Dranke, M. E., Jr. & Paulson, G. W. 1996. Event-related potentials

Bartels, A. L., Balash, Y., Gurevich, T., Schaafsma, J. D., Hausdorff, J. M. & Giladi N. 2003.

Huntington's disease rate of progression: relevance of the trinucleotide repeat

in Parkinson's disease. Electromyogr. Clin. Neurophysiol., Vol. 36, No. 4, pp. 215–

Relationship between freezing of gait (FOG) and other features of Parkinson's:

importance to prevent manifestations of PD.

for quantifying the efficacy of medication of PD patients.

Mov. Disord., Vol. 18, pp. 231-240.

length. Ann. Neurol., Vol. 43, No. 2, pp. 253-255.

assessment of the severity of PD

**6. References** 

220.

FOG is not correlated with bradykinesia. J. Clin. Neurosci., Vol. 10, No. 5, pp. 584 - 588.


Diagnosis of Parkinson's Disease by Electrophysiological Methods 57

Lukhanina, E.P., Karaban', I.N., Chivliklii, M.A., Pil'kevich, N.A. & Berezetskaya, N.M.

Marie, R.M., Barre, L., Rioux, P., Allain, P., Lechevalier, B. & Baron J.C. 1995. PET imaging of

Marshall, F. H. & Foord, S. M. 2010. Heterodimerization of the GABAB receptor-

Meigal, A.I., Rissanen, S., Tarvainen, M.P., Karjalainen, P.A., Iudina-Vassel, I.A., Airaksinen,

Nieminen, H. & Takala E.P. 1996. Evidence of deterministic chaos in the myoelectric signal.

Oishi, M., Mochizuki, Y.C. Du. & Takasu T. 1995. Contingent negative variation and

Otsuka, M., Ichiya, Y., Kuwabara, Y., Hosokawa, S., Sasaki, M., Yoshida, T., Fukumura, T.,

Pulvermuller, F., Lutzenberger, W., Muller, V., Mohr, B., Dichgans, J. & Birbaumer, N. 1996.

Ridding, M.C., Inzelberg, R. & Rothwell, J.C. 1995. Changes in excitability of motor cortical circuitry in patients with Parkinson's disease. Ann. Neurol., Vol. 37, pp. 181-188. Robichaud, J. A., Pfann, K. D., Leurgans, S. et al. 2009. Variability of EMG patterns: a

Rossini, P.M., Filippi, M.M. & Vernieri, F. 1998. Neurophysiology of sensorimotor integration in Parkinson's disease. Clin. Neurosci., Vol. 5, pp. 121-130. Ruether, E., Alvarez, X.A., Rainer, M. & Moessler, H. 2002. Sustained improvement of

Sadekov, R.A. 1997. Evoked potentials in parkinsonism. Zh. Nevropatol. Psikhiatr. Im. S. S.

Schaafsma, J. D., Balash, Y., Gurevich, T., Bartels, A.. L., Hausdorff, J. M. & Giladi N. 2003.

Semmler, J.G. & Nordstrom, M.A. 1999. A comparison of cross-correlation and surface EMG

Shammah-Lagnado, S.J., Alheid, G.F. & Heimer, L. 1996. Efferent connections of the caudal part of the globus pallidus in the rat. J. Comp. Neurol., Vol. 376, pp. 489-507.

in Parkinson's disease. Eur. J. Neurol., Vol. 10, No. 4, pp. 391 - 398.

Cerebrolysis. J. Neural. Transm., Suppl., Vol. 62, pp. 265-75.

insufficiency. Neurophysiology, Vol. 42, No 1, pp. 39-49.

Electromyogr. Clin. Neurophysiol, Vol. 36, pp. 49-58.

symptoms. J. Neurol. Sci., Vol. 136, pp. 169-173.

Neurophysiol., Vol. 98, No. 6, pp. 456–467.

Park. Dis. Dement. Sect., Vol. 9, pp. 55-71.

vol. 57, C, pp. 63– 91.

Vol. 19, pp e206-e213.

Vol. 95, No. 5, pp. 346–349.

Vol. 120, No. 2, pp. 390-397.

Korsakova, Vol. 97, pp. 64-65.

Methods, Vol. 90, pp. 47-55.

2010. Electromyographic manifestations of hereditary signs of extrapiramidal

neocortical monoaminergic terminals in Parkinson's disease. J. Neural. Transm.

implications for GPCR signaling and drug discovery. Advances in Pharmacology,

O. & Kankaanpää, M. 2009. Novel parameters of surface EMG in patients with Parkinson's disease and healthy young and old controls. J. Electromyogr. Kinesiol,

movement-related cortical potentials in parkinsonism. EEG Clin. Neurophysiol.,

Masuda, K. & Kato, M. 1996. Differences in the reduced 18F-Dopa uptakes of the caudate and the putamen in Parkinson's disease: correlations with the three main

P3 and contingent negative variation in Parkinson's disease. EEG Clin.

potential neurophysiological marker of Parkinson's disease? Clin. Neurophysiol.,

cognition and global function in patients with moderately severe Alzheimer's disease: a double-blind, placebo-controlled study with the neurotrophic agent

Characterization of freezing of gait subtypes and the response of each to levolopa

techniques used to quantify motor unit Synchronization in humans. J. Neurosci.


Gonzalez-Islas, C. & Hablitz, J.J. 2001. Dopamine inhibition of evoked IPSCs in rat prefrontal

Hanajima, R. & Ugawa, Y. 2000. Intracortical inhibition of the motor cortex in movement

Holloway, R.G., Shoulson, I., Fahn, S. et al. 2004. Parkinson Study Group. Pramipexole *vs*

Houk, J.C. 1979. Regulation of stiffness by skeletomotor reflexes. Ann. Rev. Physio.l, Vol.

Hughes, A.I., Ben-Shlomo, Y., Daniel, S.E. & Lees, A.I. 1992. What features improve the

Hutter-Paier, B., Grygar, E., Fruhwirth, M., Temmel, I. & Windisch, M. 1998. Further

Ikeda, A., Shibasaki, H., Kaji, R., Terada, K., Nagamine, T., Honda, M. & Kimura, J. 1997.

Illarioshkin, S. N. 2002. Conformational Disease of the Brain [in Russian], Yanus-K, Moscow. Illarioshkin, S. N. 2008. Molecular basis of Parkinson's disease, in: Parkinson's Disease and

Jenner, P. 1994. The contribution of dopamine receptor subtypes to the therapeutic actions

Kaji, R., Urushihara, R., Murase, N., Shimazu, H. & Goto, S. 2005. Abnormal sensory gating in basal ganglia disorders. J. Neurol. 2005, Vol. 252, Suppl 4, 1V/13-1V/16. Karaban, N., Lukhanina, E. P., Melnik, N. A. & Berezetskaya, N. M. 2006. Influence of course

Katayama, Y., Kasai, M., Oshima, H., Fukaya, C. & Yamamoto T. 2000. Effects of

Kawaguchi, Y. & Shindou, T. 1998. Noradrenergic excitation and inhibition of GABAergic

Krnjevic, K., Randic, M. & Straughan, D.W. 1966. Nature of a cortical inhibitory process. J.

Kryzhanovskii, G. N., Karaban', I. N., Magayeva, S. V. et al. 2002. Parkinson's Disease [in

Lewis, G.N. & Byblow, W.D. 2002. Altered sensorimotor integration in Parkinson's disease.

disease. Stereotact. Funct. Neurosurg., Vol. 74, No. 3-4, pp. 99-105.

cell types in rat frontal cortex. J. Neurosci., Vol. 18, pp. 6963-6976.

levodopa as initial treatment for Parkinson's disease: a 4-year randomized

accuracy of clinical diagnosis in Parkinson's disease: a clinico-pathologic study.

evidence that Cerebrolysin protects cortical neurons from neurodegeneration in

Dissociation between contingent negative variation (CNV) and Bereitschaftspotential (BP) in patients with parkinsonism. Electroencephalogr. Clin.

Motor Disorders, pp. 8-17, Handbook for Physicians Based on Proceedings of I

and side-effects of anti-parkinsonian drugs. In: Beyond the Decade of the Brain, Stern, M.B., editor, pp. 131-156, Wells Medical Limited Chapel Place, Royal

treatment with Noofen on motor activity, cognitive functions and emotional state in patients with Parkinson's disease. Ukrainskiy Vestnik Psihonevrologii, vol. 14,

anterodorsal pallidal stimulation on gait freezing (Kinesia paradoxa) in Parkinson's

cortex. J. Neurophysiol., Vol. 86, pp. 2911-2918.

41, pp. 99-114.

Neurology, Vol. 42, pp. 1142-1146.

National Congress, Moscow.

Tunbridge Wells, Kent.

No. 46, pp. 26–30.

Physiol., Vol. 184, pp. 49-77.

Russian], Meditsina, Moscow.

Brain, Vol. 125, pp. 2089-2099.

vitro. J. Neural. Transm., Vol. 53, pp. 363-372.

Neurophysiol., Vol. 102, No. 2, pp. 142 - 151..

disorders. Brain Dev., Vol. 22, Suppl. 1, pp. 132-135.

controlled trial. Arch. Neurol., Vol. 61, pp. 1044-1053.


**3** 

Juliana Dushanova

*Bulgaria* 

**Brain Event - Related Oscillations** 

**in Parkinsonian Patients During Discrimination Task Conditions** 

*Institute of Neurobiology, Bulgarian Academy of Sciences* 

Parkinson's disease (PD) is caused by a disruption of dopaminergic neurotransmission in the basal ganglia, which serve as an integrative centre for the sensory and cognitive processing of information and as a mutual link between this processing and disturbed motor performance. The basal ganglia and the cerebellum transmit information via the thalamus to the cerebral cortex in order to regulate movement. The neurotransmitter changes affect the output of the striatum into the globus pallidus as well as into the thalamus and cerebral cortex beyond. The disease is a common and disabling disorder of movement characterized by poverty, slowness and impaired scaling of voluntary movements (akinesia and bradykinesia), muscle rigidity, and tremor of the limbs at rest. Alterations of the basal ganglia with proven neuronal degenerative disorders of dopaminergic neurons and a reduction in activity in frontostriatal neural circuitry have been suggested to play a role in the executive dysfunction of PD (Taylor et al., 1990; Innis et al., 1993; Lewis et al., 2003; Owen, 2004; Leblois et al., 2006; Anik et al., 2007). The slowed information processing, insufficient encoding strategies and planning, and attentional setshifting are related to memory deficits and cognitive impairment in PD (Daum et al., 1995; Pillon et al., 1997; Knoke et al., 1998; Robertson & Empson, 1999; Sawamoto et al., 2002; Cools, 2006). Neuropsychological studies of PD patients report cognitive deficits even during the early stages of the disease (van Spaendonck et al., 2006). The primary working memory deficit

in PD is associated with impaired free recall performances (Higginson et al., 2003).

stimulus classification and attentional processing (Robertson & Empson, 1999).

Many electroencephalographic (EEG) studies on PD have used the event-related potential (ERP) method, where the early modal dependent and obligatory N1 and P2 components permit analysis of sensory events while the later N2 and P3 potentials reflect the cognitive processes involving the assessment of stimuli, decision making, strategy selection and recognition memory. ERP investigations have shown P3 predominantly with prolonged latencies and/or diminished amplitudes for Parkinsonian patients (PP) when compared to healthy subjects (HS) (Evarts et al., 1981; Tachibana et al., 1992; Philipova et al., 1997; Wascher et al., 1997; Minamoto et al., 2001; Antal et al., 2002; Wang et al., 2002). Such results have been interpreted as electrophysiological signs of cognitive slowing with respect to

One valuable means of assessing deviations from the normal state in PD is to study oscillatory brain processes. In the ERP method, however, the functional significance of the

**1. Introduction**


## **Brain Event - Related Oscillations in Parkinsonian Patients During Discrimination Task Conditions**

Juliana Dushanova *Institute of Neurobiology, Bulgarian Academy of Sciences Bulgaria* 

## **1. Introduction**

58 Diagnostics and Rehabilitation of Parkinson's Disease

Sherman M. Y. & Goldberg A. L. 2001. Cellular defenses against unfolded proteins: a cell

Stanley, H.E., Amaral, L.A.N., Goldberger, A.L., Havlin, S., Ivanov P.Ch. & Peng, C.-K. 1999.

Swie, Y. W., Sakamoto, K. & Ahimizu, Y. 2005. Chaotic analysis of electromyography signal

Szabo, B., Wallmichrath, I., Mathonia, P. & Pfreundtner, C. 2000. Cannabinoids inhibit

Talebinejad, M., Chan, A.D. & Miri, A. 2010. Fatigue estimation using a novel multi-fractal

Teo, Ch., Rasco, L., Al-Mefty, K., Skinner, R.D., Boop, F.A. & Garcia-Ril,l E. 1997. Decreased

Valls-Solé, J. & Valldeoriola, F. 2002. Neurophysiological correlate of clinical signs in

Zappoli, R., Versari, A., Arnetoli, G., Paganini, M., Muscas, G. C., Arneodo, M. G., Gangemi,

Parkinson' s disease. Clin. Neurophysiol., Vol. 113, pp. 792-805.

Electromyogr. Clin. Neurophysiol, Vol. 45, pp. 329-342.

Physica, Vol. 270, p. 309.

Vol. 97, No. 1, pp. 89–97.

Disord., Vol. 12, pp. 655-664.

21, No. 5-6, pp. 473-483.

№ 3, pp. 433-439.

biologist thinks about neurodegenerative disease. Neuron, Vol 29, No. 1, pp. 15-32.

Statistical physics and physiology: monofractal and multifractal approaches.

at low back and lower limb muscles during forward bending posture.

excitatory neurotransmission in the substantia nigra pars reticulate. Neuroscience,

detrended fluctuation analysis-based approach. J. Electromyogr. Kinesiol, Vol. 20,

habituation of midlatency auditory evoked responses in Parkinson's disease. Mov.

P. F. & Bartelli M. 1991. Topographic CNV activity mapping, presenile mild primary cognitive decline and Alzheimer - type dementia. Neurophysiol. Clin., Vol. Parkinson's disease (PD) is caused by a disruption of dopaminergic neurotransmission in the basal ganglia, which serve as an integrative centre for the sensory and cognitive processing of information and as a mutual link between this processing and disturbed motor performance. The basal ganglia and the cerebellum transmit information via the thalamus to the cerebral cortex in order to regulate movement. The neurotransmitter changes affect the output of the striatum into the globus pallidus as well as into the thalamus and cerebral cortex beyond. The disease is a common and disabling disorder of movement characterized by poverty, slowness and impaired scaling of voluntary movements (akinesia and bradykinesia), muscle rigidity, and tremor of the limbs at rest. Alterations of the basal ganglia with proven neuronal degenerative disorders of dopaminergic neurons and a reduction in activity in frontostriatal neural circuitry have been suggested to play a role in the executive dysfunction of PD (Taylor et al., 1990; Innis et al., 1993; Lewis et al., 2003; Owen, 2004; Leblois et al., 2006; Anik et al., 2007). The slowed information processing, insufficient encoding strategies and planning, and attentional setshifting are related to memory deficits and cognitive impairment in PD (Daum et al., 1995; Pillon et al., 1997; Knoke et al., 1998; Robertson & Empson, 1999; Sawamoto et al., 2002; Cools, 2006). Neuropsychological studies of PD patients report cognitive deficits even during the early stages of the disease (van Spaendonck et al., 2006). The primary working memory deficit in PD is associated with impaired free recall performances (Higginson et al., 2003).

Many electroencephalographic (EEG) studies on PD have used the event-related potential (ERP) method, where the early modal dependent and obligatory N1 and P2 components permit analysis of sensory events while the later N2 and P3 potentials reflect the cognitive processes involving the assessment of stimuli, decision making, strategy selection and recognition memory. ERP investigations have shown P3 predominantly with prolonged latencies and/or diminished amplitudes for Parkinsonian patients (PP) when compared to healthy subjects (HS) (Evarts et al., 1981; Tachibana et al., 1992; Philipova et al., 1997; Wascher et al., 1997; Minamoto et al., 2001; Antal et al., 2002; Wang et al., 2002). Such results have been interpreted as electrophysiological signs of cognitive slowing with respect to stimulus classification and attentional processing (Robertson & Empson, 1999).

One valuable means of assessing deviations from the normal state in PD is to study oscillatory brain processes. In the ERP method, however, the functional significance of the

Brain Event - Related Oscillations

2003; Marsden et al., 2001).

**2. Methods** 

**2.1.1 Subjects** 

**2.1 Experimental procedure** 

**2.1.2 Stimuli and task** 

obtained according to the declaration of Helsinki.

in Parkinsonian Patients During Discrimination Task Conditions 61

hence the suggestion that synchronization of the activity of populations of basal ganglia neurons in the gamma band may facilitate motor processing (Brown, 2003). Investigations based on local field potentials recorded from the STN in PP show increased power in the beta range (13–35 Hz) while the patient is at rest (Cassidy et al., 2002; Levy et al., 2002; Brown et al., 2001; Priori et al., 2004). This suggests that there is excessive synchrony in the basal ganglia networks in PD and some of the clinical signs of the disease, it is proposed, stem from this abnormal synchrony between basal ganglia and cortical circuits (Brown,

The aim of the present study was to investigate the functional relationships between oscillatory EEG-dominant components with ERD/ERS method for PP and HS during auditory discrimination tasks within two poststimulus intervals of 0–250 and 250–600 ms. We first focused on time-frequency analysis of delta, theta and alpha rhythms, the appearance of which is well established in PD and is thought to reflect the degree of cortical activation during the information processing. We later shift our focus to the beta and gamma bands, where our aim is to assess the differences between PP and HS in these frequency bands and check an assumption that some PP clinical symptoms stem from excessive synchrony between the basal ganglia and cortical circuits. This investigation of the oscillatory processes and ERD/ERS in HS and PP could contribute to clarification of the

We investigated eleven voluntary patients with a mean age of 61 ± 12.2 years (±s.d.; 7 males, 4 females) with a diagnosis of idiopathic Parkinson's disease for no longer than 2.8 years, assessed by a neurologist at the University Neurological Hospital, with score of I on the Hoehn–Yahr scale of motor function (Hoehn & Yahr, 1967). Patients receiving levodopa (L– dopa) drugs (Sinemet) were included in order to reduce the heterogeneity in the medication. During the experimental session, all patients were in off-phase of the medication. None of the patients had dementia, depression, a presence of atherosclerosis, attendant neurological complications or pronounced tremor. The same number of healthy volunteers was included as aged-matched healthy controls with a mean age of 59.5 ± 9.5 years. Screening confirmed that subjects were free of past or current psychiatric and neurological disorders. All participants were right handed and without deficits in hearing. Handedness was assessed by a questionnaire adapted from the Edinburgh Handedness Inventory (Oldfield, 1971). The study was performed with the approval of the local ethics committee. The subjects were introduced to the nature of the investigation and their informed written consent was

Each subject was comfortably seated in an ergonomically designed chair inside a Faraday cage, monitored by a Canon Video system. The experimental design included a binary sensory-motor reaction task. Each sensory-motor series consisted of 50 computer generated low frequency (LT – 800 Hz) and 50 high frequency (HT – 1000 Hz) acoustic stimuli with an intensity of 60 dB, duration of 50 ms, and an inter-stimulus interval of 2.5–3.5 s presented to the subject in a randomized order. PP and HS were asked to press a key with the index

disturbances of the neurophysiological mechanisms of this disease.

responses in different frequency bands is lost. More clarification could be expected when attentional processes in PD during a representation of discrimination tasks (Vieregge et al., 1994) are examined using event-related desynchronization/synchronization (ERD/ERS) method. In the early stages, PD also affects cognitive functions (Cools, 2006). Cognitive processes require transient integration between different brain areas. Hence, dynamic links are formed, mediated by the ERS or ERD of neuronal assemblies. ERD is defined as a relative decrease in the power of a certain frequency band during stimulus processing, while ERS is a relative increase in the power of the same frequency (Pfurtscheller & Klimesch, 1991). The ERD/ERS method has been used to study auditory and visual working memory encoding and categorization processes in PP; studies indicate less theta-ERS and upper alpha-ERD reflected disturbance of both the basal ganglia activity as well as activity related to their thalamo-cortical neuronal nets at frontal electrode locations (Schmiedt et al., 2005; Ellfolk et al., 2006). The encoding of auditory stimuli elicits alpha- and theta-ERS, while memory retrieval during the presentation of a target stimulus elicits theta-ERS and alpha-ERD (Karrasch et al., 1998, 2004; Krause et al., 1996, 1999). Oscillations in the beta frequency band are associated with cognitive control of behaviour or "executive functions" (Pfurtscheller & Lopes da Silva, 1999; Engel et al., 2001). By means of an auditory stop-signal task, the differential participation of beta subbands in voluntary motor control can be revealed: ERD in the 20–30 Hz band is related to initiation of movement, while ERS in a low frequency beta band (12–16 Hz) is exclusively linked to the stopping of planned action (Pfurtscheller & Lopes da Silva, 1999; Engel et al., 2001). One proposed hypothesis for these observations is that lower-frequency beta subbands represent inhibitory components of cognitive control and are more generalized, while higher frequency beta subbands take part in response choice and activation and are more specialized in terms of both function and cortical distribution.

Some recent investigations (Basar, 2001; Ozgoren et al., 2005; Sutoh et al., 2000; Gurtubay et al., 2001) propose that beta and gamma cortical rhythms may serve cognitive processes such as linking perception to action or movement planning (Donoghue et al., 1998). Research both on animals and humans has suggested that gamma-frequency activity also plays an important role in attention as well as working and long-term memory (Herrmann et al., 2004). Current investigations using intracranial and high-density electro- and magnetoencephalographic recordings explore the involvement of gamma-band synchronization in various cognitive paradigms in humans (Engel et al., 2001; Basar et al., 2001; Herrmann et al., 2004; Pantev, 1995; Tallon-Baudry et al., 1996; Farmer, 1998; Fries et al., 2001). Other works associate the changes in EEG spectral power in the gamma frequency band to interactions between the cortex and basal ganglia (Gatev & Wichmann, 2008). Additionally, akinesia in PP has been related in some studies to abnormally increased beta (15–30 Hz) and decreased gamma (35–80 Hz) synchronous oscillatory activity in the basal ganglia (Weinberger et al., 2006). Other results suggest that resting tremor in PD is associated with an altered balance between beta and gamma oscillations in the motor circuits of the subthalamic nucleus (STN) and is exhibited as increased oscillatory activity in the low gamma frequency range (35–55 Hz) during periods with stronger tremor (Weinberger et al., 2008). Therapeutic doses of dopaminergic medication in PP attenuate the beta band power in the STN, giving rise to the hypothesis that the beta prominence is pathological in PD (Cassidy et al, 2002; Kühn et al., 2006; Levy et al., 2001; 2002; Priori et al., 2002). Treatment of PP with dopaminergic therapy leads to increased gamma band activity in the basal ganglia and thus to improvement in motor performance (Brown et al., 2001), hence the suggestion that synchronization of the activity of populations of basal ganglia neurons in the gamma band may facilitate motor processing (Brown, 2003). Investigations based on local field potentials recorded from the STN in PP show increased power in the beta range (13–35 Hz) while the patient is at rest (Cassidy et al., 2002; Levy et al., 2002; Brown et al., 2001; Priori et al., 2004). This suggests that there is excessive synchrony in the basal ganglia networks in PD and some of the clinical signs of the disease, it is proposed, stem from this abnormal synchrony between basal ganglia and cortical circuits (Brown, 2003; Marsden et al., 2001).

The aim of the present study was to investigate the functional relationships between oscillatory EEG-dominant components with ERD/ERS method for PP and HS during auditory discrimination tasks within two poststimulus intervals of 0–250 and 250–600 ms. We first focused on time-frequency analysis of delta, theta and alpha rhythms, the appearance of which is well established in PD and is thought to reflect the degree of cortical activation during the information processing. We later shift our focus to the beta and gamma bands, where our aim is to assess the differences between PP and HS in these frequency bands and check an assumption that some PP clinical symptoms stem from excessive synchrony between the basal ganglia and cortical circuits. This investigation of the oscillatory processes and ERD/ERS in HS and PP could contribute to clarification of the disturbances of the neurophysiological mechanisms of this disease.

## **2. Methods**

60 Diagnostics and Rehabilitation of Parkinson's Disease

responses in different frequency bands is lost. More clarification could be expected when attentional processes in PD during a representation of discrimination tasks (Vieregge et al., 1994) are examined using event-related desynchronization/synchronization (ERD/ERS) method. In the early stages, PD also affects cognitive functions (Cools, 2006). Cognitive processes require transient integration between different brain areas. Hence, dynamic links are formed, mediated by the ERS or ERD of neuronal assemblies. ERD is defined as a relative decrease in the power of a certain frequency band during stimulus processing, while ERS is a relative increase in the power of the same frequency (Pfurtscheller & Klimesch, 1991). The ERD/ERS method has been used to study auditory and visual working memory encoding and categorization processes in PP; studies indicate less theta-ERS and upper alpha-ERD reflected disturbance of both the basal ganglia activity as well as activity related to their thalamo-cortical neuronal nets at frontal electrode locations (Schmiedt et al., 2005; Ellfolk et al., 2006). The encoding of auditory stimuli elicits alpha- and theta-ERS, while memory retrieval during the presentation of a target stimulus elicits theta-ERS and alpha-ERD (Karrasch et al., 1998, 2004; Krause et al., 1996, 1999). Oscillations in the beta frequency band are associated with cognitive control of behaviour or "executive functions" (Pfurtscheller & Lopes da Silva, 1999; Engel et al., 2001). By means of an auditory stop-signal task, the differential participation of beta subbands in voluntary motor control can be revealed: ERD in the 20–30 Hz band is related to initiation of movement, while ERS in a low frequency beta band (12–16 Hz) is exclusively linked to the stopping of planned action (Pfurtscheller & Lopes da Silva, 1999; Engel et al., 2001). One proposed hypothesis for these observations is that lower-frequency beta subbands represent inhibitory components of cognitive control and are more generalized, while higher frequency beta subbands take part in response choice and activation and are more specialized in terms of both function and

Some recent investigations (Basar, 2001; Ozgoren et al., 2005; Sutoh et al., 2000; Gurtubay et al., 2001) propose that beta and gamma cortical rhythms may serve cognitive processes such as linking perception to action or movement planning (Donoghue et al., 1998). Research both on animals and humans has suggested that gamma-frequency activity also plays an important role in attention as well as working and long-term memory (Herrmann et al., 2004). Current investigations using intracranial and high-density electro- and magnetoencephalographic recordings explore the involvement of gamma-band synchronization in various cognitive paradigms in humans (Engel et al., 2001; Basar et al., 2001; Herrmann et al., 2004; Pantev, 1995; Tallon-Baudry et al., 1996; Farmer, 1998; Fries et al., 2001). Other works associate the changes in EEG spectral power in the gamma frequency band to interactions between the cortex and basal ganglia (Gatev & Wichmann, 2008). Additionally, akinesia in PP has been related in some studies to abnormally increased beta (15–30 Hz) and decreased gamma (35–80 Hz) synchronous oscillatory activity in the basal ganglia (Weinberger et al., 2006). Other results suggest that resting tremor in PD is associated with an altered balance between beta and gamma oscillations in the motor circuits of the subthalamic nucleus (STN) and is exhibited as increased oscillatory activity in the low gamma frequency range (35–55 Hz) during periods with stronger tremor (Weinberger et al., 2008). Therapeutic doses of dopaminergic medication in PP attenuate the beta band power in the STN, giving rise to the hypothesis that the beta prominence is pathological in PD (Cassidy et al, 2002; Kühn et al., 2006; Levy et al., 2001; 2002; Priori et al., 2002). Treatment of PP with dopaminergic therapy leads to increased gamma band activity in the basal ganglia and thus to improvement in motor performance (Brown et al., 2001),

cortical distribution.

## **2.1 Experimental procedure**

#### **2.1.1 Subjects**

We investigated eleven voluntary patients with a mean age of 61 ± 12.2 years (±s.d.; 7 males, 4 females) with a diagnosis of idiopathic Parkinson's disease for no longer than 2.8 years, assessed by a neurologist at the University Neurological Hospital, with score of I on the Hoehn–Yahr scale of motor function (Hoehn & Yahr, 1967). Patients receiving levodopa (L– dopa) drugs (Sinemet) were included in order to reduce the heterogeneity in the medication. During the experimental session, all patients were in off-phase of the medication. None of the patients had dementia, depression, a presence of atherosclerosis, attendant neurological complications or pronounced tremor. The same number of healthy volunteers was included as aged-matched healthy controls with a mean age of 59.5 ± 9.5 years. Screening confirmed that subjects were free of past or current psychiatric and neurological disorders. All participants were right handed and without deficits in hearing. Handedness was assessed by a questionnaire adapted from the Edinburgh Handedness Inventory (Oldfield, 1971). The study was performed with the approval of the local ethics committee. The subjects were introduced to the nature of the investigation and their informed written consent was obtained according to the declaration of Helsinki.

#### **2.1.2 Stimuli and task**

Each subject was comfortably seated in an ergonomically designed chair inside a Faraday cage, monitored by a Canon Video system. The experimental design included a binary sensory-motor reaction task. Each sensory-motor series consisted of 50 computer generated low frequency (LT – 800 Hz) and 50 high frequency (HT – 1000 Hz) acoustic stimuli with an intensity of 60 dB, duration of 50 ms, and an inter-stimulus interval of 2.5–3.5 s presented to the subject in a randomized order. PP and HS were asked to press a key with the index

Brain Event - Related Oscillations

ms) and T2 (250–600 ms).

largest peaks for each pair of channels.

**2.2.1 Statistics** 

**3. Experimental results 3.1 Response parameters** 

in Parkinsonian Patients During Discrimination Task Conditions 63

interval (T2: 250–600 ms), defined as beginning when a tone ends. The peak amplitude modulations, defined in 10 ms bins, were specified as dominating components. Afterwards, the high amplitudes in each frequency band, respectively, were added over trials and across subjects to compare their amplitude changes in the post-stimulus intervals with those in the reference interval. Thus, we calculated the ERD/ERS of delta (δ ~ 1.5–4 Hz), theta (θ ~ 4.1–7 Hz) and alpha (α ~ 7.1–13 Hz) waves as percentage power differences in each frequency band compared to the reference interval for both 0–250 ms and 250–600 ms post-stimulus intervals. We also defined the ERD/ERS of beta (β1 ~ 13.1–20 Hz), (β 2 ~ 20.1–32 Hz) and gamma (γ ~ 32.1–50 Hz) frequency rhythms during the post-stimulus intervals T1 (0–250

We employed the detection of the temporally dynamic processes similar to the approach applied by Foffani et al. (Foffani et al., 2005) that describes the behaviour of β1 -, β<sup>2</sup> -, γ - ERS rhythms in zones which vary both in amplitude and frequency. Each zone *i* is characterized by a value *ERSi* (*t*) and a frequency value *Fi* (*t*), both dependent on time, which separately describe event-related synchronization and corresponding frequency modulations for the beta1, beta2 and gamma rhythms in two post-stimulus time windows. The significance of the observed ERS values was tested for each frequency band and EEG channel using a permutation test, including corrections for multiple comparisons between time points for time course analysis (Mason & Newton, 1990). The latencies of ERS and corresponding frequency (relative to stimulus onset) were measured as the last zero-crossing before a significant modulation, after subtracting the baseline mean ERS value. Although the ERS clearly occurred, the relationship between the ERS peaks and the maximum of the average event-related synchronization (AERS) was not always evident. Since the AERSs of different channels were not identical, an exact coincidence between the peaks times was not observed. The probabilities for the amplitudes and latencies were not uniform and the activity distribution was clearly not Gaussian. We estimated the latency shift between the

We performed statistical analyses of the ERD/ERS for the two post-stimulus intervals and assessed the statistical difference between the groups (PP and HS) for each tone type and interval by means of a bootstrap nonparametric procedure (Mason & Newton, 1990). The characteristics were grouped by tone, interval, patients and healthy controls and analyzed by means of a permutation test for multiple comparisons (Mason & Newton, 1990). The computed random distribution for interval was analyzed with a nonparametric test (Kruskal-Wallis [KW] test, *p* < 0.05) for pairs comparison of the scalp leads between patient and control group. This procedure reduces the influence of random variations in experimental conditions between trials. The ERD/ERS analysis served to identify the most robust differences between groups and was generally done for the two time windows. The parameters of the movement performance (onset of reaction time, force peak latency and

Parkinson's patients showed a longer reaction time onset, but the difference between the two groups was not significant: in response to a low tone—440.5 ± 135.8 ms in HS and 508.4

error of performance) were processed statistically by Mann–Whitney U test.

finger of each hand and make rapid and accurate choice responses with the left hand to the high frequency (HT) or with the right hand to the low frequency (LT) tone. The movement performance from the stimulus presentation to the onset of voluntary force production (onset of reaction time) and from the stimulus presentation to the force peak (force peak latency) were measured by a force transducer. A surface electromyographic activity pattern of the first dorsal interosseus muscles was also registered.

## **2.1.3 EEG recording**

An electroencephalogram (EEG) was recorded from Fz, Cz, Pz, C3' and C4' (10/20, system), using Ag/AgCl Nihon-Kohden electrodes with a reference to both processi mastoidei and a ground electrode, placed on the forehead. An oculogram (EOG) was recorded from m. orbicularis oculi dex. We placed two EOG electrodes next to the eyes to register eye movements. EEG and EOG data were recorded using a Nihon-Kohden EEG-4314F (cut-off frequencies of 0.3–70 Hz) and recorded together with markers of the movement performance as a force profile and a surface electromyographic activity pattern of the first dorsal interosseus muscles (bandpass filtered 0.03–500 Hz). The signals were digitized online (10 bit A/D converter, 256 samples/s). The data recordings for the sensorimotor task were synchronized to the marker of the stimulus onset (-0.2 s before and 0.8 s after the stimulus). Only recordings that were artifact-free with respect to event-related potentials were processed. We applied a Chebyshev Type II bandpass filter (1-70 Hz) and secondorder notch filter at frequency 50 Hz (AC) component. We defined an independent reference interval in order to quantify the changes in the time-frequency energy density of the signal. We used stimulus-nonrelated subepochs within the resting condition series, distant enough (-1.4 s) from the stimulus onset, not including event-related properties, and exceeding the period of the lowest frequency studied in the signal (1.5 Hz, 0.67 s). We preselected trials by applying a bootstrap estimation within the reference period and a false discovery rate correction for multiple comparisons (0.05) to the available data across the indexes corresponding to time and number of the trial, eliminating the need for a strict assumption of ergodicity (Durka et al., 2004).

#### **2.2 Analysis**

The time-frequency analysis (TF) represented the power of a continuous EEG signal as a function of both time and frequency (Matlab®, Mathworks, Inc.). For time amplitudefrequency distributions, the filtered signal was analyzed with a sliding-window fast Fourier transform with length 200 ms and step 10 ms. The amplitude was computed for every time window t and frequency bin f by the real and imaginary Fourier coefficients. The amplitude modulations obtained for each frequency band for each subject in a group were added across trials in order to compare amplitude changes in the post-stimulus intervals with respect to pre-stimulus interval reference amplitudes, i.e. to derive ERD/ERS. This method characterizes the relative amplitude decrement/increment of the given frequency during the post-stimulus period in relation to pre-stimulus amplitude modulation of the same frequency (Pfurtscheller G, Klimesch, 1991). This resulted in ERD/ERS values which could then be presented as percentage changes with respect to the reference interval. Negative values indicate a relative power decrease (ERD), whereas positive values point to a relative power increase (ERS). Relatively, sensory processing takes place during the first poststimulus interval (T1: 0–250 ms) and cognitive processing during the second post-stimulus interval (T2: 250–600 ms), defined as beginning when a tone ends. The peak amplitude modulations, defined in 10 ms bins, were specified as dominating components. Afterwards, the high amplitudes in each frequency band, respectively, were added over trials and across subjects to compare their amplitude changes in the post-stimulus intervals with those in the reference interval. Thus, we calculated the ERD/ERS of delta (δ ~ 1.5–4 Hz), theta (θ ~ 4.1–7 Hz) and alpha (α ~ 7.1–13 Hz) waves as percentage power differences in each frequency band compared to the reference interval for both 0–250 ms and 250–600 ms post-stimulus intervals. We also defined the ERD/ERS of beta (β1 ~ 13.1–20 Hz), (β 2 ~ 20.1–32 Hz) and gamma (γ ~ 32.1–50 Hz) frequency rhythms during the post-stimulus intervals T1 (0–250 ms) and T2 (250–600 ms).

We employed the detection of the temporally dynamic processes similar to the approach applied by Foffani et al. (Foffani et al., 2005) that describes the behaviour of β1 -, β<sup>2</sup> -, γ - ERS rhythms in zones which vary both in amplitude and frequency. Each zone *i* is characterized by a value *ERSi* (*t*) and a frequency value *Fi* (*t*), both dependent on time, which separately describe event-related synchronization and corresponding frequency modulations for the beta1, beta2 and gamma rhythms in two post-stimulus time windows. The significance of the observed ERS values was tested for each frequency band and EEG channel using a permutation test, including corrections for multiple comparisons between time points for time course analysis (Mason & Newton, 1990). The latencies of ERS and corresponding frequency (relative to stimulus onset) were measured as the last zero-crossing before a significant modulation, after subtracting the baseline mean ERS value. Although the ERS clearly occurred, the relationship between the ERS peaks and the maximum of the average event-related synchronization (AERS) was not always evident. Since the AERSs of different channels were not identical, an exact coincidence between the peaks times was not observed. The probabilities for the amplitudes and latencies were not uniform and the activity distribution was clearly not Gaussian. We estimated the latency shift between the largest peaks for each pair of channels.

## **2.2.1 Statistics**

62 Diagnostics and Rehabilitation of Parkinson's Disease

finger of each hand and make rapid and accurate choice responses with the left hand to the high frequency (HT) or with the right hand to the low frequency (LT) tone. The movement performance from the stimulus presentation to the onset of voluntary force production (onset of reaction time) and from the stimulus presentation to the force peak (force peak latency) were measured by a force transducer. A surface electromyographic activity pattern

An electroencephalogram (EEG) was recorded from Fz, Cz, Pz, C3' and C4' (10/20, system), using Ag/AgCl Nihon-Kohden electrodes with a reference to both processi mastoidei and a ground electrode, placed on the forehead. An oculogram (EOG) was recorded from m. orbicularis oculi dex. We placed two EOG electrodes next to the eyes to register eye movements. EEG and EOG data were recorded using a Nihon-Kohden EEG-4314F (cut-off frequencies of 0.3–70 Hz) and recorded together with markers of the movement performance as a force profile and a surface electromyographic activity pattern of the first dorsal interosseus muscles (bandpass filtered 0.03–500 Hz). The signals were digitized online (10 bit A/D converter, 256 samples/s). The data recordings for the sensorimotor task were synchronized to the marker of the stimulus onset (-0.2 s before and 0.8 s after the stimulus). Only recordings that were artifact-free with respect to event-related potentials were processed. We applied a Chebyshev Type II bandpass filter (1-70 Hz) and secondorder notch filter at frequency 50 Hz (AC) component. We defined an independent reference interval in order to quantify the changes in the time-frequency energy density of the signal. We used stimulus-nonrelated subepochs within the resting condition series, distant enough (-1.4 s) from the stimulus onset, not including event-related properties, and exceeding the period of the lowest frequency studied in the signal (1.5 Hz, 0.67 s). We preselected trials by applying a bootstrap estimation within the reference period and a false discovery rate correction for multiple comparisons (0.05) to the available data across the indexes corresponding to time and number of the trial, eliminating the need for a strict assumption

The time-frequency analysis (TF) represented the power of a continuous EEG signal as a function of both time and frequency (Matlab®, Mathworks, Inc.). For time amplitudefrequency distributions, the filtered signal was analyzed with a sliding-window fast Fourier transform with length 200 ms and step 10 ms. The amplitude was computed for every time window t and frequency bin f by the real and imaginary Fourier coefficients. The amplitude modulations obtained for each frequency band for each subject in a group were added across trials in order to compare amplitude changes in the post-stimulus intervals with respect to pre-stimulus interval reference amplitudes, i.e. to derive ERD/ERS. This method characterizes the relative amplitude decrement/increment of the given frequency during the post-stimulus period in relation to pre-stimulus amplitude modulation of the same frequency (Pfurtscheller G, Klimesch, 1991). This resulted in ERD/ERS values which could then be presented as percentage changes with respect to the reference interval. Negative values indicate a relative power decrease (ERD), whereas positive values point to a relative power increase (ERS). Relatively, sensory processing takes place during the first poststimulus interval (T1: 0–250 ms) and cognitive processing during the second post-stimulus

of the first dorsal interosseus muscles was also registered.

**2.1.3 EEG recording** 

of ergodicity (Durka et al., 2004).

**2.2 Analysis** 

We performed statistical analyses of the ERD/ERS for the two post-stimulus intervals and assessed the statistical difference between the groups (PP and HS) for each tone type and interval by means of a bootstrap nonparametric procedure (Mason & Newton, 1990). The characteristics were grouped by tone, interval, patients and healthy controls and analyzed by means of a permutation test for multiple comparisons (Mason & Newton, 1990). The computed random distribution for interval was analyzed with a nonparametric test (Kruskal-Wallis [KW] test, *p* < 0.05) for pairs comparison of the scalp leads between patient and control group. This procedure reduces the influence of random variations in experimental conditions between trials. The ERD/ERS analysis served to identify the most robust differences between groups and was generally done for the two time windows. The parameters of the movement performance (onset of reaction time, force peak latency and error of performance) were processed statistically by Mann–Whitney U test.

## **3. Experimental results**

#### **3.1 Response parameters**

Parkinson's patients showed a longer reaction time onset, but the difference between the two groups was not significant: in response to a low tone—440.5 ± 135.8 ms in HS and 508.4

Brain Event - Related Oscillations

(right) at all channels.

**3.2.3 Alpha** 

in Parkinsonian Patients During Discrimination Task Conditions 65

Fig. 1. Group means (±SE) and statistical results of δ-, θ-, α-ERD/ERS over all HS and PP trials after the low tone (A) and high tone type (B) for the early (left) and late time period

the HS and a weak θ-ERD in the PP following the high frequency tone.

pronounced θ-ERS in the PP and smaller θ-ERD in the HS.

right motor areas (*p* < 0.05), parietal side (*p* < 0.001), and centrally was non-significantly different (Cz, *p* > 0.05; Fig. 1A, 2nd row, left). Following the high frequency tone, PP produced a significantly lower θ-ERS than the HS at right motor area, fronto-central and parietal sides (*p* < 0.001; Fig. 1B, left). The left motor area showed a pronounced θ-ERS for

The comparison of the groups during the late period following the low frequency tone showed θ-ERD in both groups at most electrodes with the following exceptions. PP recorded a large parietal θ-ERS while a less prominent θ-ERS appeared in HS at frontal and right motor areas (Fig. 1A, 2nd row, right). The most pronounced θ-ERDs for HS were at central and left motor areas (*p* < 0.001). The signal at the parietal area was characterized by a very prominent θ-ERS in the patients but by θ-ERD in the control group. In response to the high frequency tone, we found a significant θ-ERD for the PP as compared with the HS at frontocentral, left and right motor areas (*p* < 0.001; Fig. 1B, right). The signal at the parietal area was characterized in a similar manner to that elicited by a low tone, but with a less

The PP showed fronto-central, left and right motor α-ERS responses following the LT, while the HS had synchronization only at the fronto-central sides but α-ERD at the parietal side,

± 148.5 ms in PP (mean ± S.D., *p* > 0.05), in response to a high tone—455.7 ± 134.6 ms in HS and 500.4 ± 146.5 ms in PP (*p* > 0.05). Parkinson's patients had significantly longer force peak latency (FPL) in response to the two tone types. The FPLs were 672.8 ± 154.7 ms in HS and 919.96 ± 163.7 ms in PP in response to the low tone (mean ± s.d, *p* < 0.02). In response to the high tone, the FPLs were 690.7 ± 148.7 ms in HS and 934.9 ± 160.2 ms in PP (*p* < 0.05). The mean errors (false and missing responses) were 4.5 and 5.1 in PP, respectively, in response to the low and high tone. The mean errors were 3.5 and 4.4 in HS, respectively, in response to the low and high tone. The differences of errors between the two groups were not significant (*p* > 0.05).

#### **3.2 Frequency components**

The grand average ERD/ERS values as a function of time and frequency band at the frontal, central, parietal, left and right motor areas were used for assessment of group means with the standard errors (±SE) for the post-stimulus intervals T1 (0–250 ms) and T2 (250–600 ms). The statistical group comparison for pair channels are shown graphically to illustrate the delta-, theta-, alpha-, beta-, gamma- ERD/ERS following the low frequency tone type (Figs. 1A, 2A) and high frequency tone type (Figs. 1B, 2B) for the early (0–250 ms) and late (250– 600 ms) post-stimulus intervals.

### **3.2.1 Delta**

The patterns of δ-ERD/ERS were different between the groups for both intervals in response to both tone types (Dushanova et al., 2009). In the early post-stimulus interval, central δ-ERS amplitude responses were most pronounced in the HS after both tones (Cz, Fig. 1A, B, 1st row, left plots) and in the PP at the frontal side for the high tone (Fig. 1B, left) and parietal area for the low tone (Fig. 1A, left). The least pronounced δ-ERS were those appearing at the frontal side in the HS and in the left motor area in PP following both tones (Fig. 1A, B, left). The comparison by the bootstrap procedure of δ-ERS between the two groups after the low frequency tone determined that the control δ-ERS was significantly higher than that of the PP for all channels (Fz, Pz, *p* < 0.05; Cz, C3', C4', *p* < 0.001). Both groups displayed δ-ERS following the high frequency tone (Fig. 1B, left). This was significantly higher for the HS than the PP at centro-parietal, left and right motor areas (Cz, C3', C4' *p* < 0.001; Pz, *p* < 0.05). The PD patients' frontal δ-ERS, however, was greater than that of the HS (*p* < 0.001).

The HS maintained δ-ERS at all electrodes during the late post-stimulus interval T2 following the low frequency tone, while in the PP the early post-stimulus δ-ERS was reversed to become δ-ERD in the late post-stimulus interval (Fig. 1A, B, 1st row, right). The highest δ-ERS was located at parietal side for the HS, whereas the PP had a less enhanced parietal δ-ERD. The PP showed the most enhanced δ-ERD at the left motor area (Fig. 1B, right). Following the high frequency tone, δ-ERS was elicited at all electrodes in the HS (Fig. 1B, right). The PP group, in comparison with the HS, showed a less pronounced central δ-ERS (Cz, *p* < 0.001) and specific δ-ERD at parietal and left motor areas (*p* < 0.001).

#### **3.2.2 Theta**

In the first post-stimulus interval following both tone types, the θ-ERS responses were most prominent at parietal electrodes for both groups (Fig. 1, A, B, 2nd row, left). Following the low tone, the θ-ERS elicited was significantly higher for HS than for PP at frontal, left and

± 148.5 ms in PP (mean ± S.D., *p* > 0.05), in response to a high tone—455.7 ± 134.6 ms in HS and 500.4 ± 146.5 ms in PP (*p* > 0.05). Parkinson's patients had significantly longer force peak latency (FPL) in response to the two tone types. The FPLs were 672.8 ± 154.7 ms in HS and 919.96 ± 163.7 ms in PP in response to the low tone (mean ± s.d, *p* < 0.02). In response to the high tone, the FPLs were 690.7 ± 148.7 ms in HS and 934.9 ± 160.2 ms in PP (*p* < 0.05). The mean errors (false and missing responses) were 4.5 and 5.1 in PP, respectively, in response to the low and high tone. The mean errors were 3.5 and 4.4 in HS, respectively, in response to the low and high tone. The differences of errors between the two groups were

The grand average ERD/ERS values as a function of time and frequency band at the frontal, central, parietal, left and right motor areas were used for assessment of group means with the standard errors (±SE) for the post-stimulus intervals T1 (0–250 ms) and T2 (250–600 ms). The statistical group comparison for pair channels are shown graphically to illustrate the delta-, theta-, alpha-, beta-, gamma- ERD/ERS following the low frequency tone type (Figs. 1A, 2A) and high frequency tone type (Figs. 1B, 2B) for the early (0–250 ms) and late (250–

The patterns of δ-ERD/ERS were different between the groups for both intervals in response to both tone types (Dushanova et al., 2009). In the early post-stimulus interval, central δ-ERS amplitude responses were most pronounced in the HS after both tones (Cz, Fig. 1A, B, 1st row, left plots) and in the PP at the frontal side for the high tone (Fig. 1B, left) and parietal area for the low tone (Fig. 1A, left). The least pronounced δ-ERS were those appearing at the frontal side in the HS and in the left motor area in PP following both tones (Fig. 1A, B, left). The comparison by the bootstrap procedure of δ-ERS between the two groups after the low frequency tone determined that the control δ-ERS was significantly higher than that of the PP for all channels (Fz, Pz, *p* < 0.05; Cz, C3', C4', *p* < 0.001). Both groups displayed δ-ERS following the high frequency tone (Fig. 1B, left). This was significantly higher for the HS than the PP at centro-parietal, left and right motor areas (Cz, C3', C4' *p* < 0.001; Pz, *p* < 0.05). The PD patients' frontal δ-ERS, however, was greater than

The HS maintained δ-ERS at all electrodes during the late post-stimulus interval T2 following the low frequency tone, while in the PP the early post-stimulus δ-ERS was reversed to become δ-ERD in the late post-stimulus interval (Fig. 1A, B, 1st row, right). The highest δ-ERS was located at parietal side for the HS, whereas the PP had a less enhanced parietal δ-ERD. The PP showed the most enhanced δ-ERD at the left motor area (Fig. 1B, right). Following the high frequency tone, δ-ERS was elicited at all electrodes in the HS (Fig. 1B, right). The PP group, in comparison with the HS, showed a less pronounced central δ-

In the first post-stimulus interval following both tone types, the θ-ERS responses were most prominent at parietal electrodes for both groups (Fig. 1, A, B, 2nd row, left). Following the low tone, the θ-ERS elicited was significantly higher for HS than for PP at frontal, left and

ERS (Cz, *p* < 0.001) and specific δ-ERD at parietal and left motor areas (*p* < 0.001).

not significant (*p* > 0.05).

**3.2 Frequency components** 

600 ms) post-stimulus intervals.

that of the HS (*p* < 0.001).

**3.2.2 Theta** 

**3.2.1 Delta** 

Fig. 1. Group means (±SE) and statistical results of δ-, θ-, α-ERD/ERS over all HS and PP trials after the low tone (A) and high tone type (B) for the early (left) and late time period (right) at all channels.

right motor areas (*p* < 0.05), parietal side (*p* < 0.001), and centrally was non-significantly different (Cz, *p* > 0.05; Fig. 1A, 2nd row, left). Following the high frequency tone, PP produced a significantly lower θ-ERS than the HS at right motor area, fronto-central and parietal sides (*p* < 0.001; Fig. 1B, left). The left motor area showed a pronounced θ-ERS for the HS and a weak θ-ERD in the PP following the high frequency tone.

The comparison of the groups during the late period following the low frequency tone showed θ-ERD in both groups at most electrodes with the following exceptions. PP recorded a large parietal θ-ERS while a less prominent θ-ERS appeared in HS at frontal and right motor areas (Fig. 1A, 2nd row, right). The most pronounced θ-ERDs for HS were at central and left motor areas (*p* < 0.001). The signal at the parietal area was characterized by a very prominent θ-ERS in the patients but by θ-ERD in the control group. In response to the high frequency tone, we found a significant θ-ERD for the PP as compared with the HS at frontocentral, left and right motor areas (*p* < 0.001; Fig. 1B, right). The signal at the parietal area was characterized in a similar manner to that elicited by a low tone, but with a less pronounced θ-ERS in the PP and smaller θ-ERD in the HS.

#### **3.2.3 Alpha**

The PP showed fronto-central, left and right motor α-ERS responses following the LT, while the HS had synchronization only at the fronto-central sides but α-ERD at the parietal side,

Brain Event - Related Oscillations

**Fz ERS(%)/ F/ t**  (±SE) **D** 

**64.1 (±1.2)/** 13 Hz 40 (±7.3) ms [8–72] ms (**a**)

[13–15] Hz 36 (±6.9) ms [8–64] ms (**a**)

**51 (±3.2)**/20 Hz 516 (±11.5) ms [424–600] ms (**f**)

**69.4(±0.9)\*/** 13 Hz 12 (±4)\*\* ms [8–80] ms (**a**)

15.8(±0.2)/14(±0.6) Hz 124 (±5.2) ms [112–136] ms (**b**)

–

–

–

– –

**68.6(±10.2)**/13.9(±0.3)Hz

**Channels** 

**Subjects Tone(period)** 

**LT (T1)** 

**HT (T1)** 

**LT (T2)** 

**HT (T2)** 

**Subjects Tone(period)** 

**LT (T1)** 

**HT (T1)** 

**LT (T2) HT (T2)** 

period.

in Parkinsonian Patients During Discrimination Task Conditions 67

**Pz ERS(%)/ F/ t**  (±SE) **D HS**

18.8 (±0.5)/13 Hz 20 (±5.2) ms [8–32] ms (**a**)

30.49(±3.3)\*/13Hz 28(±6.1)\*\* ms [8–48] ms (**a**)

–

–

**PP** 

–

–

–

– –

**Note:** Bold-marked mean ERS β1 bursts (%) (±SE) are maximum across the channels for each condition and subject separately for LT(T1), HT(T1), LT(T2), HT(T2); **F** (Hz) – mean frequency peaks (±SE) of the maximum ERS β1 bursts across trials; **t** (ms) – mean times (±SE) of maximum ERS β1 bursts across trials during T1 (or T2) with respect to stimulus onset; **D** (ms) – time duration of these short–term zones with ERS β1 bursts; 1 \* =a significant difference in ERS between the groups HS, PP for each channel and tone separately, marked the higher value (p<0.001, KW test); 2 \*\* =a significant difference in **t** between the groups for each channel and tone separately, marked the shorter value (p<0.001, KW test). Lower case letters in Table 1 marked consecutive sub-intervals with ERS β1 bursts: **a**, **b** during T1; **f** during T2. Table 1. Mean ERS β1 bursts (%) across the trials with mean frequency peaks **F (**Hz) for HS and PP after LT and HT presented in short–term zones **D** (ms) during early T1 and late T2

21.3(±0.8)/ 13 Hz 46 (±8.7) ms [8–88] ms (**a**)

**C3' ERS(%)/ F/ t**  (±SE) **D** 

41 (±3.3)\* /13 Hz 36 (±28) ms [8–80] ms (**a**)

[13–15] Hz 32(±6.5) ms [8–56] ms (**a**)

–

–

–

– –

36.8(±2.8)/13 Hz 12 (±4)\*\* ms [8–16] ms (**a**)

19.3(±1.3)/18 Hz 148 (±5.2) ms [136–160] ms (**b**) 35(±1.9)/ 13 Hz 20(±5.2)\*\* ms [8–32] ms (**a**)

54.7(±8.3)\*/13.5(±0.3)Hz

**C4' ERS(%)/F /t**  (±SE) **D** 

44.9(±5.2)\*/13 Hz 32.0 (±6.5)\*\* ms [8–72] ms (**a**)

50.4(±6.6)\*/13 Hz 32(±6.5) ms [8–56] ms (**a**)

–

–

–

– –

33.9(±5)/13Hz 48 (±6.5) ms [24–72] ms (**a**)

**40.2(±2.8)**/13 Hz 24 (±5.7) \*\* ms [8–40] ms (**a**)

25(±1.1)/16 Hz 92 (±5.2) ms [80–112] ms (**b**)

 **Cz ERS(%)/ F/ t**  (±SE) **D** 

50.1 (±5.8) /13 Hz 36 (±6.9) ms [8–64] ms (**a**)

[13–15] Hz 32 (±6.5)\*\* ms [8–56] ms (**a**)

57.2(±3)**\***/13 Hz 16 (±4.6)\*\* ms [8–64] ms (**a**)

34.4(±3.8)/13.8(±0.3) Hz 53.6 (±11.2) ms [8–104] ms (**a**)

–

–

–

–

– –

62.7(±8.5)\*/13.7(±0.3)Hz

right and left motor areas (Fig. 1A, 3rd row, left). The alpha frequency band in the first interval after the LT showed an enhanced central α-ERS for the PP in comparison with the HS (*p* < 0.001; Fig. 1A, left). The right and left motor areas manifested different reversal alteration as α-ERS for the PP and weakly elicited α-ERD for the HS. Following the high tone, we detected significantly different processes, α-ERS in the PP contrasted with α-ERD in the HS at all electrodes, most prominently at the left motor side (Fig. 1B, left).

Alpha-ERD differences between the groups were observed after low as well as high frequency tones during the late period (Fig. 1A, B, right). After the low tone, the HS α-ERD means were significantly more pronounced than those of the PP at centro-parietal sides, left and right motor areas (*p* < 0.001; Fig. 1A, right). The PP displayed a higher frontal α-ERD than that in the HS (*p* < 0.001). Alpha-ERD differences were observed between the groups following the high tone (Fig. 1B, right). The α-ERD signals of HS at parietal and left motor areas were more pronounced than in the PP (*p* < 0.001). The PP fronto-central and right motor α-ERD signals had greater respective magnitudes than those in HS (*p* < 0.001).

#### **3.2.4 Beta1**

During the early post-stimulus interval T1, ERD/ERS β1 patterns appeared in the lower frequency portions of this band for each electrode and group (Dushanova et al., 2010). The maximum synchronized β1 bursts across the channels were localized over fronto-central sides for both groups after LT, but had significantly higher amplitude and shorter durations in PP than HS (*p* < 0.001, bootstrap, KW test; Table 1, **a**). Later β1 ERS bursts were found only in PP over the frontal side and left motor area (Table 1, LT (T1), **b**). Synchronized β1 bursts were centered on right motor side for PP and fronto-central sides for HS, following HT (Table 1, HT (T1)**, a**, **b**). In PP, the peaks were of a significantly lower amplitude and peaked later at centro-parietal sides than in HS (Table 1, HT (T1)**, a**). During the late poststimulus interval T2, frontal synchronized β1 peaks at 20 Hz were extracted only in HS after HT and had a prolonged latency of 176 ± 11.5 ms (Table 1, HT (T2)**, f**).

#### **3.2.5 Beta2**

During T1, ERS β2 bursts were present only in PP after either tone (Table 2, LT(T1), HT(T1)). Their maximum amplitude across the scalp was localized over frontal side for either tone, right and left motor areas respectively for LT and HT (Table 2, **c**). They peaked earlier at the right than at the left motor area, fronto-central leads (Table 2, **c**), and parietal side (Table 2, **d**) for either tone, but in higher β2 frequency range after LT than after HT. During T2, the maximum value of β2 bursts across all recorded areas was centered on right motor area in PP (Table 2, **g**) but on parietal side in HS following LT (Table 2, LT(T2), **j**). More widely distributed, prolonged β2 ERS bursts were pronounced over all locations for PP following HT, but limited to frontal-central and right motor areas for HS (Table 2, HT(T2)). The maximum scalp β2 burst was localized at the right motor area in PP and frontal area in HS after HT (Table 2, **g** - **j**). At the right motor area, spectral peaks of low β2 exhibited significantly more exaggerated and prolonged bursts in PP than in HS (*p* < 0.001, Table 2, **g** - **j**). The frontal synchronized high frequency β2 bursts appeared in PP during two subintervals, the first one with a significantly shorter latency than the low frequency β2 bursts in HS (*p* < 0.001, Table 2, **g**). In PP, synchronized bursts of high β2 were generated on the left motor area (Table 2, **g**) earlier than larger synchronized bursts at low β2 (Table 2, **i**, **j**).

right and left motor areas (Fig. 1A, 3rd row, left). The alpha frequency band in the first interval after the LT showed an enhanced central α-ERS for the PP in comparison with the HS (*p* < 0.001; Fig. 1A, left). The right and left motor areas manifested different reversal alteration as α-ERS for the PP and weakly elicited α-ERD for the HS. Following the high tone, we detected significantly different processes, α-ERS in the PP contrasted with α-ERD

Alpha-ERD differences between the groups were observed after low as well as high frequency tones during the late period (Fig. 1A, B, right). After the low tone, the HS α-ERD means were significantly more pronounced than those of the PP at centro-parietal sides, left and right motor areas (*p* < 0.001; Fig. 1A, right). The PP displayed a higher frontal α-ERD than that in the HS (*p* < 0.001). Alpha-ERD differences were observed between the groups following the high tone (Fig. 1B, right). The α-ERD signals of HS at parietal and left motor areas were more pronounced than in the PP (*p* < 0.001). The PP fronto-central and right

in the HS at all electrodes, most prominently at the left motor side (Fig. 1B, left).

motor α-ERD signals had greater respective magnitudes than those in HS (*p* < 0.001).

HT and had a prolonged latency of 176 ± 11.5 ms (Table 1, HT (T2)**, f**).

synchronized bursts at low β2 (Table 2, **i**, **j**).

During the early post-stimulus interval T1, ERD/ERS β1 patterns appeared in the lower frequency portions of this band for each electrode and group (Dushanova et al., 2010). The maximum synchronized β1 bursts across the channels were localized over fronto-central sides for both groups after LT, but had significantly higher amplitude and shorter durations in PP than HS (*p* < 0.001, bootstrap, KW test; Table 1, **a**). Later β1 ERS bursts were found only in PP over the frontal side and left motor area (Table 1, LT (T1), **b**). Synchronized β1 bursts were centered on right motor side for PP and fronto-central sides for HS, following HT (Table 1, HT (T1)**, a**, **b**). In PP, the peaks were of a significantly lower amplitude and peaked later at centro-parietal sides than in HS (Table 1, HT (T1)**, a**). During the late poststimulus interval T2, frontal synchronized β1 peaks at 20 Hz were extracted only in HS after

During T1, ERS β2 bursts were present only in PP after either tone (Table 2, LT(T1), HT(T1)). Their maximum amplitude across the scalp was localized over frontal side for either tone, right and left motor areas respectively for LT and HT (Table 2, **c**). They peaked earlier at the right than at the left motor area, fronto-central leads (Table 2, **c**), and parietal side (Table 2, **d**) for either tone, but in higher β2 frequency range after LT than after HT. During T2, the maximum value of β2 bursts across all recorded areas was centered on right motor area in PP (Table 2, **g**) but on parietal side in HS following LT (Table 2, LT(T2), **j**). More widely distributed, prolonged β2 ERS bursts were pronounced over all locations for PP following HT, but limited to frontal-central and right motor areas for HS (Table 2, HT(T2)). The maximum scalp β2 burst was localized at the right motor area in PP and frontal area in HS after HT (Table 2, **g** - **j**). At the right motor area, spectral peaks of low β2 exhibited significantly more exaggerated and prolonged bursts in PP than in HS (*p* < 0.001, Table 2, **g** - **j**). The frontal synchronized high frequency β2 bursts appeared in PP during two subintervals, the first one with a significantly shorter latency than the low frequency β2 bursts in HS (*p* < 0.001, Table 2, **g**). In PP, synchronized bursts of high β2 were generated on the left motor area (Table 2, **g**) earlier than larger

**3.2.4 Beta1** 

**3.2.5 Beta2** 


**Note:** Bold-marked mean ERS β1 bursts (%) (±SE) are maximum across the channels for each condition and subject separately for LT(T1), HT(T1), LT(T2), HT(T2); **F** (Hz) – mean frequency peaks (±SE) of the maximum ERS β1 bursts across trials; **t** (ms) – mean times (±SE) of maximum ERS β1 bursts across trials during T1 (or T2) with respect to stimulus onset; **D** (ms) – time duration of these short–term zones with ERS β1 bursts; 1 \* =a significant difference in ERS between the groups HS, PP for each channel and tone separately, marked the higher value (p<0.001, KW test); 2 \*\* =a significant difference in **t** between the groups for each channel and tone separately, marked the shorter value (p<0.001, KW test). Lower case letters in Table 1 marked consecutive sub-intervals with ERS β1 bursts: **a**, **b** during T1; **f** during T2.

Table 1. Mean ERS β1 bursts (%) across the trials with mean frequency peaks **F (**Hz) for HS and PP after LT and HT presented in short–term zones **D** (ms) during early T1 and late T2 period.

Brain Event - Related Oscillations

**3.2.6 Gamma** 

**Channels** 

**Fz ERS(%)/ F/ t**  (±SE) **D** 

–

22.7(±2)/ 47 Hz (**c**)

18.4(±1)/42.2(±2.5)Hz [36,37,47,48] Hz (**a**, **b**) 28 (±6.1)\*\* ms [8–48] ms

20 (±5.2)\*\* ms [8–32] ms

**Subjects tone(period)**

**LT (T1)** 

**HT (T1)** 

in Parkinsonian Patients During Discrimination Task Conditions 69

The scalp burst topography was localized on frontal area for PP following either tone (Table 3, LT(T1), HT(T1), **a**) and for HS – on right motor area after LT and left motor area after HT during T1 (Table 3, **c**). The ERS bursts in PP peaked later than in HS at the frontocentral for either tone, and at right and left motor areas after LT (Table 3, **a, b, c**). They were of significantly greater amplitude and more prolonged duration in PP than in HS at the fronto-central and parietal sides after either tone (*p* < 0.001, Table 3). Later burst ERS was also found during T1 over right motor and parietal areas in PP after LT (Table 3, **b; c**), and over left motor area in PP, but parietal and right motor sides in HS, after HT (Table 3, **d**). During T2, synchronized frontal bursts were extracted from both groups after either tone (Table 3, **f**) and peaked later in PP. The scalp bursts were with significantly higher

In sum, despite the early short-term β1 synchrony during the two periods, both groups exhibited mean β1 ERD following either tone type and interval (Fig. 2A, B, 1st row), which was significantly greater for HS in comparison with PP in all channels (*p* < 0.001, bootstrap, KW test), except frontal mean β1 ERS for the control group following HT during T2 (Fig. 2B, 1st row; Pz, C3', C4', *p* < 0.001; Cz, *p* < 0.05). The prolonged β2 synchronized bursts for PP during T1 had an effect on the β2 band behavior during the entire early time period (T1, Fig. 2, 2nd row, left plot). The mean β2 ERS were prominent only in PP at frontal-parietal and right motor areas after LT and at parietal and left motor areas following HT during T1 (Fig. 2A, B). The comparison of the groups also showed mean β2 ERD in HS and β2 ERS in PP during T2 following either tone in all channels except for the frontal area, which showed mean β2 ERD in both groups after LT, significantly more prominently in HS (LT, *p* < 0.05; Fig. 2A, B). The results for closely resembled the β2-frequency band behaviour. The mean ERS were more prominent than those for β2 in PP. During the sensory processing (T1), PP showed mean ERS responses at fronto-parietal and right motor areas after LT, but not after HT (Fig. 2A, B, 3rd row, left). The ERD in HS and ERS in PP were observed in all channels after either tone during the cognitive processing (T2), except frontal ERS after HT, which

amplitudes in PP than the equivalent responses from HS (*p* < 0.001, Table 3).

was more pronounced in HS than in PP (*p* < 0.001, Fig. 2B, right plot).

19.8(±2.8)/ 33 Hz (**a**)

19.4(±1.4)/36 Hz (**a**) 20 (±5.2)\*\* ms [8–32] ms

12 (±4)\*\* ms [8–16] ms

–

**Pz ERS(%)/ F/ t**  (±SE) **D HS**

20.4(±1.1)/39.4(±1.3)Hz [35–48] Hz (**a, b, c**) 86.4 (±9.9) ms [24–144] ms 

19.1(±0.5)/40.8(±0.2)Hz 72 (±8.6)\*\* ms (**b**) [24–120] ms

32.2(±4)/ 42.6(±0.4) Hz 208 (±8.6) ms [160–250] ms (**d**)

**C3' ERS(%)/ F/ t**  (±SE) **D**

–

18.9(±1.4)/41(±5.2) Hz [32,50] Hz (**a, c**) 28 (±9.5)\*\* ms [8–48] ms

**33.8(±2)/**43.8(±0.4) Hz 120 (±12.6) ms (**c**) [8–250] ms

**C4' ERS(%)/F /t**  (±SE) **D**

**30.7(±1.9)**/47.6(±0.4)Hz [46–49] Hz (**c**) 40 (±7.3)\*\* ms [8–72] ms

30.1(±2.3)\*/39.9(±0.5)Hz 44 (±7.7) ms (**b**) [8–80] ms

32.8(±3.3)/41.5(±0.2) Hz 216(±8) ms [176–250] ms (**d**)

 **Cz ERS(%)/ F/ t**  (±SE) **D**


**Note:** Lower case letters in Table 2 marked consecutive sub-intervals with ERS β2 bursts: **c**, **d**, **e** during T1 and **g**, **h**, **i**, **j** during T2.

Table 2. Mean ERS β2 bursts (%) across the trials with mean frequency peaks **F** (Hz) for HS and PP after LT and HT presented in short–term zones **D** (ms) during early T1 and late T2 period (same format as **Table 1).**

#### **3.2.6 Gamma**

68 Diagnostics and Rehabilitation of Parkinson's Disease

**C3' ERS(%)/ F/ t**  (±SE) **D**

– –

–

–

– 

19.6(±1.3)/30.5(±0.2)Hz 588 (±6.1) ms [576–600] ms (**j**)

22.4(±3.7)/29.7(±0.3)Hz 32 (±4.6) ms [24–40] ms (**c**)

**46(±2.8)**/23.9(±0.4) Hz 64 (±6.5) ms [40–88] ms (**c**) 19.3(±0.8)/32 Hz 168 (±4.6) ms [160–176] ms (**e**)

26(±1.5)/29.3(±0.5)Hz 492 (±9.5) ms [432–552] ms (**h**) 34.8(±2)/32Hz 352 (±11.8) ms [256–448] ms (**g**)

44.8(±2.5)/25Hz 532 (±10.6) ms [456–600] ms (**i**, **j**) **C4' ERS(%)/F /t**  (±SE) **D**

– –

–

28.2(±2.3)/ 32 Hz 384 (±6.5)\*\* ms [360–408] ms (**g**) 40.6(±3.1)/23.4(±0.4)Hz 532 (±10.6) ms [456–600] ms (**h**–**j**)

**60.3(±1.1)**/32Hz 12 (±4) ms [8–16] ms (**c**)

38.1(±1.2)/22Hz 56 (±4.6) ms [48–72] ms (**c**)

**30.7(±2.1)**/32 Hz 368 (±8.6) ms [320–416] ms (**g**)

**60.6(±2)**\*/22.2(±0.1)Hz 444 (±15.1) ms [280–600] ms (**g**–**j**)

–

–

–

**Pz ERS(%)/ F/ t**  (±SE) **D HS**

– –

–

–

**PP** 

–

–

**Note:** Lower case letters in Table 2 marked consecutive sub-intervals with ERS β2 bursts: **c**, **d**, **e** during

Table 2. Mean ERS β2 bursts (%) across the trials with mean frequency peaks **F** (Hz) for HS and PP after LT and HT presented in short–term zones **D** (ms) during early T1 and late T2

23.2(±1.7)/25.2(±1.6)Hz 64 (±9.2) ms [8–120] ms (**c**, **d**)

28(±2.7)/24.2(±1.1)Hz 121.6 (±9.1) ms [64–160] ms (**d**)

17.9(±1)/ 32 Hz 376 (±4.6) ms [368–384] ms (**g**) 24.7(±1.7)/30.9(±0.1)Hz 528 (±10.8) ms [448–600] ms (**h**) 25.3(±0.7)/30(±0.03)Hz 452 (±14.8) ms [296–600] ms (**g**-**j**)

**25.8(±1.6)**/32Hz 584 (±6.5) ms [560–600] ms (**j**)

 **Cz ERS(%)/F/ t**  (±SE) **D**

– –

–

–

18.1(±0.6)/21Hz 488(±5.7) ms [472–504] ms (**h**)

30.6(±4.7)/32Hz 48 (±5.7) ms [32–64] ms (**c**)

29.8(±1.1)/23Hz 64 (±4.6) ms [56–72] ms (**c**)

15.5(±0.1)/32Hz 392 (±4.6) ms [384–400] ms (**g**)

23.8(±1.3)\*/32Hz 376 (±7.3)\*\* ms [344–408] ms (**g**)

–

–

–

**Channels** 

**Fz ERS(%)/ F/ t**  (±SE) **D** 

– –

–


–

–

–

T1 and **g**, **h**, **i**, **j** during T2.

period (same format as **Table 1).**

36.7(±2.9)/ 32Hz 388(±8.9)\*\* ms [336–440] ms (**g**)

26.1(±1.0)/ 32Hz 548 (±9.5) ms [488–600] ms (**i**)

**43.9(±3.1)**/23.1(±0.1)Hz 516 (±11.5) ms [424–600] ms (**g**–**j**)

**61.1 (±1.1)**/ 32 Hz 40 (±5.7) ms [24–56] ms (**c**)

41.2(±2)/24.3(±0.5)Hz 60 (±8.8) ms [32–88] ms (**c**)

**Subjects tone(period)**

**LT (T1) HT (T1)** 

**LT (T2)** 

**HT (T2)** 

**Subjects tone(period)** 

 **LT (T1)** 

**HT (T1)** 

**LT (T2)** 

**HT (T2)** 

The scalp burst topography was localized on frontal area for PP following either tone (Table 3, LT(T1), HT(T1), **a**) and for HS – on right motor area after LT and left motor area after HT during T1 (Table 3, **c**). The ERS bursts in PP peaked later than in HS at the frontocentral for either tone, and at right and left motor areas after LT (Table 3, **a, b, c**). They were of significantly greater amplitude and more prolonged duration in PP than in HS at the fronto-central and parietal sides after either tone (*p* < 0.001, Table 3). Later burst ERS was also found during T1 over right motor and parietal areas in PP after LT (Table 3, **b; c**), and over left motor area in PP, but parietal and right motor sides in HS, after HT (Table 3, **d**). During T2, synchronized frontal bursts were extracted from both groups after either tone (Table 3, **f**) and peaked later in PP. The scalp bursts were with significantly higher amplitudes in PP than the equivalent responses from HS (*p* < 0.001, Table 3).

In sum, despite the early short-term β1 synchrony during the two periods, both groups exhibited mean β1 ERD following either tone type and interval (Fig. 2A, B, 1st row), which was significantly greater for HS in comparison with PP in all channels (*p* < 0.001, bootstrap, KW test), except frontal mean β1 ERS for the control group following HT during T2 (Fig. 2B, 1st row; Pz, C3', C4', *p* < 0.001; Cz, *p* < 0.05). The prolonged β2 synchronized bursts for PP during T1 had an effect on the β2 band behavior during the entire early time period (T1, Fig. 2, 2nd row, left plot). The mean β2 ERS were prominent only in PP at frontal-parietal and right motor areas after LT and at parietal and left motor areas following HT during T1 (Fig. 2A, B). The comparison of the groups also showed mean β2 ERD in HS and β2 ERS in PP during T2 following either tone in all channels except for the frontal area, which showed mean β2 ERD in both groups after LT, significantly more prominently in HS (LT, *p* < 0.05; Fig. 2A, B). The results for closely resembled the β2-frequency band behaviour. The mean ERS were more prominent than those for β2 in PP. During the sensory processing (T1), PP showed mean ERS responses at fronto-parietal and right motor areas after LT, but not after HT (Fig. 2A, B, 3rd row, left). The ERD in HS and ERS in PP were observed in all channels after either tone during the cognitive processing (T2), except frontal ERS after HT, which was more pronounced in HS than in PP (*p* < 0.001, Fig. 2B, right plot).


Brain Event - Related Oscillations

<0.05) or \*\* (*p* <0.001, KW test).

**4. Discussion** 

in Parkinsonian Patients During Discrimination Task Conditions 71

Fig. 2. Beta 1, Beta 2 and gamma band ERD/ERS over T1 and T2. Group means (±SE) are shown graphically to illustrate statistical results of β1-, β2-, -ERD/ERS following LT (A) and HT (B) for the early T1 (left) and late T2 (right) time periods. The significant difference in the ERD/ERS of HS and PP are presented for each pair of channels and marked by \* (*p*

The data obtained confirmed that event-related oscillatory responses in different frequency bands vary with sensory and cognitive processes. We found functional differences between event-related oscillatory activity for cognitive and sensory-motor information processing, and a clear distinction between PP and HS in both the stimuli encoding (0–250 ms) and cognitive processing (250–600 ms) intervals. Attended stimuli produced theta response synchronizations in both groups, more markedly in HS, in the first period up to 250 ms after stimulation. Enhanced theta waves in the early period (up to 250 ms) of visual and auditory stimuli have also been described by Basar (1980), Schurmann and Basar-Eroglu (1994). Theta frequency rhythms are dominant oscillations within the hippocampal formation, which is of crucial importance for the encoding of new information (Klimesch, 1997; Klimesch et al., 2005). In the late post-stimulus period functionally related to cognitive processing, θ-ERD response was predominant, excluding the parietal θ-ERS in PP in response to both

frequency tones and the frontal θ-ERS in HS in response to the low frequency tone.

Prominent differences in the α-ERD/ERS responses between the groups were observed during the first time period of 0–250 ms. A widely distributed fronto-central α-ERS


**Note:** Lower case letters in Table 3 marked consecutive sub-intervals with ERS γ bursts: **a**, **b**, **c** during T1 and **f**, **g**, **h**, **i**, **j** during T2.

Table 3. Mean ERS γ bursts (%) across the trials with mean spectral peaks **F** (Hz) for HS and PP after LT and HT presented in short–term zones **D** (ms) during early T1 and late T2 period (same format as **Table 1).** 

Fig. 2. Beta 1, Beta 2 and gamma band ERD/ERS over T1 and T2. Group means (±SE) are shown graphically to illustrate statistical results of β1-, β2-, -ERD/ERS following LT (A) and HT (B) for the early T1 (left) and late T2 (right) time periods. The significant difference in the ERD/ERS of HS and PP are presented for each pair of channels and marked by \* (*p* <0.05) or \*\* (*p* <0.001, KW test).

## **4. Discussion**

70 Diagnostics and Rehabilitation of Parkinson's Disease

–

–

–

**PP** 

–

**Note:** Lower case letters in Table 3 marked consecutive sub-intervals with ERS γ bursts: **a**, **b**, **c** during

Table 3. Mean ERS γ bursts (%) across the trials with mean spectral peaks **F** (Hz) for HS and PP after LT and HT presented in short–term zones **D** (ms) during early T1 and late T2

**36.8(±1.6)**/38.8(±0.4)Hz [35–41] Hz 464 (±14.2) ms [320–600] ms (**g**-**j**)

24.3(±1.6)/34.1(±0.3)Hz [33–35] Hz 552.6 (±9.9) ms [496–600] ms (**i**-**j**)

16.6(±1.1)/39Hz 524(±5.2) ms [512-536] ms (**h**)

32.4(±1.9)/40.7(±0.4) Hz [39–44] Hz 376 (±13.1)\*\* ms [250–496] ms (**f**) 46(±5.7)/39.3(±0.6) Hz 568 (±8) ms [528–600] ms (**j**)

32.4(±4.3)/43.2(±2.2) Hz [32,46,48,49,50]Hz (**a,d**) 56 (±8.6) ms [8–104] ms

30.1(±2.7)/37.9(±0.1) Hz

192 (±8) ms (**b**) [152–232] ms

21.1(±2)/41(±3.5)Hz [34,36,47] Hz (**a**) 20 (±5.2)\*\* ms [8–32] ms

54.3(±2.9)\*/41.6(±0.7)Hz [38–50] Hz 452 (±14.8) ms [296–600] ms (**g**-**j**) 41.6(±2.3)/42.4(±1.2)Hz [34–36, 49,50] Hz 448 (±15) ms [288–600] ms ( **g**, **h**, **i**, **j**)

–

–

–

**55(±3)**/ 44.1(±0.3) Hz [39–45] Hz 432 (±15.7) ms [250–600] ms (**f**, **g**, **i**, **j**)

28.2(±3.4)\*/37.5(±1.6)Hz [34,35,43] Hz (**a**) 36 (±6.9) ms [8–64] ms

29.5(±2.6)/34.4(±0.2) Hz 36 (±7)\*\* ms (**a**) [8–64] ms

29.9(±1.5)/33.4(±0.1) Hz 188 (±10.1) ms (**d**) [120–250] ms

**82.1(±5.2)** \*/47(±0.7) Hz [36–50] Hz 468 (±14.0) ms [328–600] ms (**g**-**j**) **55.5(±3.2)**/36.5(±0.9) Hz [33, 50] Hz 432 (±15.7) ms [250–600] ms (**f**, **g**, **i**, **j**)

–

–

–

–

53.7(±2)/ 43.2(±0.2) Hz

32.5(±2.1)\*/45.5(±1.1)Hz [42–50] Hz (**c**) 56 (±8.6)\*\* ms [8–104] ms

21.9(±0.8)/47.8(±1.4)Hz [33,49,50 ] Hz (**c**) 196 (±9.5) ms [136–250] ms 

34.3(±3)\*/ 34.1 (±0.1) Hz 112 (±11.8) ms (**a**) [16–208] ms

68.9(±5.4) \*/ 45(±0.7) Hz [35–50] Hz 432 (±15.7) ms [250–600] ms (**g**-**j**) 52(±2.1)/ 37.8(±0.9) Hz [34,35, 41,45,49,50] Hz 436 (±15.5) ms [264–600] ms (**g**, **i**, **j**)

432(±15.7) ms [250–600] ms (**f**, **j**)

21.8(±1.2)/40 (±0.5) Hz [38, 42] Hz 548 (±9.5) ms [488–600] ms (**i**-**j**)

25.6(±2.2)/45.9(±0.6) Hz [45, 49] Hz 312 (±7.3)\*\* ms [280–344] ms (**f**) 39(±3)/ 45 Hz 444 (±7.7) ms [408–488] ms (**g**) 48(±7)/ 38.6 (±0.2) Hz 572 (±7.7) ms [536–600] ms (**j**)

36.4(±4)\*/32.3(±0.7)Hz(**a**)

27.6(±2.4)\*/34.9(±0.1)Hz

48.8(±4.2)\*/44.3(±1.2)Hz [35–50] Hz 468 (±14) ms [328–600] ms (**g**-**j**) 43(±1.5)/44.9(±0.7)Hz, [47,50, 35,38,39] Hz 432 (±15.7) ms [250–600] ms (**f**, **g**, **i**, **j**)

44 (±7) ms (**a**) [16–72] ms

 36 (±6.9) ms [8–64] ms

–

 

–

–

**LT (T2)** 

16.9(±0.5)/ 38 Hz 260 (±4)\*\* ms [250–264] ms (**f**) 24.6(±2)/ 35 Hz 568 (±6.5) ms [544–592] ms (**j**) 36.3(±1.4)/43(±0.6)Hz [37,38,40,46,47] Hz 432 (±15.7) ms [250–600] ms (**f**, **g**, **j**)

**HT (T2)** 

**Subjects tone(period)** 

 **LT (T1)** 

**HT (T1)** 

**LT (T2)** 

**HT (T2)** 

–

–

–

**49(±5.5)\***/ 32 Hz (**a**) 44 (±7.7) ms [8–80] ms

**49.7(±3.5)\***/34 Hz (**a**)

39(±2.6)\*/42.7(±0.9)Hz [34–42] Hz 468 (±14.0) ms [328–600] ms (**g**-**j**) **58(±3.5)\***/40.3(±1.1)Hz [34,48,49] Hz 444 (±15.1)ms [280–600] ms (**f**, **g**, **j**)

40 (±7.3) ms [8–72] ms –

T1 and **f**, **g**, **h**, **i**, **j** during T2.

period (same format as **Table 1).** 

The data obtained confirmed that event-related oscillatory responses in different frequency bands vary with sensory and cognitive processes. We found functional differences between event-related oscillatory activity for cognitive and sensory-motor information processing, and a clear distinction between PP and HS in both the stimuli encoding (0–250 ms) and cognitive processing (250–600 ms) intervals. Attended stimuli produced theta response synchronizations in both groups, more markedly in HS, in the first period up to 250 ms after stimulation. Enhanced theta waves in the early period (up to 250 ms) of visual and auditory stimuli have also been described by Basar (1980), Schurmann and Basar-Eroglu (1994). Theta frequency rhythms are dominant oscillations within the hippocampal formation, which is of crucial importance for the encoding of new information (Klimesch, 1997; Klimesch et al., 2005). In the late post-stimulus period functionally related to cognitive processing, θ-ERD response was predominant, excluding the parietal θ-ERS in PP in response to both frequency tones and the frontal θ-ERS in HS in response to the low frequency tone.

Prominent differences in the α-ERD/ERS responses between the groups were observed during the first time period of 0–250 ms. A widely distributed fronto-central α-ERS

Brain Event - Related Oscillations

associative posterior cortex.

Lopes da Silva, 199; Engel et al., 2001).

et al., 2005).

1998).

in Parkinsonian Patients During Discrimination Task Conditions 73

motor area as θ-ERS in the HS and θ-ERD in the PP during sensory-motor processing (early period) following the high frequency tone. We also detected different processes of θ-ERS for the PP and ERD for the HS in the parietal lead during the cognitive information processing (late period) following both tones, which reflects different task-related activation of the

These findings are probably due to the auditory cortex being located in the dorsal and lateral part of the superior temporal gyrus as well as in the inferior parietal lobule (Konig et al., 2005). The absence of α-ERD at the frontal electrode locations in the patients with PD indicated that the PP, compared with HS, used different cognitive strategies for stimulus response processing which are normally implemented by fronto-striatal circuits (Krause, 2006). The late higher fronto-central α-ERD in PP accompanied by a lower P3 component amplitude, especially in the fronto-central sides, reflects a disturbance in the frontal regulation of attentional processes as well a disturbance of the basal ganglia activity and their related thalamo-cortical neuronal nets (Stam et al., 1993; Piccirilli et al., 1989; Schmiedt

In PP, we found hemispheric lateralization for sensory and cognitive processing concerning θ-ERD/ERS at left and right motor areas as well as a significantly higher α-ERS at left compared to right motor area. This finding corresponds with the results of Magnani et al., 1998, Defebvre et al., 1996. These authors suggested that other cortical areas may be activated both to compensate for a dysfunction of motor preparation and to increase the level of cortical activity necessary for the realization of the movement. Another possible explanation is that this hemispheric lateralization is connected with auditory attention and hemispheric differences in the processing of high and low frequencies (Ivry & Robertson,

Post-stimulus β1 ERD was elicited from both groups during sensory (T1) and cognitive information (T2) processing, though this was significantly more pronounced in HS in response to both tone types at all electrodes. The greater β1 ERD in HS can be explained by the increased excitability level of the neurons (Pfurtscheller & Lopes da Silva, 1999; Brown & Marsden, 1998). Late post-stimulus frontal β1 ERS (T2) was evident only in HS following HT. This HS ERS, comprising components in the band between 13 and 20 Hz, may represent an inhibited frontal cortical network, at least under certain circumstances (Pfurtscheller &

A frontal β2 ERD was maintained in both groups during the cognitive information processing (T2) following LT, though this was weaker in PP. β2 ERS was only observed in PP. These were weakly elicited during the sensory stimuli processing (T1) and appeared at fronto-parietal and left motor areas (LT: Fz, C3', Pz; HT: C3', Pz). β2 ERS in PP was more prominent during cognitive processing (T2) after either tone type, but particularly so following HT. The β2 change reversals compared to β1 which we observed for the PD patients support the hypothesis of Marceglia et al. (2009), that two distinct information channels in the cortico-basal ganglia–thalamo-cortical loop, involved in motor and nonmotor information processing, are formed in the parkinsonian brain. The frontal β synchronization at 20–30 Hz arises both from communication with, and also from within, the STN (Williams et al., 2003). The β synchrony has been ascribed predominantly to a lack of dopaminergic activity in the striatum which, together with the STN, is the recipient of cortical input to the basal ganglia (Fogelson et al., 2006; Williams et al., 2002). Studies with unmedicated PD patients have revealed prominent oscillations in 'basal ganglia β frequency band' (Weinberger et al., 2006; Kühn et al., 2006; Priori et al., 2004; Fogelson et al., 2006). The

manifested in PP in response to both tone types during the first 250 ms after stimulus was absent in HS. In the second period after stimulus absence, α-ERD was found in both groups in response to both tone types. This was generally more prominent in HS, with the exception of frontal side in response to either a high or low frequency tone. Alpha-ERD was more prominent at central and right motor areas in response to the high frequency tone in PP.

Theta and alpha frequency ERD/ERS were significantly different between subject groups. It is known that the oscillatory alterations to θ-ERS are related to memory encoding (Klimesch et al., 2001; Jensen & Tesche, 2002).

Alpha-ERS most probably demonstrates active working memory or attentional processes (Klimesch, 1997; Jensen et al., 2002), whereas α-ERD is functionally related to mental activity (Basar, 1980) and reflects memory search processes (Klimesch, 1997; Klimesch et al., 2005; Pesonen et al., 2006). The recognition of auditory stimuli elicits widespread α-ERD responses (Krause et al., 1994). It is accepted that alpha oscillations are mainly generated by cortico-cortical and thalamo-cortical neuronal networks (Lopes da Silva et al., 1980; Schmiedt et al., 2005; Ellfolk et al., 2006). This fact, together with the changes in the metabolic patterns of thalamic, premotor and prefrontal cortex, parieto-occipital regions, etc., that occur in PP (Fukuda et al., 2001) could explain the abnormality of early time period α-ERS in the PP compared to the HS. Observed slight activity of the basal ganglia–thalamic and cerebellar–thalamic pathways might be implicated in the development of parkinsonian symptoms (Rolland et al., 2007).

Schmiedt et al. (2005) also found differences between PP and HS in the θ- and α-frequency ERD/ERS responses during working memory encoding but in a visual working memory paradigm. We cannot draw direct parallels between their results and ours in the present study because of the different stimulus modality. The early and late δ post-stimulus activities were enhanced in HS. The late period, related to cognitive information processing, exhibited δ-ERS in HS and δ-ERD in PP at most electrodes in response to a low frequency tone, and at parietal and left motor areas in response to the high frequency tone. Many authors agree that the main power of P300 is in the delta range (Demiralp et al., 1999; Karakas et al., 2000; Klimesch et al., 2000; Klimesch et al., 2006). The lower δ-ERS in PP in the late post-stimulus period, which becomes δ-ERD in some recordings, could explain the lower P3 amplitude observed in PP (Philipova et al., 1997). Our patients were medicated by L-dopa drugs and this medication may have had some effect on the present findings. One of the models (Leblois et al., 2006) supposed that high dopamine depletion could modify the network dynamic state from an imbalance between the feedbacks and lead to synchronous oscillations driven by a hyperdirect loop appearing in basal ganglia after inactivation of the striatum.

A reduction in this α-ERD/ERS abnormality and a consequent improvement in PP performance during working memory tasks have been found as the result of L-dopa (Lewis et al., 2003; Marini et al., 2003; Shohamy et al., 2005; Devos et al., 2004). Nevertheless, we found some differences between the two groups. The memory related and stimulus categorized ERD/ERS responses at all these frequencies reflected different underlying neuropathological and cognitive changes in this neurodegenerative disease. Theta activity is suggested to be mostly engaged in memory operations (Klimesch et al., 1996; Karakas et al., 2000; Jensen et al., 2002b) and this θ pathological synchronized enhancement in PD could explain the cognitive dysfunction commonly occurring even in the early stages of Parkinson's disease (Lewis et al., 2003). We found specific significant differences at left

manifested in PP in response to both tone types during the first 250 ms after stimulus was absent in HS. In the second period after stimulus absence, α-ERD was found in both groups in response to both tone types. This was generally more prominent in HS, with the exception of frontal side in response to either a high or low frequency tone. Alpha-ERD was more prominent at central and right motor areas in response to the high frequency tone in

Theta and alpha frequency ERD/ERS were significantly different between subject groups. It is known that the oscillatory alterations to θ-ERS are related to memory encoding (Klimesch

Alpha-ERS most probably demonstrates active working memory or attentional processes (Klimesch, 1997; Jensen et al., 2002), whereas α-ERD is functionally related to mental activity (Basar, 1980) and reflects memory search processes (Klimesch, 1997; Klimesch et al., 2005; Pesonen et al., 2006). The recognition of auditory stimuli elicits widespread α-ERD responses (Krause et al., 1994). It is accepted that alpha oscillations are mainly generated by cortico-cortical and thalamo-cortical neuronal networks (Lopes da Silva et al., 1980; Schmiedt et al., 2005; Ellfolk et al., 2006). This fact, together with the changes in the metabolic patterns of thalamic, premotor and prefrontal cortex, parieto-occipital regions, etc., that occur in PP (Fukuda et al., 2001) could explain the abnormality of early time period α-ERS in the PP compared to the HS. Observed slight activity of the basal ganglia–thalamic and cerebellar–thalamic pathways might be implicated in the development of parkinsonian

Schmiedt et al. (2005) also found differences between PP and HS in the θ- and α-frequency ERD/ERS responses during working memory encoding but in a visual working memory paradigm. We cannot draw direct parallels between their results and ours in the present study because of the different stimulus modality. The early and late δ post-stimulus activities were enhanced in HS. The late period, related to cognitive information processing, exhibited δ-ERS in HS and δ-ERD in PP at most electrodes in response to a low frequency tone, and at parietal and left motor areas in response to the high frequency tone. Many authors agree that the main power of P300 is in the delta range (Demiralp et al., 1999; Karakas et al., 2000; Klimesch et al., 2000; Klimesch et al., 2006). The lower δ-ERS in PP in the late post-stimulus period, which becomes δ-ERD in some recordings, could explain the lower P3 amplitude observed in PP (Philipova et al., 1997). Our patients were medicated by L-dopa drugs and this medication may have had some effect on the present findings. One of the models (Leblois et al., 2006) supposed that high dopamine depletion could modify the network dynamic state from an imbalance between the feedbacks and lead to synchronous oscillations driven by a hyperdirect loop appearing in basal ganglia after inactivation of the

A reduction in this α-ERD/ERS abnormality and a consequent improvement in PP performance during working memory tasks have been found as the result of L-dopa (Lewis et al., 2003; Marini et al., 2003; Shohamy et al., 2005; Devos et al., 2004). Nevertheless, we found some differences between the two groups. The memory related and stimulus categorized ERD/ERS responses at all these frequencies reflected different underlying neuropathological and cognitive changes in this neurodegenerative disease. Theta activity is suggested to be mostly engaged in memory operations (Klimesch et al., 1996; Karakas et al., 2000; Jensen et al., 2002b) and this θ pathological synchronized enhancement in PD could explain the cognitive dysfunction commonly occurring even in the early stages of Parkinson's disease (Lewis et al., 2003). We found specific significant differences at left

PP.

striatum.

et al., 2001; Jensen & Tesche, 2002).

symptoms (Rolland et al., 2007).

motor area as θ-ERS in the HS and θ-ERD in the PP during sensory-motor processing (early period) following the high frequency tone. We also detected different processes of θ-ERS for the PP and ERD for the HS in the parietal lead during the cognitive information processing (late period) following both tones, which reflects different task-related activation of the associative posterior cortex.

These findings are probably due to the auditory cortex being located in the dorsal and lateral part of the superior temporal gyrus as well as in the inferior parietal lobule (Konig et al., 2005). The absence of α-ERD at the frontal electrode locations in the patients with PD indicated that the PP, compared with HS, used different cognitive strategies for stimulus response processing which are normally implemented by fronto-striatal circuits (Krause, 2006). The late higher fronto-central α-ERD in PP accompanied by a lower P3 component amplitude, especially in the fronto-central sides, reflects a disturbance in the frontal regulation of attentional processes as well a disturbance of the basal ganglia activity and their related thalamo-cortical neuronal nets (Stam et al., 1993; Piccirilli et al., 1989; Schmiedt et al., 2005).

In PP, we found hemispheric lateralization for sensory and cognitive processing concerning θ-ERD/ERS at left and right motor areas as well as a significantly higher α-ERS at left compared to right motor area. This finding corresponds with the results of Magnani et al., 1998, Defebvre et al., 1996. These authors suggested that other cortical areas may be activated both to compensate for a dysfunction of motor preparation and to increase the level of cortical activity necessary for the realization of the movement. Another possible explanation is that this hemispheric lateralization is connected with auditory attention and hemispheric differences in the processing of high and low frequencies (Ivry & Robertson, 1998).

Post-stimulus β1 ERD was elicited from both groups during sensory (T1) and cognitive information (T2) processing, though this was significantly more pronounced in HS in response to both tone types at all electrodes. The greater β1 ERD in HS can be explained by the increased excitability level of the neurons (Pfurtscheller & Lopes da Silva, 1999; Brown & Marsden, 1998). Late post-stimulus frontal β1 ERS (T2) was evident only in HS following HT. This HS ERS, comprising components in the band between 13 and 20 Hz, may represent an inhibited frontal cortical network, at least under certain circumstances (Pfurtscheller & Lopes da Silva, 199; Engel et al., 2001).

A frontal β2 ERD was maintained in both groups during the cognitive information processing (T2) following LT, though this was weaker in PP. β2 ERS was only observed in PP. These were weakly elicited during the sensory stimuli processing (T1) and appeared at fronto-parietal and left motor areas (LT: Fz, C3', Pz; HT: C3', Pz). β2 ERS in PP was more prominent during cognitive processing (T2) after either tone type, but particularly so following HT. The β2 change reversals compared to β1 which we observed for the PD patients support the hypothesis of Marceglia et al. (2009), that two distinct information channels in the cortico-basal ganglia–thalamo-cortical loop, involved in motor and nonmotor information processing, are formed in the parkinsonian brain. The frontal β synchronization at 20–30 Hz arises both from communication with, and also from within, the STN (Williams et al., 2003). The β synchrony has been ascribed predominantly to a lack of dopaminergic activity in the striatum which, together with the STN, is the recipient of cortical input to the basal ganglia (Fogelson et al., 2006; Williams et al., 2002). Studies with unmedicated PD patients have revealed prominent oscillations in 'basal ganglia β frequency band' (Weinberger et al., 2006; Kühn et al., 2006; Priori et al., 2004; Fogelson et al., 2006). The

Brain Event - Related Oscillations

are not fixed, and may vary with task demands.

specific content of short-term memory in each group.

ms (Pantev, 1995; Arnott et al., 2004).

task demands.

in Parkinsonian Patients During Discrimination Task Conditions 75

anticipatory activations both for LT and HT, one might assume that the same cortical networks should underlie the same stimulus representations. However, while all components were mainly localized over fronto-central areas, there was some variation between the conditions, showing significant effects on the parietal components for LT, but not for parietal activity for HT. This suggests that networks encoding the stimulus features

The assessing EEG stimulus-specific oscillatory activity yielded insights into the temporal dynamics of sound processing in short-term memory. Contrasting oscillatory activity between the two stimuli, such as between LT ERS and HT ERD during the sensory processing (T1) in PP, as well as between LT ERD and HT ERS during the cognitive processing (T2) in HS, revealed stimulus-specific activity behavior in the 30–50 Hz range over the HS's frontal and PP's fronto-parietal cortex. This suggests that band activity reflects the general involvement of cortical networks in particular tasks but may index the

The pronounced and well-synchronized burst in HS was present in the very short-term phases around 25–60 ms after the stimulus onset, with spectral peaks ranging from 30 to 45 Hz (Gurtubay et al., 2001). Despite this short-term high synchrony in HS, the common behavior during sensory stimuli processing (T1) was desynchronization. However, the PP processes were with higher short-term energy, which is a prerequisite for all maintenancerelated processes, and thus defined a persistent synchronization during sensory stimuli processing, mainly at fronto-parietal and right motor areas following LT. It is clear that there is a difference between groups in the early and well-synchronized response that is basically a sensory phenomenon important to preparing the brain for the subsequent processing. This evidence suggests that oscillations may be modulated by attentional processes. Several cognitive paradigms for the auditory system have shown early spectral peak responses in the band between 30 and 40 Hz at around 25 ms after stimulus onset that last for about 100

 Later peaking activity has been recorded in the 200–400 ms interval following an experimental task. The latency and scalp topography vary according to the type of stimulus, indicating task-dependent local network activation. The significantly varying magnitudes of differentiation demonstrated that the topography of stimulus-specific -band activity is also task-dependent within the groups (Tiitinen et al., 1993). We found restricted energy changes over all recorded areas in HS, but not in the frontal area after HT. The significant differences between groups in T2 recorded after both stimuli could be due to memory retrieval processes that are activated during the performance of the paradigms. The lower energy in HS during cognitive processing (T2) could be related to fewer attentional processes required to eventually perform a task. The relative strength of differentiation in the band may suggest that performance depends on the different group's ability to retain a representation in memory of the relevant stimulus feature and thus to be able to neglect the irrelevant stimuli*.* The acquisition and retention of sound frequency information was accompanied by frontal gamma band activity components (Karakaş & Başar, 1998). The high-frequency stimuli were accompanied by more exaggerated, well–synchronized frontal band components in performing the tasks. Memory for low versus high frequency tones selectively enhanced oscillatory activity for the posterior versus the frontal components, thus directly demonstrating differentiation of the group's modulation of cortical activity by

engagement of the basal ganglia in β band synchronization is found when there is acute or chronic dopaminergic hypoactivity, and while primarily associated with bradykinesia and rigidity, it has also been associated with impairments to complex movements and motor related cognitive behaviour because of the widespread basal ganglia connectivity with the cerebral cortex (Terman et al., 2002). Further, the pathological β synchrony in the cerebellum might lead to a purer breakdown of simple motor tasks because of more focal cerebellothalamic projections into the cerebral cortex that are concentrated on the primary motor cortex (Leblois et al., 2007). A relative functional division between activities in the β band might be supported by the evidence for different patterns of pharmacological sensitivity (Priori et al., 2004) and cortico-subthalamic coupling (Fogelson et al., 2006). The dopaminergic drug treatment suppressed mainly β1 synchrony, graded by the amount of drug-induced suppression in the STN (Kühn et al., 2006; Wang et al., 2005) and cerebral cortex, correlating with the level of improvement in bradykinesia and rigidity but not in parkinsonian rest tremor (Weinberger et al., 2006; Silberstein et al., 2005), the latter of which probably has an independent pathophysiological substrate (Rivlin-Etzion et al., 2006).

Our group of patients showed a significantly reduced ERD compared with HS over central and left motor areas, and only PP showed ERS over fronto-parietal and right motor areas following LT during the sensory stimuli processing (T1). A widespread ERS appeared during later cognitive processing (T2), and then only in PP, following either tone type, with the exception of a more prominent frontal ERS in HS following HT. In our study, we observed switches between cortical activity in the β2 and band oscillations. Hence we concluded that a reduction in β2-band synchronized activity allows higher frequency oscillatory activity in the range leading to its synchronization. The observed energy changes in the β2 and bands indicate that an increase in one is accompanied by a decrease in the other. These T2 changes in PP were more pronounced in the motor cortex than in the parietal and even frontal cortex data. In the parkinsonian state, there was a tendency towards increased synchronized higher frequency fluctuations, specifically in the motor cortex, where instances of peaks were found after both tone types. Except for the β2 band series of data during cognitive processing (T2) after HT, the difference between magnitude of the peaks in the frontal, parietal and contralateral motor areas did not reach significance. Recent data demonstrate that the disruptions of the beta and gamma range cortical rhythms are based on the disturbed temporal relationship between cortical oscillatory activity and basal ganglia activity in Parkinsonism (Gatev & Wichmann, 2008). This finding is also in agreement with studies of PP following dopaminergic medication, which promoted synchronized oscillatory activity at higher frequencies () predominantly at the level of the frontal cortex and striatum (Levy et al., 2001; 2002; Brown et al., 2001; Williams et al., 2002; Leblois et al., 2007).

In recent MEG investigations of various cognitive and sensory tasks (Kaiser et al., 2003; Lutzenberger et al., 2002) the reported band activity over the higher sensory areas has not shown a sustained activation, but rather, a peaking activity. In our sensori-motor study, these transient responses were functionally dissociable between the two groups. We observed stimulus-specific band activity components over the fronto-parietal cortex, but this was differently manifest in each group and varied over the time course. The topography was compatible with the notion of an auditory dorsal space processing stream involving the posterior temporal, parietal and superior frontal cortex (Rauschecker, 1998; Arnott et al., 2004). If the cognitive processing (T2) band activity components represent similar

engagement of the basal ganglia in β band synchronization is found when there is acute or chronic dopaminergic hypoactivity, and while primarily associated with bradykinesia and rigidity, it has also been associated with impairments to complex movements and motor related cognitive behaviour because of the widespread basal ganglia connectivity with the cerebral cortex (Terman et al., 2002). Further, the pathological β synchrony in the cerebellum might lead to a purer breakdown of simple motor tasks because of more focal cerebellothalamic projections into the cerebral cortex that are concentrated on the primary motor cortex (Leblois et al., 2007). A relative functional division between activities in the β band might be supported by the evidence for different patterns of pharmacological sensitivity (Priori et al., 2004) and cortico-subthalamic coupling (Fogelson et al., 2006). The dopaminergic drug treatment suppressed mainly β1 synchrony, graded by the amount of drug-induced suppression in the STN (Kühn et al., 2006; Wang et al., 2005) and cerebral cortex, correlating with the level of improvement in bradykinesia and rigidity but not in parkinsonian rest tremor (Weinberger et al., 2006; Silberstein et al., 2005), the latter of which probably has an independent pathophysiological substrate (Rivlin-Etzion et al., 2006). Our group of patients showed a significantly reduced ERD compared with HS over central and left motor areas, and only PP showed ERS over fronto-parietal and right motor areas following LT during the sensory stimuli processing (T1). A widespread ERS appeared during later cognitive processing (T2), and then only in PP, following either tone type, with the exception of a more prominent frontal ERS in HS following HT. In our study, we observed switches between cortical activity in the β2 and band oscillations. Hence we concluded that a reduction in β2-band synchronized activity allows higher frequency oscillatory activity in the range leading to its synchronization. The observed energy changes in the β2 and bands indicate that an increase in one is accompanied by a decrease in the other. These T2 changes in PP were more pronounced in the motor cortex than in the parietal and even frontal cortex data. In the parkinsonian state, there was a tendency towards increased synchronized higher frequency fluctuations, specifically in the motor cortex, where instances of peaks were found after both tone types. Except for the β2 band series of data during cognitive processing (T2) after HT, the difference between magnitude of the peaks in the frontal, parietal and contralateral motor areas did not reach significance. Recent data demonstrate that the disruptions of the beta and gamma range cortical rhythms are based on the disturbed temporal relationship between cortical oscillatory activity and basal ganglia activity in Parkinsonism (Gatev & Wichmann, 2008). This finding is also in agreement with studies of PP following dopaminergic medication, which promoted synchronized oscillatory activity at higher frequencies () predominantly at the level of the frontal cortex and striatum (Levy et al., 2001; 2002; Brown et al., 2001;

In recent MEG investigations of various cognitive and sensory tasks (Kaiser et al., 2003; Lutzenberger et al., 2002) the reported band activity over the higher sensory areas has not shown a sustained activation, but rather, a peaking activity. In our sensori-motor study, these transient responses were functionally dissociable between the two groups. We observed stimulus-specific band activity components over the fronto-parietal cortex, but this was differently manifest in each group and varied over the time course. The topography was compatible with the notion of an auditory dorsal space processing stream involving the posterior temporal, parietal and superior frontal cortex (Rauschecker, 1998; Arnott et al., 2004). If the cognitive processing (T2) band activity components represent similar

Williams et al., 2002; Leblois et al., 2007).

anticipatory activations both for LT and HT, one might assume that the same cortical networks should underlie the same stimulus representations. However, while all components were mainly localized over fronto-central areas, there was some variation between the conditions, showing significant effects on the parietal components for LT, but not for parietal activity for HT. This suggests that networks encoding the stimulus features are not fixed, and may vary with task demands.

The assessing EEG stimulus-specific oscillatory activity yielded insights into the temporal dynamics of sound processing in short-term memory. Contrasting oscillatory activity between the two stimuli, such as between LT ERS and HT ERD during the sensory processing (T1) in PP, as well as between LT ERD and HT ERS during the cognitive processing (T2) in HS, revealed stimulus-specific activity behavior in the 30–50 Hz range over the HS's frontal and PP's fronto-parietal cortex. This suggests that band activity reflects the general involvement of cortical networks in particular tasks but may index the specific content of short-term memory in each group.

The pronounced and well-synchronized burst in HS was present in the very short-term phases around 25–60 ms after the stimulus onset, with spectral peaks ranging from 30 to 45 Hz (Gurtubay et al., 2001). Despite this short-term high synchrony in HS, the common behavior during sensory stimuli processing (T1) was desynchronization. However, the PP processes were with higher short-term energy, which is a prerequisite for all maintenancerelated processes, and thus defined a persistent synchronization during sensory stimuli processing, mainly at fronto-parietal and right motor areas following LT. It is clear that there is a difference between groups in the early and well-synchronized response that is basically a sensory phenomenon important to preparing the brain for the subsequent processing. This evidence suggests that oscillations may be modulated by attentional processes. Several cognitive paradigms for the auditory system have shown early spectral peak responses in the band between 30 and 40 Hz at around 25 ms after stimulus onset that last for about 100 ms (Pantev, 1995; Arnott et al., 2004).

 Later peaking activity has been recorded in the 200–400 ms interval following an experimental task. The latency and scalp topography vary according to the type of stimulus, indicating task-dependent local network activation. The significantly varying magnitudes of differentiation demonstrated that the topography of stimulus-specific -band activity is also task-dependent within the groups (Tiitinen et al., 1993). We found restricted energy changes over all recorded areas in HS, but not in the frontal area after HT. The significant differences between groups in T2 recorded after both stimuli could be due to memory retrieval processes that are activated during the performance of the paradigms. The lower energy in HS during cognitive processing (T2) could be related to fewer attentional processes required to eventually perform a task. The relative strength of differentiation in the band may suggest that performance depends on the different group's ability to retain a representation in memory of the relevant stimulus feature and thus to be able to neglect the irrelevant stimuli*.* The acquisition and retention of sound frequency information was accompanied by frontal gamma band activity components (Karakaş & Başar, 1998). The high-frequency stimuli were accompanied by more exaggerated, well–synchronized frontal band components in performing the tasks. Memory for low versus high frequency tones selectively enhanced oscillatory activity for the posterior versus the frontal components, thus directly demonstrating differentiation of the group's modulation of cortical activity by task demands.

Brain Event - Related Oscillations

Elsevier, Amsterdam

ISSN 0270-6474

ISSN 0885-3185

ISSN 0149-7634

pp. 241-248, ISSN 0167-8760

Jun 13), pp. 1801-1804, ISSN 0140-6736

125(Pt 6), (2002 Jun), pp. 1235-1246, ISSN 0006-8950

*Cortex* 31(3), (1995 Sep), pp. 413–432, ISSN 0010-9452

Jan), pp. 108–128, ISSN 0093-934X

408–419, ISSN 0006-8950

ISSN 0018-9294

in Parkinsonian Patients During Discrimination Task Conditions 77

Basar, E. (1980) EEG brain dynamics. Relations between EEG and brain evoked potentials.

Başar, E., Basar-Eroglu, C., Karakas, S., Schurmann, M. (2001) Gamma, alpha, delta, and

Brown, P. & Marsden, C. D. (1998) What do the basal ganglia do? Lancet 351(9118), (1998

Brown, P., Oliviero, A., Mazzone, P., Insola, A., Tonali, P. & Di Lazzaro, V. (2001)

Brown, P. (2003) Oscillatory nature of human basal ganglia activity: relationship to the

Cassidy, M., Mazzone, P., Oliviero, A., Insola, A., Tonali, P., Di Lazzaro, V., Brown, P. (2002)

Cools R (2006) Dopaminergic modulation of cognitive functionimplications for L-DOPA

Gurtubay, I. G., Alegre, M., Labarga, A., Malanda, A., Iriarte, J., Artieda, J. (2001) Gamma

Daum, I., Schugens, M. M., Spieker, S., Poser, U., Schönle, P. W., Birbaumer, N. (1995)

Defebvre, L, Bouniez, JL, Destke, A, Guieu, JD (1996) Movement related ERD pattern

Demiralp, T., Ademoglu, A., Schürmann, M., Bas¸ar-Eroglu, C., Başar, E. (1999) Detection of

Devos, D., Labyt, E., Derambure, P., Bourriez, J. L., Cassim, F., Reyns, N., Blond, S., Guieu, J.

Donoghue, J. P., Sanes, J. N., Hatsopoulos, N. G., Gaal, G. (1998) Neural discharge and local

Durka, P. J., Zygierewicz, J., Klekowicz, H., Ginter, J., Blinowska, K. J. (2004) On the

Dushanova, J., Philipova, D., Nikolova., G. (2009) Event-related synchronization/

*Neurophysiol* 79(1), (1998 Jan), pp. 159-173, ISSN 0022-3077

*Clin Neurophysiol* 112(7), (2001 Jul), pp. 1219-1228, ISSN 1388-2457

*Neurosurg Psychiatry* 60(3), (1996 Mar), pp. 307–312, ISSN 0022-3050

theta oscillations govern cognitive processes. *Int J Psychophysiol* 39(2-3), (2001 Jan)

Dopamine dependency of oscillations between subthalamic nucleus and pallidum in Parkinson's disease. *J Neurosci* 21 (3), (2001 Feb 1) pp. 1033-1038,

pathophysiology of Parkinson's disease. *Mov Disord* 18(4), (2003 Apr), pp. 357-363,

Movement-related changes in synchronization in the human basal ganglia. Brain

treatment in Parkinson's disease. *Neurosci Biobehav Rev* 30(1), (2005 Jun 1), pp. 1–23,

band activity in an auditory oddball paradigm studied with the wavelet transform.

Memory and skill acquisition in Parkinson's disease and frontal lobe dysfunction.

preceding voluntary movement in untreated Parkinson's disease. *J Neurol* 

P300 waves in single trials by the wavelet transform (WT). *Brain Lang* 66(1), (1999

D., Destée, A., Defebvre, L. (2004) Subthalamic nucleus stimulation modulates motor cortex oscillatory activity in Parkinson's disease. *Brain* 127(2), (2004 Feb), pp.

field potential oscillations in primate motor cortex during voluntary movements. *J* 

statistical significance of event-related EEG desynchronization and synchronization in the time-frequency plane. *IEEE Trans Biomed Eng* 51(7), (2004 Jul), pp. 1167–1175,

desynchronization during discrimination task conditions in patients with

A mechanism that underlies many of the immediately aforementioned cognitive functions is the match of sensory information with memory contents (Kaiser et al., 2009a; 2009b; Visscher et al., 2007). The 'early' -band activity occurring 150 ms after stimulus presentation reflects such a match with memory. The 'late' activity, which typically emerges with a latency of more than 150 ms, is a temporal signature of utilization processes such as response selection or context updating. We also found a later (250–400 ms) ERS response, following only HT in HS, over the frontal location, where this activity peaked in the 33–45 Hz range. In PP, the specific β2 and bursts (30–38 Hz) exhibited maximal scalp projection covering areas to the left and right of the motor areas**,** and with frontal, central or parietal participation that depended on the stimuli. This oscillatory burst reflects a later stimulus context process, although it has also been associated with the motor responses later in the task (Brown, 2003; Kaiser et al., 2009b). The β2/ oscillation in the groups points to a direct relation to aspects of post-discrimination processes related to the P300 wave **(**Haig et al., 2000). This oscillatory burst (letters **g-j** for intervals from 320 to 550 ms) also showed a variable relationship to attention, as it was significantly different during the HT and LT task. The results also showed that EEG activity in the frontal, parietal and motor cortex is significantly different between groups, not only in temporal variations (always with a delay in PP) but also in frequency shifts (β2/ ERD in HS compared to the ERS in PP). Although these shifts do not follow a simple pattern, they are significantly different from HS, raising the possibility that the interactions between basal ganglia activity and cortical rhythms are functionally relevant. Therefore, the normal higher frequency relationships between cortical and basal ganglia activity are strongly altered in the parkinsonian state (Gatev & Wichmann, 2008). The shifts of β/ patterns occurring in the groups are probably associated with specific types of basal ganglia events related to transitions between cortical idling and more active states (Williams et al., 2003).

#### **5. Conclusion**

Our investigation further demonstrates the close relationship between physiological abnormalities in PD and disturbances in the EEG frequency characteristics. The results of this investigation in PD patients of both sensory and cognitive processing of auditory stimuli suggests that PD should be characterized by multiple impairments in oscillatory networks, which in turn indicates the presence of task-specific disturbances in the temporal and regional integration of all frequency components.

### **6. References**


A mechanism that underlies many of the immediately aforementioned cognitive functions is the match of sensory information with memory contents (Kaiser et al., 2009a; 2009b; Visscher et al., 2007). The 'early' -band activity occurring 150 ms after stimulus presentation reflects such a match with memory. The 'late' activity, which typically emerges with a latency of more than 150 ms, is a temporal signature of utilization processes such as response selection or context updating. We also found a later (250–400 ms) ERS response, following only HT in HS, over the frontal location, where this activity peaked in the 33–45 Hz range. In PP, the specific β2 and bursts (30–38 Hz) exhibited maximal scalp projection covering areas to the left and right of the motor areas**,** and with frontal, central or parietal participation that depended on the stimuli. This oscillatory burst reflects a later stimulus context process, although it has also been associated with the motor responses later in the task (Brown, 2003; Kaiser et al., 2009b). The β2/ oscillation in the groups points to a direct relation to aspects of post-discrimination processes related to the P300 wave **(**Haig et al., 2000). This oscillatory burst (letters **g-j** for intervals from 320 to 550 ms) also showed a variable relationship to attention, as it was significantly different during the HT and LT task. The results also showed that EEG activity in the frontal, parietal and motor cortex is significantly different between groups, not only in temporal variations (always with a delay in PP) but also in frequency shifts (β2/ ERD in HS compared to the ERS in PP). Although these shifts do not follow a simple pattern, they are significantly different from HS, raising the possibility that the interactions between basal ganglia activity and cortical rhythms are functionally relevant. Therefore, the normal higher frequency relationships between cortical and basal ganglia activity are strongly altered in the parkinsonian state (Gatev & Wichmann, 2008). The shifts of β/ patterns occurring in the groups are probably associated with specific types of basal ganglia events related to transitions between cortical idling and more

Our investigation further demonstrates the close relationship between physiological abnormalities in PD and disturbances in the EEG frequency characteristics. The results of this investigation in PD patients of both sensory and cognitive processing of auditory stimuli suggests that PD should be characterized by multiple impairments in oscillatory networks, which in turn indicates the presence of task-specific disturbances in the temporal

Anik, Y., Iseri, P., Demirci, A., Komsuoglu, S., Inan, N. (2007) Magnetization transfer ratio in

Antal, A., Kéri, S., Dibó, G., Benedek, G., Janka, Z., Vécsei, L., Bodis-Wollner, I. (2002)

Arnott, S. R., Binns, M. A., Grady, C. L., Alain, C. (2004) Assessing the auditory

early period of Parkinson disease. *Acad Radiol* 14(2), (2007 Feb), pp. 189–192, ISSN

Electrophysiological correlates of visual categorization, pp. evidence for cognitive dysfunctions in early Parkinson's disease. *Brain Res Cogn Brain Res* 13(2), (2002 Apr)

dualpathway model in humans. *NeuroImage* 22(1), (2004 May), pp. 401-408, ISSN

active states (Williams et al., 2003).

and regional integration of all frequency components.

,pp. 153–158, ISSN 0926-6410

**5. Conclusion** 

**6. References** 

1076-6332

1053-8119


Brain Event - Related Oscillations

ISSN 0027-8424

0953-816X

ISSN 1053-8119

1732, ISSN 1388-2457

ISSN 0304-3940

5), pp. 18–23, ISSN 0304-3940

Jun), pp. 319–340, ISSN 0167-8760

May 17), pp. 1235–1240, ISSN 0959-4965

ISBN 0-262-09034-1

in Parkinsonian Patients During Discrimination Task Conditions 79

Ivry, R. B., Robertson, L. C. (1998) The two sides of perception. MIT Press, Cambridge, MA,

Jensen, O., Tesche, C. D. (2002) Frontal theta activity in humans increases with memory load

Jensen, O., Gelfand, J., Kounios, J., Lisman, J. E. (2002) Oscillations in the alpha band (9–12

Kaiser, J., Ripper, B., Birbaumer, N., Lutzenberger, W. (2003) Dynamics of gamma-band

Kaiser, J., Rahm, B., Lutzenberger, W. (2009a) Temporal dynamics of stimulus-specific

Kaiser, J., Lutzenberge, W., Decker, C., Wibral, M., Rahm, B. (2009b) Task- and performance-

Karakaş ,S., & Başar, E. (1998) Early gamma response is sensory in origin: a conclusion based

Karakas, S., Erzengin, O. U., & Başar, E. (2000) A new strategy involving multiple cognitive

Karrasch, M., Krause, C. M., Laine, M., Lang, H. A., Lehto, M. (1998) Eventrelated ERD and

Karrasch, M., Laine, M., Rapinoja, P., Krause, C. M. (2004) Effects of normal ageing on event-

Klimesch, W. (1997) EEG a-rhythms and memory processes. *Int J Psychophysiol* 26(1-3), (1997

Klimesch, W., Doppelmay, M., Russegger, H., Pachinger, T. (1996) Theta band power in the

Klimesch, W., Doppelmayr, M., Schwaiger, J., Winkler, T., Gruber W. (2000) Theta

Klimesch, W., Doppelmayr, M., Stadler, W., Pollhuber, D., Sauseng, P., Rohm, D. (2001)

*Neurophysiol* 111(5), (2000 May), pp. 781–793, ISSN 1388-2457

memory. *NeuroImage* 20(2), (2003 Oct), pp. 816-827, ISSN 1053-8119

*Cereb Cortex* 12(8), (2002 Aug), pp. 877– 882, ISSN 1047-3211

*Psychophysiol* 31(1), (1998 Dec), pp. 13-31, ISSN 0167-8760

44(1), (2009 Jan 1), pp. 257-264, ISSN 1053-8119

107(2), (1998 Aug), pp. 112–121, ISSN 0013-4694

Parkinson disease. *Proc Natl Acad Sci USA* 90(24), (1993 Dec 15), pp. 11965–11969,

in a working memory task. *Eur J Neurosci* 15(8), (2002 Apr), pp. 1395–1399, ISSN

Hz) increase with memory load during retention in a short-term memory task.

activity in human magnetoencephalogram during auditory pattern working

gamma-band activity components during auditory short-term memory. *Neuroimage*

related modulation of domain-specific auditory short-term memory representations in the gamma band. *Neuroimage* 46(4), (2009 Jul 15), pp. 1127-36,

on cross-comparison of results from multiple experimental paradigms. *Int J* 

paradigms demonstrates that ERP components are determined by the superposition of oscillatory signals. *Clin Neurophysiol* 111(10), (2000 Oct), pp. 1719–

ERS during an auditory lexical matching task. *Electroencephalogr Clin Neurophysiol*

related ERD/ERS during a memory task in humans. *Neurosci Lett* 366(1), (2004 Aug

human scalp EEG and the encoding of new information. *Neuroreport* 7(7), (1996

oscillations and the ERP old/new effect: independent phenomena? *Clin* 

Episodic retrieval is reflected by a process specific increase in human electroencephalographic h activity. *Neurosci Lett* 302(1), (2001 Apr 13), pp. 49–52,

Parkinson's disease. *Cell Mol Neurobiol* 29(6-7), (2009 Sep), pp. 971-980, ISSN 0272- 4340


Dushanova, J., Philipova, D., Nikolova, G. (2010) Beta and gamma frequency-range

*the Neurological Sciences* 293(1-2), (2010 Jun 15), pp. 51–58, ISSN 0022-510X Ellfolk, U., Karrasch, M., Laune, M., Pesonen, M., Krause, K. (2006) Event related

Engel, A. K., Fries, P., & Singer, W. (2001) Dynamic predictions: oscillations and synchrony

Evarts, E. V., Teräväinen, H., & Calne, D. B. (1981) Reaction time in Parkinson's disease.

Farmer, S. F. (1998) Rhythmicity, synchronization and binding in human and primate motor

Fries, P., Reynolds, J. H., Rorie, A. E., Desimone, R. (2001) Modulation of oscillatory

Foffani, G., Bianchi ,A. M., Baselli, G., Priori, A. (2005) Movement-related frequency

Fogelson, N., Williams, D., Tijssen, M., van Bruggen, G., Speelman, H., Brown, P. (2006)

Haig, A. R., Gordon, E., Pascalis, V. D., Meares, R. A., Bahramali, H., Harris, A. (2000)

Herrmann, C. S., Munk, M. H. J., Engel, A. K. (2004) Cognitive functions of gamma-band

Higginson, C. I., King, D. S., Levine, D., Wheelock , V. L., Khamphay, N. O., Sigvardt, K. A.

Innis, R. B., Seibyl, J. P., Scanley, B. E., Laruelle, M., Abi-Dargham, A., Wallace, E., Baldwin,

systems. *J Physiol* 509(1), (1998 May 15), pp. 3-14, ISSN 0022-3751

*Brain* 104(1), (1981 Mar), pp. 167–186, ISSN 0006-8950

568(2), (2005 Oct 15), pp 699–711, ISSN 0022-3751

*Cereb Cortex* 19(6), (2009 Jun), pp. 1330-44, ISSN 1047-3211

17(5), (1967 May), pp. 427-442, ISSN 0028-3878

*Neurophysiol* 111(8), (2000 Aug), pp. 1461–1468, ISSN 1388-2457

23), pp. 1560-1563, ISSN 0036-8075

355, ISSN 1364-6613

4340

2457

003X

Parkinson's disease. *Cell Mol Neurobiol* 29(6-7), (2009 Sep), pp. 971-980, ISSN 0272-

abnormalities in parkinsonian patients under cognitive sensorimotor task. *Journal of* 

desynchronisation/ERS during an auditory-verbal memory task in mild Parkinson's disease. *Clin Neurophysiol* 117(8), (2006 Aug), pp. 1737–1745, ISSN 1388-

in top-down processing. *Nat Rev Neurosci* 2(10), (2001 Oct), pp. 704-716, ISSN 1471-

neuronal synchronization by selective visual attention. *Science* 291(5508), (2001 Feb

modulation of beta oscillatory activity in the human subthalamic nucleus, *J Physiol* 

Different functional loops between cerebral cortex and the subthalmic area in Parkinson's disease. *Cereb Cortex* 16(1), (2006 Jan), pp. 64-75, ISSN 1047-3211 Fukuda, M., Edward, C., Eidelberg, D. (2001) Functional brain networks in Parkinson's disease. *Parkinsonism Relat Disord* 8(2), (2001 Oct), pp. 91–94, ISSN 1353-8020 Gatev, P., & Wichmann, T. (2008) Interactions between Cortical Rhythms and Spiking

Activity of Single Basal Ganglia Neurons in the Normal and Parkinsonian State.

Gamma activity in schizophrenia: evidence of impaired network binding? *Clin* 

activity: memory match and utilization. *Trends Cogn Sci* 8(8), (2004 Aug), pp. 347-

(2003) The relationship between executive function and verbal memory in Parkinson's disease. *Brain Cogn* 52(3), (2003 Aug), pp. 343–352, ISSN 0278-2626 Hoehn, M. M., Yahr, M. D. (1967) Parkinsonism: onset progression and mortality. *Neurology*

R. M., Zea-Ponce, Y., Zoghbi, S., Wang, S. (1993) Single photon emission computed tomographic imaging demonstrates loss of striatal dopamine transporters in Parkinson disease. *Proc Natl Acad Sci USA* 90(24), (1993 Dec 15), pp. 11965–11969, ISSN 0027-8424


Brain Event - Related Oscillations

0885-3185

0167-8760

in Parkinsonian Patients During Discrimination Task Conditions 81

Lutzenberger, W., Ripper, B., Busse, L., Birbaumer, N., Kaiser, J. (2002) Dynamics of gamma-

Magnani, G., Cursi, M., Leocani, L., Volonté, M. A., Locatelli, T., Elia, A., Comi, G. (1998)

Marsden, J. F., Limousin-Dowsey, P., Ashby, P., Pollak, P. & Brown, P. (2001) Subthalamic

Mason, M., & Newton, M. (1990) A rank statistics approach to the consistency of the general

Marceglia, S., Fiorio, M., Foffani, G., Mrakic-Sposta, S., Tiriticco, M., Locatelli, M., Caputo,

Minamoto, H., Tachibana, H., Sugita, M., Okita, T. (2001) Recognition memory in normal

Owen, A. M. (2004) Cognitive dysfunction in Parkinson's disease: the role of frontostriatal circuitry. *Neuroscientist* 10(6), (2004 Dec), pp. 525–537, ISSN 1073-8584 Ozgoren, M., Basar-Eroglu, C., Basar, E. (2005) Beta oscillations in face recognition. *Int J* 

Pantev, C. (1995) Evoked and induced gamma-band activity of the human cortex. *Brain* 

Pesonen, M., Björnberg, C. H., Hämäläinen, H., Krause, C. M. (2006) Brain oscillatory 1–30

search task. *Neurosci Lett* 399(1–2), (2006 May 15), pp. 45–50, ISSN 0304-3940 Philipova, D., Gatchev, G., Vladova, T., Georgiev, D (1997) Event related potentials under

Piccirilli M., D'Alessandro P., Finali G., Piccinin G. L., Agostini L. (1989) Frontal lobe

Pillon, B., Ertle, S., Deweer, B., Bonnet, A., Vidailhet, M., Dubois, B. (1997) Memory for

Priori, A., Foffani, G., Pesenti, A., Bianchi, A., Chiesa, V., Baselli, G. (2002) Movement-

Hz EEG ERD/ERS responses during the different stages of an auditory memory

auditory discrimination tasks. *Int J Psychophysiol* 27(1), (1997 Jul), pp. 69–78, ISSN

dysfunction in Parkinson's disease: prognostic value for dementia? *Eur Neurol* 

spatial location in 'de novo' parkinsonian patients. *Neuropsychologia* 35(3), (1997

related modulation of neural activity in human basal ganglia and its l-DOPA dependency: recordings from deep brain stimulation electrodes in patients with

*Brain Res Cogn Brain Res* 11(1), (2001 Mar), pp. 23–32, ISSN 0926-6410 Oldfield, R. C. (1971) The assessment and analysis of handedness: the Edinburgh inventory.

*Neuropsychologia* 9(1), (1971 Mar), pp. 97-113, ISSN 0028-3932

*Psychophysiol* 55(1), (2005 Jan), pp. 51-59, ISSN 0167-8760

*Topogr* 7(4), (1995 Summer), pp. 321-330, ISSN 0896-0267

50(5-6), (1980 Dec), pp. 449–456, ISSN 0013-4694

22(13), (2002 Jul 1), pp. 5630-5638, ISSN 0270-6474

*Brain* 124(2), (2001 Feb), pp. 378-388, ISSN 0006-8950

21), pp. 1027-1036, ISSN 0306-4522

29(2), pp. 71–76, ISSN 0014-3022

Mar), pp. 221–228, ISSN 0028-3932

bootstrap, *Ann. Statist* 20, pp. 1611–1624, ISSN 0090-5364

rhythms, revealed by partial coherence analysis. *Electroencephalogr Clin Neurophysiol*

band activity during an audiospatial working memory task in humans. J Neurosci

Event-related ERD to contingent negative variation and self-paced movement paradigms in Parkinson's disease. *Mov Disord* 13(4), (1998 Jul), pp. 653–660, ISSN

nucleus, sensorimotor cortex and muscle interrelationships on Parkinson's disease.

E., Tinazzi M., Priori, A. (2009) Modulation of beta oscillations in the subthalamic area during action observation in Parkinson's disease. *Neuroscience* 161(4), (2009 Jul

ageing and Parkinson's disease: behavioural and electrophysiological measures.


Klimesch, W., Schnack, B., Sauseng, P. (2005) The functional significance of h and upper a-

Klimesch, W., Hanslmayr, S., Sauseng, P., Gruber, W., Brozinsky, C. J., Kroll, N. E.,

Krause, C. M. (2006) Cognition- and memory-related ERD/ERS responses in the auditory

Krause, C. M., Lang, A. H., Laine, M., Kuusisto, M., Pörn, B. (1996) Eventrelated EEG

Knoke, D., Taylor, A. E., Saint-Cyr, J. A. (1998) The differential defects of ceing on recall in

Kühn, A. A., Kupsch, A., Schneider, G. H., Brown, P. (2006) Reduction in subthalamic 8–35

Leblois, A., Boraud, T., Meissner, W., Bergman, H., Hansel, D. (2006) Competition between

Leblois, A., Meissner, W., Bioulac, B., Gross, C. E., Hansel, D., Boraud, T. (2007) Late

Levy, R., Dostrovsky, J. O., Lang, A. E., Sime, E., Hutchison, W. D., Lozano, A. M. (2001)

Levy, R., Hutchison, W. D., Lozano, A. M., Dostrovsky, J. O. (2002) Synchronized neuronal

activity. *J Neuroscience* 22(7), (2002 Apr 1), pp. 2855-2861, ISSN 0270-6474 Lopes da Silva, F. H., Vos, J. E., Mooibroek, J., Van Rotterdam, A. (1980) Relative

stimulus modality. *Prog Brain Res* 159, pp. 197–207, ISSN 0079-6123 Krause, C. M., Lang, H. A., Laine, M., Helle, S. I., Kuusisto, M. J., Pörn, B. (1994) Event-

*Neurophysiol* 110(8), (1999 Aug), pp. 1371– 1377, ISSN 1388-2457

*Neurosci* 26(13), (2006 Mar 29), pp. 3567– 3583, ISSN 0270-6474

23(7), (2006 Apr), pp. 1956-1960, ISSN 0953-816X

Winter), pp. 107–112, ISSN 0896-0267

trace decay. *Cereb Cortex* 16(2), (2006 Feb), pp. 280–290, ISSN 1047-3211 Konig, R., Heil, P., Budinger, E., Scheich, H. (2005) The auditory cortex: a synthesis of

Yonelinas, A. P., Doppelmayr, M. (2006) Oscillatory EEG correlates of episodic

human and animal research. Lawrence Erlbaum Associates, NJ. ISBN 0-8058-

related desynchronization evoked by auditory stimuli. *Brain Topogr* 7(2), (1994

desynchronization and synchronization during an auditory memory task. *Electroencephalogr Clin Neurophysiol* 98(4), (1996 Apr), pp. 319–326, ISSN 0013-4694 Krause, C. M., Aström, T., Karrasch, M., Laine, M., Sillanmäki, L. (1999) Cortical activation

related to auditory semantic matching of concrete versus abstract words. *Clin* 

Parkinson's Disease and normal subjects. *Brain Cogn* 38(2), (1998 Nov), pp. 261–274,

Hz oscillatory activity correlat with clinical improvement in PD. *Eur J Neurosci*

feedback loops underlies normal and pathological dynamics in the basal ganglia. *J* 

emergence of synchronized oscillatory activity in the pallidum during progressive parkinsonism. *Eur J Neurosci* 26(6), (2007 Sep), pp. 1701-1713, ISSN 0953-816X Lewis, S. J., Dove, A., Robbins, T. W., Barker, R. A., Owen, A. M. (2003) Cognitive

impairments in early Parkinson's disease are accompanied by reductions in activity in frontostriatal neural circuitry. *J Neurosci* 23(15), (2003 Jul 16), pp. 6351–6356, ISSN

Effects of apomorphine on subtalamic nucleus and globus pallidus internus neurons in patients with Parkinson's disease. *J Neurophysiology* 86(1), (2001 Jul), pp.

discharge in the basal ganglia of parkinsonian patients is limited to oscillatory

contributions of intracortical and thalamo-cortical processes in the generation of a-

oscillations. *Exp Psychol* 52(2), pp. 99–108, ISSN 1618-3169

4938-6

ISSN 0278-2626

0270-6474

249-260, ISSN 0022-3077

rhythms, revealed by partial coherence analysis. *Electroencephalogr Clin Neurophysiol* 50(5-6), (1980 Dec), pp. 449–456, ISSN 0013-4694


Brain Event - Related Oscillations

134–142, ISSN 0022-510X

211–232, ISSN 0278-2626

2963-2976, ISSN 0270-6474

0167-9228

9173

0022-3077

in Parkinsonian Patients During Discrimination Task Conditions 83

Tachibana, H., Toda, K., Sugita, M. (1992) Actively and passively evoked P3 latency of

Taylor, A. E., Saint-Cyr, J. A., Lang, A. E. (1990) Memory and learning in early Parkinson's

Tallon-Baudry, C., Bertrand, O., Delpuech, C., Pernier, J. (1996) Stimulus specificity of

Terman, D., Rubin, J. E., Yew, A. C., Wilson, C. J. (2002) Activity patterns in a model for the

Tiitinen, H., Sinkkonen, J., Reinikainen, K., Alho, K., Lavikainen, J., Näätänen, R. (1993)

van Spaendonck, K. P., Berger, H. J., Horstink, M. W., Cools, A. R. (1998) Cognitive deficits

Vieregge, P., Verleger, R., Wascher, E., Stüven, F., Kömpf, D. (1994) Auditory selective

Wang, H., Wang, Y., Wang, D., Cui L., Tian, S., Zhang, Y. (2002) Cognitive impairment in

Wang, S. Y., Aziz, T. Z., Stein, J. F. & Liu, X. (2005) Time-frequency analysis of transient

Wascher, E., Verleger, R., Vieregge, P., Jaskowski ,P., Koch, S., Kömpf, D. (1997) Responses

Weinberger, M., Mahant, N., Hutchison, W. D., Lozano, A. M., Moro, E., Hodaie, M., Lang,

Weinberger, M., Hutchison, W. D., Lozano, A. M., Hodaie, M., Dostrovsky, J. O. (2008)

Williams, D., Kühn, A., Kupsch, A., Tijssen, M., van Bruggen, G., Speelman, H., Hotton, G.,

16(13), (1996 Jul 1), pp. 4240-4249, ISSN 0270-6474

*Nature* 364(6432), (1993 Jul 1), pp. 59-60, ISSN 0028-0836

(2002 Feb 15), pp. 49–53, ISSN 0022-510X

145(1-2), (2005 Jun 30), pp. 151-158, ISSN 0165-0270

96(6), (2006 Dec), pp. 3248-3256, ISSN 0022-3077

event-related potentials in Parkinson's disease. *J Neurol Sci* 111(2), (1992 Sep), pp.

disease: evidence for a ''frontal lobe syndrome''. *Brain Cogn* 13(2), (1990 Jul), pp.

phase-locked and non-phase-locked 40 Hz visual responses in human. *J Neurosci*

subthalamopallidal network of the basal ganglia. *J Neurosci* 22(7), (2002 Apr 1), pp.

Selective attention enhances the auditory 40-Hz transient response in humans.

in Parkinson's disease. *Tijdschr Gerontol Geriatr* 29(4), (1998 Aug), pp. 189-195, ISSN

attention is impaired in Parkinson's disease event-related evidence from EEG potentials. *Brain Res Cogn Brain Res* 2(2), (1994 Sep), pp. 117–129, ISSN 0926-6410 Visscher, K. M., Kaplan, E., Kahana, M. J., Sekuler, R. (2007) Auditory short-term memory

behaves like visual short-term memory. *PLoS Biol* 5(3), (2007 Mar) e56, ISSN 1544-

Parkinson's disease revealed by event-related potential N270. *J Neurol Sci* 194(1),

neuromuscular events: dynamic changes in activity of the subthalamic nucleus and forearm muscles related to the intermittent resting tremor. *J Neurosci Methods*

to cued signals in Parkinson's disease. Distinguishing between disorders of cognition and of activation. *Brain* 120(8), (1997 Aug), pp. 1355–1375, ISSN 0006-8950

A E., Dostrovsky, J. O. (2006) Beta oscillatory activity in the subthalamic nucleus and its relation to dopaminergic response in Parkinson's disease. *J Neurophysiol*

Increased gamma oscillatory activity in the subthalamic nucleus during tremor in Parkinson's disease patients. *J Neurophysiol* 101(2), (2009 Feb), pp. 789-802, ISSN

Yarrow, K. & Brown, P. (2003) Behavioural cues are associated with modulations of

Parkinson's disease. *Neurol Sci* 23(Suppl. 2), (2002 Sep), pp. S101-S102, ISSN 1590- 1874


Priori ,A., Foffani, G., Pesenti, A., Tamma, F., Bianchi, A. M., Pellegrini, M., Locatelli, M.,

Pfurtscheller, G., & Klimesch, W. (1991) Event-related ERD during motor behaviour and

Pfurtscheller, G., & Lopes da Silva, F. H. (1999) Event-related desynchronization Handbook

Rauschecker J. P. (1998) Cortical processing of complex sounds. *Curr Opin Neurobiol* 8(4),

Rivlin-Etzion, M., Marmor, O., Heimer, G., Raz, A., Nini, A., Bergman, H. (2006) Basal

Robertson, C., Empson, J. (1999) Slowed cognitive processing and high workload in Parkinson's disease. *J Neurol Sci* 162(1), (1999 Jan 1), pp. 27–33, ISSN 0022-510X Rolland, A.-S., Herrero, M.-T., Garcia-Martinez, V., Ruberg, M., Hirsch, E. C., Francois, C.

slowing. *J Neurosci* 22(12), (2002 Jun 15), pp. 5198–5203, ISSN 0270-6474 Schmiedt, C., Meistrowitz, A., Schwendemann, G., Herrmann, M., Basar-Eroglu, C. (2005)

Schurmann, M., Basar-Eroglu, E. (1994) Topography of alpha and theta oscillatory responses

Shohamy, D., Myers, C. E., Grossman, S., Sage, J., Gluck, M. A. (2005) The role of dopamine

Silberstein, P., Pogosyan, A., Kühn, A. A., Hotton, G., Tisch, S., Kupsch, A., Dowsey-

Stam, C. J., Visser, S. L., Op de Coul, A. A., De Sonneville, L. M., Schellens, R. L., Brunia, C.

Parkinson's disease. *Brain* 116(5), (1993 Oct), pp. 1139–1158, ISSN 0006-8950 Sutoh T., Yabe H., Sato Y., Hiruma T., Kaneko S. (2000) Event-related desynchronization

1874

379, ISSN 0014-4886

65, ISSN 0424-8155

ISBN 0-444-82999-7

0340-1200

0006-8950

ISSN 1388-2457

(1998 Aug), pp. 516-521, ISSN 0959-4388

*Neurobiol* 16(6), (2006 Dec), pp. 629-637, ISSN 0959-4388

388(3), (2005 Nov 18), pp. 138–143, ISSN 0304-3940

156(2), (2005 Jan 30), pp. 191–199, ISSN 0166-4328

Parkinson's disease. *Neurol Sci* 23(Suppl. 2), (2002 Sep), pp. S101-S102, ISSN 1590-

Moxon, K. A., Villani, R. M. (2004) Rhythm-specific pharmacological modulation of subthalamic activity in Parkinson's disease. *Exp Neurol* 189(2), (2004 Oct), pp. 369-

visual information processing. *Electroencephalogr Clin Neurophysiol Suppl* 42, pp. 58–

of Electroencephalography and Clinical Neurophysiulogy. Revised Series, Vol. 6,

ganglia oscillations and pathophysiology of movement disorders. *Curr Opin* 

(2007) Metabolic activity of cerebellar and basal ganglia-thalamic neurons is reduced in parkinsonism. *Brain* 130(1), (2007 Jan), pp. 265–275, ISSN 0006-8950 Sawamoto, N., Honda, M., Hanakawa, T., Fukuyama, H., Shibasaki, H. (2002) Cognitive

slowing in Parkinson's disease: a behavioural evaluation independent of motor

Theta and alpha oscillations reflect differences in memory strategy and visual discrimination performance in patients with Parkinson's disease. *Neurosci Lett*

upon auditory and visual stimuli in humans. *Biol Cybern* 72(2), pp. 161–174, ISSN

in cognitive sequence learning: evidence from Parkinson's disease. *Behav Brain Res*

Limousin, P., Hariz, M. I., Brown, P. (2005) Cortico-cortical coupling in Parkinson's disease and its modulation by therapy. *Brain* 128(6), (2005 Jun), pp. 1277-1291, ISSN

H., de Smet. J. S., Gielen, G. (1993) Disturbed frontal regulation of attention in

during an auditory oddball task. *Clin Neurophysiol* 111(5), (2000 May), pp. 858-862,


**4** 

*Taiwan* 

**Extraction of Single-Trial Post-Movement** 

**MEG Beta Synchronization in Normal and** 

*3Center for Dynamical Biomarkers and Translational Medicine, National Central University,* 

The human brain is a dynamic system that frequently changes functional mode (Lopes da Silva, 1991; Lopes da Silva, 1996). Spatiotemporal analysis of brain activities with regard to distinct spatial locations and frequency bands reveals task-specific brain activation which changes in a fraction of a second (Jensen & Vanni, 2002). At rest, Rolandic EEG and MEG rhythms are dominated by rhythmic activity around 10 (alpha band) and 20 (beta band) Hz. Electrocorticographic (Pfurtscheller et al., 1994) and neuromagnetic recordings have shown that the ~20-Hz rhythm mainly originates in the anterior bank of the central sulcus while the ~10-Hz rhythm is concentrated predominantly in the post-central cortex (Pfurtscheller & Lopes da Silva, 1999). These two frequency components appear to have different functional roles, with the ~20-Hz rhythm being more closely connected to movements and their termination and the ~10-Hz component behaving more like a classical "idling" rhythm (Salmelin et al., 1995). Voluntary movement is composed of three phases: planning, execution and recovery (Pfurtscheller et al., 1998a). It has been suggested that localized event-related alpha desynchronization (ERD) upon movement can be viewed as an EEG/MEG correlate of an activated cortical sensorimotor network, servicing planning and execution, while beta event-related synchronization (ERS) may reflect deactivation/inhibition during the recovery

Movement-related ERD and ERS have been used as probes to study neurophysiology in normal brains and pathophysiology in the diseased (Tamas et al., 2003). It has been reported that the diagnostic features of patients with Parkinson's disease, in comparison with controls, are a slowing and suppression of the post-movement beta ERS independent of the amount of beta activity in the reference period (Pfurtscheller et al., 1998a). These findings imply that slowed and reduced recovery after the motor act impedes cortical preparation of the next movement (Pfurtscheller et al., 1996). Patients with Unverricht-Lundborg type myoclonic epilepsy demonstrate little rebound of beta activities contingent upon median nerve stimulation (Silen et al., 2000). The diminished beta ERS indicates that the myoclonic

phase in the underlying cortical network (Pfurtscheller et al., 1996).

**1. Introduction** 

**Parkinson's Patient Using ICA-Based** 

*1Department of Electrical Engineering, National Central University,* 

**Spatiotemporal Approach** 

Po-Lei Lee1,2,3, Yu-Te Wu2 and Jen-Chuen Hsieh2

*2Institute of Brain Science, National Yang-Ming University,* 

synchronous oscillations in the human subthalamic nucleus. *Brain* 126(9), (2003 Sep), pp. 1975-1985, ISSN 0006-8950

Williams, D., Tijssen, M., van Bruggen, G., Bosch, A., Insola, A., Di Lazzaro, V., Mazzone, P., Oliviero, A., Quartarone, A., Speelman, H., & Brown, P (2002) Dopamine dependent changes in the functional connectivity between basal ganglia and cerebral cortex in the human. *Brain* 125(7), (2002 Jul), pp. 1558-1569, ISSN 0006-8950

## **Extraction of Single-Trial Post-Movement MEG Beta Synchronization in Normal and Parkinson's Patient Using ICA-Based Spatiotemporal Approach**

Po-Lei Lee1,2,3, Yu-Te Wu2 and Jen-Chuen Hsieh2

*1Department of Electrical Engineering, National Central University, 2Institute of Brain Science, National Yang-Ming University, 3Center for Dynamical Biomarkers and Translational Medicine, National Central University, Taiwan* 

## **1. Introduction**

84 Diagnostics and Rehabilitation of Parkinson's Disease

Williams, D., Tijssen, M., van Bruggen, G., Bosch, A., Insola, A., Di Lazzaro, V., Mazzone, P.,

Sep), pp. 1975-1985, ISSN 0006-8950

synchronous oscillations in the human subthalamic nucleus. *Brain* 126(9), (2003

Oliviero, A., Quartarone, A., Speelman, H., & Brown, P (2002) Dopamine dependent changes in the functional connectivity between basal ganglia and cerebral cortex in the human. *Brain* 125(7), (2002 Jul), pp. 1558-1569, ISSN 0006-8950

> The human brain is a dynamic system that frequently changes functional mode (Lopes da Silva, 1991; Lopes da Silva, 1996). Spatiotemporal analysis of brain activities with regard to distinct spatial locations and frequency bands reveals task-specific brain activation which changes in a fraction of a second (Jensen & Vanni, 2002). At rest, Rolandic EEG and MEG rhythms are dominated by rhythmic activity around 10 (alpha band) and 20 (beta band) Hz. Electrocorticographic (Pfurtscheller et al., 1994) and neuromagnetic recordings have shown that the ~20-Hz rhythm mainly originates in the anterior bank of the central sulcus while the ~10-Hz rhythm is concentrated predominantly in the post-central cortex (Pfurtscheller & Lopes da Silva, 1999). These two frequency components appear to have different functional roles, with the ~20-Hz rhythm being more closely connected to movements and their termination and the ~10-Hz component behaving more like a classical "idling" rhythm (Salmelin et al., 1995). Voluntary movement is composed of three phases: planning, execution and recovery (Pfurtscheller et al., 1998a). It has been suggested that localized event-related alpha desynchronization (ERD) upon movement can be viewed as an EEG/MEG correlate of an activated cortical sensorimotor network, servicing planning and execution, while beta event-related synchronization (ERS) may reflect deactivation/inhibition during the recovery phase in the underlying cortical network (Pfurtscheller et al., 1996).

> Movement-related ERD and ERS have been used as probes to study neurophysiology in normal brains and pathophysiology in the diseased (Tamas et al., 2003). It has been reported that the diagnostic features of patients with Parkinson's disease, in comparison with controls, are a slowing and suppression of the post-movement beta ERS independent of the amount of beta activity in the reference period (Pfurtscheller et al., 1998a). These findings imply that slowed and reduced recovery after the motor act impedes cortical preparation of the next movement (Pfurtscheller et al., 1996). Patients with Unverricht-Lundborg type myoclonic epilepsy demonstrate little rebound of beta activities contingent upon median nerve stimulation (Silen et al., 2000). The diminished beta ERS indicates that the myoclonic

Extraction of Single-Trial Post-Movement MEG Beta

Synchronization in Normal and Parkinson's Patient Using ICA-Based Spatiotemporal Approach 87

are categorized into task-related and task-unrelated groups respectively, based on temporal and spatial characteristics. This temporal template is the grand average of hundreds of vector-norm envelopes of the band-pass filtered, single-trial MEG measurements obtained from right index finger lifting. The spatial template can be derived from the spatial distribution at beta rebound activity either from the grand average of the generation group (for signal extraction) or from each individual (for verification). Correlations between the temporal template and component waveforms, as well as between the spatial template and spatial maps, are computed, and coupled component waveforms and spatial maps that conjointly survive with high correlation values are taken as task-related information and subjected to data reconstruction. In this way the phase and amplitude information of noisefree MEG beta activities can be preserved for profound studies of temporal and spectral variation across trials. Due to the high signal-to-noise ratio (SNR) in beta activities extracted through ICA, trial-specific reactive frequency ranges can be determined by means of the comparisons of two short time spectra between the reference and post-movement periods. Beta reactivity per single trial can be quantified using the amplitude modulation (AM) method (Clochon et al., 1996), and insignificant epochs can be determined using a nonparametric sign test (Brovelli et al., 2002). Source estimation and localization techniques can be successfully

applied to single-trial epoch to estimate the source locations of beta modulation.

and/or loss of synchronization in Parkinson's disease.

**2. Materials and methods** 

**2.1 Subjects and task** 

The current study presents: 1) a novel ICA-based spatiotemporal approach for single-trial analysis of event-related beta oscillatory modulations with a high extraction rate; 2) the prospect of trial-specific frequency bandpass filtering that takes into account subtle trial-bytrial brain dynamics; 3) the feasibility of using sophisticated source estimation/localization methods demanding high signal-to-noise ratio (SNR) on single trial data; and 4) a common template approach permitting an effective alternative in cases where lengthy procedures cannot be endured by participants or in clinical settings where patients have attention problems or are incapable of sustaining long experiments. The proposed ICA-based approach was applied to discover the mechanisms of beta ERS in one Parkinson's patient. It is helpful to investigate the reasoning of ERS vanishment due to suppression of postmovement beta rebound in each single-trial, rather than the cause of temporal jittering

The present study examined six healthy right-handed subjects (gender balanced), aged 24-30 years. Five of the healthy subjects were used in the model generation group, and MEG data from the last healthy subject were used for validation. Subjects performed self-paced lifting of the right index finger approximately once every 8 sec. Subjects were trained to perform the movement briskly for a duration of 200 to 300 ms, as monitored by surface electromyogram (EMG) on extensor digitorum communis, with a range of finger movement around 35~40, while keeping their eyes open in order to suppress the occipital alpha rhythm. In addition, somatosensory evoked fields (SEFs) for right median nerve stimulation were measured to locate the primary sensorimotor area (SMI) in each subject as part of the procedure for the generation of a temporal template (see below). Informed written consent was obtained from all subjects. This study was approved by the Institutional Review Board of Taipei Veterans General Hospital. In addition, one 56-year-old patient with idiopathic Parkinson's disease in

Hoehn and Yahr stage 1 was also recruited as a demonstration in this study.

patients have sustained motor cortex reactivity which can be attributed to impaired cortical inhibition (Pfurtscheller & Lopes da Silva, 1999).

ERD and ERS activities are time-locked, but not phase-locked, to external stimuli or tasks (Andrew & Pfurtscheller, 1995; Kalcher & Pfurtscheller, 1995; Pfurtscheller & Lopes da Silva, 1999). Existing methods for extraction of ERD/ERS signals essentially measure power or amplitude changes of corresponding frequency bands as derived from the average of dozens or hundreds of trials. The band power method squares and averages filtered brain signals within a selected frequency band (Pfurtscheller & Aranibar, 1977), and an inter-trial variance method to remove the phase-locked portion in the band power method was reported by Klimesch et al. (1998). Likewise, autoregressive and spectral decomposition methods have been used to extract significant frequency components in rhythmic signals (Florian & Pfurtscheller, 1995). Salmelin's temporal-spectral evolution method rectifies and averages filtered MEG signals (Salmelin et al., 1995). To increase the temporal resolution of the ERD/ERS technique, Clochon et al. (1996) proposed an amplitude modulation (AM) method based on the Hilbert transform to detect the envelope of filtered signals by squaring and summing their real and imaginary parts. All these approaches presume stereotypical frequency and temporal characteristics across trials and require an average of many trials for the ERD/ERS using a preset frequency filter and time window to preprocess every trial. However, non-phase-locked rhythmic signals can vary from trial-to-trial contingent upon variations in a subject's performance and state, which may be linked to fluctuations in expectation, attention, arousal, and task strategy (Bastiaansen et al., 2001; Bastiaansen et al., 1999; Earle, 1988; Haig et al., 1995; Hoffman et al., 1991; Yabe et al., 1993). Since trial-to-trial variability in amplitude, latencies, or scalp distribution might carry important information on cognitive and physiological states (Jung et al., 2001), a method that permits the extraction and analysis of the oscillatory signal on a single-trial base is crucial for the study of subtle brain dynamics. Furthermore, such a method should require fewer trials for analysis and hence shorter experiment time, which is beneficial for patients with impairment of motor and/or cognitive performance (Muller-Gerking et al., 1999).

Single-trial multi-channel EEG analysis has been developed for time-locked, phase-locked, evoked brain activities (Jung et al., 2001; Tang et al., 2002). However, approaches to singletrial movement-related oscillatory changes are less explored. Independent component analysis (ICA), a data-driven method for multivariate data analysis, has been used to reveal temporally-independent neuronal activities of EEG measurements (Jung et al., 2001; Makeig et al, 1997; McKeown et al., 1998), MEG measurements (Wu et al., 2002; Wu et al., 2003; Tang et al., 2002), fMRI (Duann et al., 2002; McKeown et al., 1998) and recently perfusion MRI (Kao et al., 2003). The present study proposes a new approach using ICA and the Hilbert transformation for the single-trial detection of movement-related beta rhythmic activity during a self-paced right finger lifting task. This study focuses on beta activity and beta ERS, centered around 20 Hz, because it has been demonstrated that the movement-related short bursts of beta oscillation have higher task and movement specificity than alpha ERD (Pfurtscheller & Aranibar, 1979b; Pfurtscheller et al., 1996).

Since brain oscillation may be expressed alone in a specific frequency band independent of artifacts (Ermer et al., 2000; Lins et al., 1993a; Lins et al., 1993b; Mosher et al., 1992), ICA is applied to transform brain signals across all channels (in a single trial) into mutually independent components by means of an unmixing matrix in which each column represents a spatial map tailoring the weights of the corresponding temporal component at each MEG sensor. The spatial maps and temporal waveforms of decomposed independent components are categorized into task-related and task-unrelated groups respectively, based on temporal and spatial characteristics. This temporal template is the grand average of hundreds of vector-norm envelopes of the band-pass filtered, single-trial MEG measurements obtained from right index finger lifting. The spatial template can be derived from the spatial distribution at beta rebound activity either from the grand average of the generation group (for signal extraction) or from each individual (for verification). Correlations between the temporal template and component waveforms, as well as between the spatial template and spatial maps, are computed, and coupled component waveforms and spatial maps that conjointly survive with high correlation values are taken as task-related information and subjected to data reconstruction. In this way the phase and amplitude information of noisefree MEG beta activities can be preserved for profound studies of temporal and spectral variation across trials. Due to the high signal-to-noise ratio (SNR) in beta activities extracted through ICA, trial-specific reactive frequency ranges can be determined by means of the comparisons of two short time spectra between the reference and post-movement periods. Beta reactivity per single trial can be quantified using the amplitude modulation (AM) method (Clochon et al., 1996), and insignificant epochs can be determined using a nonparametric sign test (Brovelli et al., 2002). Source estimation and localization techniques can be successfully applied to single-trial epoch to estimate the source locations of beta modulation.

The current study presents: 1) a novel ICA-based spatiotemporal approach for single-trial analysis of event-related beta oscillatory modulations with a high extraction rate; 2) the prospect of trial-specific frequency bandpass filtering that takes into account subtle trial-bytrial brain dynamics; 3) the feasibility of using sophisticated source estimation/localization methods demanding high signal-to-noise ratio (SNR) on single trial data; and 4) a common template approach permitting an effective alternative in cases where lengthy procedures cannot be endured by participants or in clinical settings where patients have attention problems or are incapable of sustaining long experiments. The proposed ICA-based approach was applied to discover the mechanisms of beta ERS in one Parkinson's patient. It is helpful to investigate the reasoning of ERS vanishment due to suppression of postmovement beta rebound in each single-trial, rather than the cause of temporal jittering and/or loss of synchronization in Parkinson's disease.

## **2. Materials and methods**

## **2.1 Subjects and task**

86 Diagnostics and Rehabilitation of Parkinson's Disease

patients have sustained motor cortex reactivity which can be attributed to impaired cortical

ERD and ERS activities are time-locked, but not phase-locked, to external stimuli or tasks (Andrew & Pfurtscheller, 1995; Kalcher & Pfurtscheller, 1995; Pfurtscheller & Lopes da Silva, 1999). Existing methods for extraction of ERD/ERS signals essentially measure power or amplitude changes of corresponding frequency bands as derived from the average of dozens or hundreds of trials. The band power method squares and averages filtered brain signals within a selected frequency band (Pfurtscheller & Aranibar, 1977), and an inter-trial variance method to remove the phase-locked portion in the band power method was reported by Klimesch et al. (1998). Likewise, autoregressive and spectral decomposition methods have been used to extract significant frequency components in rhythmic signals (Florian & Pfurtscheller, 1995). Salmelin's temporal-spectral evolution method rectifies and averages filtered MEG signals (Salmelin et al., 1995). To increase the temporal resolution of the ERD/ERS technique, Clochon et al. (1996) proposed an amplitude modulation (AM) method based on the Hilbert transform to detect the envelope of filtered signals by squaring and summing their real and imaginary parts. All these approaches presume stereotypical frequency and temporal characteristics across trials and require an average of many trials for the ERD/ERS using a preset frequency filter and time window to preprocess every trial. However, non-phase-locked rhythmic signals can vary from trial-to-trial contingent upon variations in a subject's performance and state, which may be linked to fluctuations in expectation, attention, arousal, and task strategy (Bastiaansen et al., 2001; Bastiaansen et al., 1999; Earle, 1988; Haig et al., 1995; Hoffman et al., 1991; Yabe et al., 1993). Since trial-to-trial variability in amplitude, latencies, or scalp distribution might carry important information on cognitive and physiological states (Jung et al., 2001), a method that permits the extraction and analysis of the oscillatory signal on a single-trial base is crucial for the study of subtle brain dynamics. Furthermore, such a method should require fewer trials for analysis and hence shorter experiment time, which is beneficial for patients with impairment of motor

Single-trial multi-channel EEG analysis has been developed for time-locked, phase-locked, evoked brain activities (Jung et al., 2001; Tang et al., 2002). However, approaches to singletrial movement-related oscillatory changes are less explored. Independent component analysis (ICA), a data-driven method for multivariate data analysis, has been used to reveal temporally-independent neuronal activities of EEG measurements (Jung et al., 2001; Makeig et al, 1997; McKeown et al., 1998), MEG measurements (Wu et al., 2002; Wu et al., 2003; Tang et al., 2002), fMRI (Duann et al., 2002; McKeown et al., 1998) and recently perfusion MRI (Kao et al., 2003). The present study proposes a new approach using ICA and the Hilbert transformation for the single-trial detection of movement-related beta rhythmic activity during a self-paced right finger lifting task. This study focuses on beta activity and beta ERS, centered around 20 Hz, because it has been demonstrated that the movement-related short bursts of beta oscillation have higher task and movement specificity than alpha ERD

Since brain oscillation may be expressed alone in a specific frequency band independent of artifacts (Ermer et al., 2000; Lins et al., 1993a; Lins et al., 1993b; Mosher et al., 1992), ICA is applied to transform brain signals across all channels (in a single trial) into mutually independent components by means of an unmixing matrix in which each column represents a spatial map tailoring the weights of the corresponding temporal component at each MEG sensor. The spatial maps and temporal waveforms of decomposed independent components

inhibition (Pfurtscheller & Lopes da Silva, 1999).

and/or cognitive performance (Muller-Gerking et al., 1999).

(Pfurtscheller & Aranibar, 1979b; Pfurtscheller et al., 1996).

The present study examined six healthy right-handed subjects (gender balanced), aged 24-30 years. Five of the healthy subjects were used in the model generation group, and MEG data from the last healthy subject were used for validation. Subjects performed self-paced lifting of the right index finger approximately once every 8 sec. Subjects were trained to perform the movement briskly for a duration of 200 to 300 ms, as monitored by surface electromyogram (EMG) on extensor digitorum communis, with a range of finger movement around 35~40, while keeping their eyes open in order to suppress the occipital alpha rhythm. In addition, somatosensory evoked fields (SEFs) for right median nerve stimulation were measured to locate the primary sensorimotor area (SMI) in each subject as part of the procedure for the generation of a temporal template (see below). Informed written consent was obtained from all subjects. This study was approved by the Institutional Review Board of Taipei Veterans General Hospital. In addition, one 56-year-old patient with idiopathic Parkinson's disease in Hoehn and Yahr stage 1 was also recruited as a demonstration in this study.

Extraction of Single-Trial Post-Movement MEG Beta

*p* mutually independent random variables, thus:

components are preserved in the FastICA calculation.

distributions of *si*, for *i* = 1,…, *p*.

*<sup>b</sup>* 1 2 [ ]*<sup>T</sup> bb b <sup>m</sup>* .

factorized as:

matrix **W**.

Synchronization in Normal and Parkinson's Patient Using ICA-Based Spatiotemporal Approach 89

Mathematically, we can consider each row of **B** as samples generated from one random variable *bi*, *i* = 1, 2, …, *m*. In other words, matrix **B** is a realization of a random vector

The ICA techniques (Jung et al., 2001; Hyvarinen et al., 2001) seek to find a *p m* ( *p m* ) matrix, **W**, which converts the random vector *b* into another vector variable, *s*, consisting of

> 2 1

*s s*

. . . 

 

*s*

 

*mmp*

The mutual independence of *si*, for *i* = 1,…, *p*, implies that if *P*(*si*) represents the probability distribution of the *i*th component, the joint probability distribution for all components can be

The ICA techniques use this assumption of mutual independence to find the un-mixing

All calculations in the present study were carried out using the FastICA algorithm which features high speed calculation (cubic convergence) and does not require selection of step size parameters or learning rate, unlike the gradient-based algorithm (Hyvarinen et al., 1997, 2001). The FastICA technique first removes means of row vectors in the **B** sample matrix such that each random variable bi has a zero mean, and then employs a whitening process using principal component analysis. After whitening, the covariance matrix of the whitened data becomes an identity matrix, and only the first *p* (*p m* ) most significant principal

The next step is to look for a matrix that transforms the whitened data into a set of components as mutually independent as possible. Mutual information, as a measure of the independence of random variables, is used as the criterion for finding such a transformation. Mutual information can be expressed in terms of negentropy, an important measure of non-Gaussianity (Hyvarinen et al., 1997, 2001). Therefore, the problem of finding the independent components (*s*) and the transform matrix (**W**) can be translated into a search for linear combinations of the whitened data that maximize the negentropy of the

After applying FastICA to the pre-processed single-trial MEG epochs, matrix **B** can be

factored into a (mixing) matrix **U** and an (independent source) matrix **S** as follows:

*p*

1

*p*

1

)()...()(),...,,( <sup>21</sup> *<sup>p</sup>* <sup>21</sup> *<sup>p</sup> sPsPsPsssP* (2)

*s* **W** (1)

*b*

### **2.2 Data recording**

Cortical magnetic signals were recorded with a 306-channel (102 sensor unit) whole-head neuromagnetometer (band-pass, 0.05-250 Hz; digitized at 1kHz; Vectorview; Neuromag Ltd., Helsinki, Finland) with subjects in sitting position. Each sensor unit was composed of a pair of planar gradiometers and a magnetometer. The magnetometer measured magnetic flux ( *Bz* ), normal to the sensor unit, while the gradiometers measured two tangential derivatives of *Bz* ( *B x <sup>z</sup>* / and *B y <sup>z</sup>* / , mutually orthogonal). Only magnetic signals measured by the gradiometers were used in this study. Bipolar horizontal and vertical electro-oculograms (EOG) were recorded using electrodes placed below and above the left eye and at the bilateral outer canthi to monitor eye movement and blinks. The exact position of the head with respect to the sensor array was determined by measuring magnetic signals from four head position indicator (HPI) coils placed on the scalp. Coil positions were identified with a three-dimensional digitizer with respect to three predetermined landmarks (naison and bilateral preauricular points) on the scalp, and this data used to superimpose MEG source signals on individual MRI images obtained with a 3.0 T Bruker MedSpec S300 system (Bruker, Kalsrube, Germany). The anatomical image was acquired using a highresolution T1-weighted, 3D gradient-echo pulse sequence (MDEFT: Modified Driven Equilibrium Fourier Transform; TR/TE/TI= 88.1ms/4.12ms/650ms, 128\*128\*128 matrix, FOV=250mm).

Empty room measurements were recorded for 3 minutes. Approximately 100 EOG-free trials of right index finger lifting were acquired and analyzed off-line. Since the focus was on beta-activities, the signals were further band-pass-filtered between 6-50 Hz (zero-phase, tenth-order, IIR Butterworth filter) to remove dc drifts and 60 Hz noise. The initial finger movement (movement onset; zero time) was registered with an optical switch (Taniguchi et al., 2000). Electromyographic (EMG) activity from the extensor digitorum communis (digitized at 1 KHz) was continuously recorded to monitor performance (see above). Each epoch comprised data points from –4s to 3s relative to the movement onset (Salmelin et al., 1995; Salmelin and Hari, 1994a) and epochs were subjected to further single-trial ICA analysis.

For SEF measurement, the right median nerve was electrically stimulated every 2 sec with constant current pulses (0.3 msec in duration) exceeding the motor threshold. Approximately 100 EOG-free trials were acquired and digitized at 1 kHz for off-line analysis.

## **2.3 Data analysis**

#### **2.3.1 Independent Component Analysis of the single-trial MEG epoch**

We take the advantages of sensitivity and localizing power of superficial sources by planar gradiometers (Rosell et al., 2001; Kajola et al., 1991). Each single-trial MEG epoch contains m channels (m = 204, 102 pairs of gradiometers) and n time points (usually m < n). The paired

gradiometer signals ( *B x <sup>z</sup>* / and *B y <sup>z</sup>* / ) are arranged into two 2 *<sup>m</sup> <sup>n</sup>* sub-matrices **B1**

and **B2** and concatenated into an *m n* matrix **B**. The *i*th rows (*i* 102) of **B1** and **B2** contain the measured gradiometer signals from the *i*th sensor location, and the *j*th column in **B** contains the measured data at the *j*th time point across all gradiometer channels.

Cortical magnetic signals were recorded with a 306-channel (102 sensor unit) whole-head neuromagnetometer (band-pass, 0.05-250 Hz; digitized at 1kHz; Vectorview; Neuromag Ltd., Helsinki, Finland) with subjects in sitting position. Each sensor unit was composed of a pair of planar gradiometers and a magnetometer. The magnetometer measured magnetic flux ( *Bz* ), normal to the sensor unit, while the gradiometers measured two tangential derivatives of *Bz* ( *B x <sup>z</sup>* / and *B y <sup>z</sup>* / , mutually orthogonal). Only magnetic signals measured by the gradiometers were used in this study. Bipolar horizontal and vertical electro-oculograms (EOG) were recorded using electrodes placed below and above the left eye and at the bilateral outer canthi to monitor eye movement and blinks. The exact position of the head with respect to the sensor array was determined by measuring magnetic signals from four head position indicator (HPI) coils placed on the scalp. Coil positions were identified with a three-dimensional digitizer with respect to three predetermined landmarks (naison and bilateral preauricular points) on the scalp, and this data used to superimpose MEG source signals on individual MRI images obtained with a 3.0 T Bruker MedSpec S300 system (Bruker, Kalsrube, Germany). The anatomical image was acquired using a highresolution T1-weighted, 3D gradient-echo pulse sequence (MDEFT: Modified Driven Equilibrium Fourier Transform; TR/TE/TI= 88.1ms/4.12ms/650ms, 128\*128\*128 matrix,

Empty room measurements were recorded for 3 minutes. Approximately 100 EOG-free trials of right index finger lifting were acquired and analyzed off-line. Since the focus was on beta-activities, the signals were further band-pass-filtered between 6-50 Hz (zero-phase, tenth-order, IIR Butterworth filter) to remove dc drifts and 60 Hz noise. The initial finger movement (movement onset; zero time) was registered with an optical switch (Taniguchi et al., 2000). Electromyographic (EMG) activity from the extensor digitorum communis (digitized at 1 KHz) was continuously recorded to monitor performance (see above). Each epoch comprised data points from –4s to 3s relative to the movement onset (Salmelin et al., 1995; Salmelin and Hari, 1994a) and epochs were subjected to further single-trial ICA

For SEF measurement, the right median nerve was electrically stimulated every 2 sec with constant current pulses (0.3 msec in duration) exceeding the motor threshold. Approximately 100 EOG-free trials were acquired and digitized at 1 kHz for off-line

We take the advantages of sensitivity and localizing power of superficial sources by planar gradiometers (Rosell et al., 2001; Kajola et al., 1991). Each single-trial MEG epoch contains m channels (m = 204, 102 pairs of gradiometers) and n time points (usually m < n). The paired

and **B2** and concatenated into an *m n* matrix **B**. The *i*th rows (*i* 102) of **B1** and **B2** contain the measured gradiometer signals from the *i*th sensor location, and the *j*th column in **B** contains the measured data at the *j*th time point across all gradiometer channels.

*<sup>m</sup> <sup>n</sup>* sub-matrices **B1**

**2.3.1 Independent Component Analysis of the single-trial MEG epoch** 

gradiometer signals ( *B x <sup>z</sup>* / and *B y <sup>z</sup>* / ) are arranged into two 2

**2.2 Data recording** 

FOV=250mm).

analysis.

analysis.

**2.3 Data analysis** 

Mathematically, we can consider each row of **B** as samples generated from one random variable *bi*, *i* = 1, 2, …, *m*. In other words, matrix **B** is a realization of a random vector *<sup>b</sup>* 1 2 [ ]*<sup>T</sup> bb b <sup>m</sup>* .

The ICA techniques (Jung et al., 2001; Hyvarinen et al., 2001) seek to find a *p m* ( *p m* ) matrix, **W**, which converts the random vector *b* into another vector variable, *s*, consisting of *p* mutually independent random variables, thus:

$$\mathbf{s} = \begin{bmatrix} s\_1 \\ s\_2 \\ \vdots \\ \vdots \\ \vdots \\ \vdots \\ \mathbf{s}\_p \end{bmatrix} = \mathbf{W} \cdot \mathbf{b} \tag{1}$$

The mutual independence of *si*, for *i* = 1,…, *p*, implies that if *P*(*si*) represents the probability distribution of the *i*th component, the joint probability distribution for all components can be factorized as:

$$P(\mathbf{s}\_1, \mathbf{s}\_2, \dots, \mathbf{s}\_p) = P(\mathbf{s}\_1)P(\mathbf{s}\_2)\dots P(\mathbf{s}\_p) \tag{2}$$

The ICA techniques use this assumption of mutual independence to find the un-mixing matrix **W**.

All calculations in the present study were carried out using the FastICA algorithm which features high speed calculation (cubic convergence) and does not require selection of step size parameters or learning rate, unlike the gradient-based algorithm (Hyvarinen et al., 1997, 2001). The FastICA technique first removes means of row vectors in the **B** sample matrix such that each random variable bi has a zero mean, and then employs a whitening process using principal component analysis. After whitening, the covariance matrix of the whitened data becomes an identity matrix, and only the first *p* (*p m* ) most significant principal components are preserved in the FastICA calculation.

The next step is to look for a matrix that transforms the whitened data into a set of components as mutually independent as possible. Mutual information, as a measure of the independence of random variables, is used as the criterion for finding such a transformation. Mutual information can be expressed in terms of negentropy, an important measure of non-Gaussianity (Hyvarinen et al., 1997, 2001). Therefore, the problem of finding the independent components (*s*) and the transform matrix (**W**) can be translated into a search for linear combinations of the whitened data that maximize the negentropy of the distributions of *si*, for *i* = 1,…, *p*.

After applying FastICA to the pre-processed single-trial MEG epochs, matrix **B** can be factored into a (mixing) matrix **U** and an (independent source) matrix **S** as follows:

Extraction of Single-Trial Post-Movement MEG Beta

**(a)**

**(a)** 

**(b)** 

**(b)**

**1200**

**Amplitude spectrum (fT/cm)** 

**0**

**600**

**(c)**

**(c)** 

**-600**

(,) *m it <sup>x</sup>* and (,) *m it <sup>y</sup>* are the AM waveforms in *Bz*

VAMW*template* (Fig. 2a).

difference between two spectra exceeds the 95% confidence level.

**Amplitude (fT/cm)** 

Synchronization in Normal and Parkinson's Patient Using ICA-Based Spatiotemporal Approach 91

preceding the onset of movement, and the other (serving as reactive target) from 0.8s to 1.8s after the onset of movement (see Fig. 1a, b). All beta-frequency components with significant modulation in terms of post-movement amplitude increase (above 95% confidence level, i.e. Z>3.09, P<0.01) in the differential amplitude spectrum (see Fig. 1c) are taken as the taskspecific frequency band for subsequent processing (Pfurtscheller G. and Berghold A., 1989).

> - Post-movement (P) - - Reference period (R)


**-4s ~ -3s 0.8s ~ 1.8s** 

**0 50Hz**

**95% Confidence level** 

**Difference of two spectra** 

**0 50Hz**

Fig. 1. Determination of task-specific frequency band using two 1-s amplitude spectra. (a) "R" represents the reference period from -4s to -3s preceding onset of movement and "P" represents the post-movement duration from 0.8s to 1.8s after onset of movement. (b) Two spectra computed over the reference (R) and post-movement periods (P), respectively. (c) The task-specific frequency band for beta-band VAMW is defined as the one where the

The vector norm of AM waveforms (VAMW) at each sensor site is computed using the square root of the AM waveforms of each gradiometer pair, i.e., 2 2 (,) (,) (,) *Vit m it m it x y* , in which *Vit* (,) is the VAMW at the *i*th sensor location, and

location. Event-related beta modulation is then computed as the difference in amplitude between the maximum amplitude of VAMW for each sensor site in the post-movement (0.8s to 1.8s) interval and mean activity between -2.5s and -2s (see Fig. 2a) (Leocani et al., 1997). Beta rebound (BR) is defined as the maximum amplitude of the computed event-related beta modulation from the subset of nine sensor sites in the vicinity of SMI (identified by SEF). The VAMWs of the BR calculation were averaged across the subjects (500 trials, 100 trials for each subject, 5 subjects pooled) to create the common temporal template, designated

*x* 

 and *Bz y* 

directions of the *i*th sensor

$$\begin{aligned} \mathbf{B}\_{\text{max}} &= \begin{bmatrix} \mathbf{B}\_1 \\ \mathbf{B}\_2 \end{bmatrix} = \mathbf{U}\_{\text{max}} \mathbf{S}\_{\text{pran}} \\\\ \begin{bmatrix} u\_{1,1} & \cdots & u\_{1,p} \\ \vdots & & \vdots \\ u\_{m,1} & \cdots & u\_{m,p} \\ \end{bmatrix} \end{aligned} \qquad \begin{aligned} \mathbf{U}\_{\text{pran}} \\ \vdots \\ u\_{m,1} & \cdots \\ \frac{u\_{m,1}}{2}, 1 \end{aligned} \qquad \begin{aligned} \mathbf{U}\_{\text{pran}} \\ \vdots \\ \mathbf{U}\_{\text{pran}} \\ \begin{bmatrix} \mathbf{s}\_1 \\ \mathbf{s}\_2 \\ \vdots \\ \mathbf{s}\_p \end{bmatrix} \end{aligned} \qquad \begin{aligned} \mathbf{U}\_{\text{pran}} \\ \begin{bmatrix} \vdots \\ \mathbf{s}\_1 \\ \vdots \\ \mathbf{s}\_p \end{bmatrix} \end{aligned} \tag{3}$$

in which each row *is* of matrix *p<sup>n</sup>* S represents samples of an independent component (IC) *si*, for *i* = 1,…, p and *m<sup>p</sup>* U is the pseudo-inverse of matrix **W** whose column vectors represent the weight distribution values of the corresponding ICs in *S* across all MEG gradiometer channels. In fact, matrix **U** is the "mixing matrix" that combines the *p* ICs to reconstruct signal **B**. These temporal ICs can be categorized into task-related ICs and taskunrelated ICs. Since the elicited brain activities or artifacts can be distributed over multiple ICs, no one-to-one correspondence between IC and source information is projected (Makeig et al., 1997). To facilitate the selection of task-related ICs, a temporal and spatial template pair was constructed prior to selection (see below). Spatial map *<sup>j</sup> x* of the *j*th IC was defined as the topographic display of all vector norms for weights of 102 gradiometer pairs in the *j*th column vector of U,

$$\bar{\mathbf{x}}\_{j} = \left[ \sqrt{{u\_{1,j}}^2 + {u\_{\frac{m}{2}+1}}^2} \quad \sqrt{{u\_{2,j}}^2 + {u\_{\frac{m}{2}+2}}^2} \quad \cdots \quad \sqrt{{u\_{\frac{m}{2},j}}^2 + {u\_{m,j}}^2} \right]^T \tag{4}$$

in which *ui j*, is the entry in the *i*th row and *j*th column of **U** in Eq. (3). The spatial map is intended for component selection (see below).

#### **2.3.2 Creation of a temporal template (VAMWtemplate) using amplitude modulation (envelope) of the MEG data**

The recorded MEG signals at each gradiometer are filtered in the task-specific frequency band (Pfurtscheller & Lopes da Silva, 1999) and rectified by computing the AM waveform (envelope) using the amplitude modulation (AM) method (Clochon et al., 1996) as follows:

$$m(t) = \sqrt{M\_{BP}(t)^2 + H(M\_{BP}(t))^2} \tag{5}$$

in which ( ) *MBP t* is the band-passed MEG signal and ( ( )) *HM t BP* is its Hilbert transform. The task-specific frequency band is determined by the contrast between two 1-s amplitude spectra calculated over about one hundred event-related EEG trials (Pfurtscheller and Lopes da Silva, 1999). One (serving as rest reference) is computed over the duration from 4s to 3s

,1 , 2 2 <sup>2</sup>

*p*

pxn

2 2 ( ) ( ) ( ( )) *mt M t H M t BP BP* (5)

(3)

of the *j*th IC was defined

T

(4)

1

*s*

*s*

*p*

*s*

of matrix *p<sup>n</sup>* S represents samples of an independent component

mxp

mxn mxp pxn

 

B U S **<sup>1</sup> 2 B B**

1,1 1,

 

*u u*

*u u*

*u u*

1,1 1, 2 2

 

*m m <sup>p</sup>*

(IC) *si*, for *i* = 1,…, p and *m<sup>p</sup>* U is the pseudo-inverse of matrix **W** whose column vectors represent the weight distribution values of the corresponding ICs in *S* across all MEG gradiometer channels. In fact, matrix **U** is the "mixing matrix" that combines the *p* ICs to reconstruct signal **B**. These temporal ICs can be categorized into task-related ICs and taskunrelated ICs. Since the elicited brain activities or artifacts can be distributed over multiple ICs, no one-to-one correspondence between IC and source information is projected (Makeig et al., 1997). To facilitate the selection of task-related ICs, a temporal and spatial template

as the topographic display of all vector norms for weights of 102 gradiometer pairs in the *j*th

2 2 2 2 22 1, 2, , ( 1), ( 2), , 2 22 *j jm j m m mj j jj x uu uu u u*

 

in which *ui j*, is the entry in the *i*th row and *j*th column of **U** in Eq. (3). The spatial map is

The recorded MEG signals at each gradiometer are filtered in the task-specific frequency band (Pfurtscheller & Lopes da Silva, 1999) and rectified by computing the AM waveform (envelope) using the amplitude modulation (AM) method (Clochon et al., 1996) as

in which ( ) *MBP t* is the band-passed MEG signal and ( ( )) *HM t BP* is its Hilbert transform. The task-specific frequency band is determined by the contrast between two 1-s amplitude spectra calculated over about one hundred event-related EEG trials (Pfurtscheller and Lopes da Silva, 1999). One (serving as rest reference) is computed over the duration from 4s to 3s

**2.3.2 Creation of a temporal template (VAMWtemplate) using amplitude modulation** 

*m m <sup>p</sup>*

,1 ,

*m m p*

*u u*

pair was constructed prior to selection (see below). Spatial map *<sup>j</sup> x*

in which each row *is*

column vector of U,

intended for component selection (see below).

**(envelope) of the MEG data** 

follows:

preceding the onset of movement, and the other (serving as reactive target) from 0.8s to 1.8s after the onset of movement (see Fig. 1a, b). All beta-frequency components with significant modulation in terms of post-movement amplitude increase (above 95% confidence level, i.e. Z>3.09, P<0.01) in the differential amplitude spectrum (see Fig. 1c) are taken as the taskspecific frequency band for subsequent processing (Pfurtscheller G. and Berghold A., 1989).

Fig. 1. Determination of task-specific frequency band using two 1-s amplitude spectra. (a) "R" represents the reference period from -4s to -3s preceding onset of movement and "P" represents the post-movement duration from 0.8s to 1.8s after onset of movement. (b) Two spectra computed over the reference (R) and post-movement periods (P), respectively. (c) The task-specific frequency band for beta-band VAMW is defined as the one where the difference between two spectra exceeds the 95% confidence level.

The vector norm of AM waveforms (VAMW) at each sensor site is computed using the square root of the AM waveforms of each gradiometer pair, i.e., 2 2 (,) (,) (,) *Vit m it m it x y* , in which *Vit* (,) is the VAMW at the *i*th sensor location, and (,) *m it <sup>x</sup>* and (,) *m it <sup>y</sup>* are the AM waveforms in *Bz x* and *Bz y* directions of the *i*th sensor

location. Event-related beta modulation is then computed as the difference in amplitude between the maximum amplitude of VAMW for each sensor site in the post-movement (0.8s to 1.8s) interval and mean activity between -2.5s and -2s (see Fig. 2a) (Leocani et al., 1997). Beta rebound (BR) is defined as the maximum amplitude of the computed event-related beta modulation from the subset of nine sensor sites in the vicinity of SMI (identified by SEF). The VAMWs of the BR calculation were averaged across the subjects (500 trials, 100 trials for each subject, 5 subjects pooled) to create the common temporal template, designated VAMW*template* (Fig. 2a).

Extraction of Single-Trial Post-Movement MEG Beta

**25**

**Magnetic Amplitude** 

**Subject IV**

**Subject V**

**(b)** 

**Subject IV**

**Subject III**

**Subject I** 

**Subject II**

**Subject V**

**(a)** 

**0**

**Subject III**

**-2.5 2.5 Time (sec)**

**45**

**Magnetic Amplitude** 

**Subject I**

**0.08**

**Spatial Weig ht** 

**0** 

**0.06**

**(Normalized)** 

**Spatial Weig ht** 

**0** 

**0.05**

**0 Spatial Weight** 

**0.07**

**Spatial Weig ht** 

**(Normalized)** 

**(Normalized)** 

**0.08**

**0** 

**(Normalized)** 

**(Norma lized)**  **Spa tial Weight** 

Fig. 2. Creation of common temporal and spatial templates. (a) The common temporal template, VAMW*template*, is created by averaging VAMWs (500 trials, 100 trials for each subject, 5 subjects pooled). Event-related beta modulation is defined as the amplitude difference between the mean amplitude of baseline activity (-2.5 to -2 s) and maximum amplitude in the post-movement interval (0.8 to 1.8 s). (b) The common spatial template is the average of the topographical distributions of event-related beta modulations of five subjects from model generation group. Only the half the spatial map (unshaded)

**0** 

contralateral to the side of finger lifting is used as the spatial template.

**0**

**(fT/cm)**

**-2.5 2.5 Time (sec)**

**25**

**Magnetic Amplitude** 

**0**

**(fT/cm)**

**-2.5 2.5 Time (sec)**

**-2.5 2.5 Time (sec)**

**Average**

**-2.5 2.5 Time (sec)**

**45**

**Magnetic Amplitude** 

**0**

**(fT/cm)**

**0**

**Subject II**

**(fT/cm)**

**Magnetic Amplitude** 

**(fT/cm)**

Synchronization in Normal and Parkinson's Patient Using ICA-Based Spatiotemporal Approach 93

**Average**

0.8s ~ 1.8s **50**

**40**

**Magnetic Amplitude** 

**0**

**-2.5 2.5 Time (sec)** *Movement onset*

**0.07** 

**Spatial Weight** 

**(Normalized)** 

**Left Right**

**0** 

**0**

**VAMWtemplate**

**Event-related beta modulation**

**(fT/cm)**

Reference period (-2.5~-2s)

#### **2.3.3 Creation of a spatial template using topographical distribution of event-related beta modulation values**

Individual spatial templates were first generated from the topographical distributions of event-related beta modulation values (see above). The five templates from the model generation group were then averaged to generate a common spatial template. In order to optimize conditions for spatial averaging, subjects' heads were carefully positioned before actual measurements to keep head positioning and orientation as similar as possible. Distances between head centers of the five subjects and the reference point (the origin of the MEG sensor array) in the horizontal plane were less than 4mm, and angles between the vertical axis of the helmet and that of the head (the normal vector of the plane constituted by the three landmark points, i.e., nasion, and both pre-auricular points) remained within 5.5 (maximum deviation 1.5) between subjects.

Only the left half of the spatial map (unshaded in Fig. 2b) was used as the spatial template because this study focused on beta event-related activities in the hemisphere contralateral to the side of finger lifting; however, the other half can be generated analogously to extract activities in the ipsilateral hemisphere. Correlations among individual spatial templates ranged from 0.92 to 0.68. Respective correlations between the common spatial template and the individual spatial templates were 0.973, 0.811, 0.881, 0.904, and 0.915. These high correlation values support the use of the spatial template in component selection for each individual's magnetic signals.

#### **2.3.4 Selection of pertinent independent components for the reconstruction of reactive beta activities**

A spatial map (Eq. (4)) and corresponding VAMWs of each IC were generated for the selection of task-related ICs. Since the original signals may be decomposed into multiple ICs, the spectrum of each IC may vary from the one in the original signal due to the decomposition process. When settings for band-pass filtering for VAMW computation cannot be optimally determined using two-spectrum comparison for the generation of a VAMW*template* (Pfurtscheller & Lopes da Silva, 1999), three standard beta bands, 12-16, 16-20 and 20-24 Hz (Pfurtscheller G., 1981), enclosing the event-related beta activities in motor task, were used to band-pass filter (zero-phase, tenth-order, IIR Butterworth filter) for each single-trial IC such that the three frequency-laden resultant VAMW*IC*s (the VAMWs bandpass filtered in three frequency bands of each IC) retained all task-related information. These VAMW*IC*s were subsequently used in the selection of task-related ICs, which must fulfill the following dual criteria: 1) at least one of three corresponding VAMW*IC*s has a correlation with the VAMW*template* higher than 95% (Z>1.63, P<0.05) among VAMW*IC*s of all the ICs for that single epoch, and 2) correlation between the spatial map and spatial template is above 95% (Z>1.63, P<0.05) for the spatial maps of all ICs. Data processed via 3 standard band filtering are not used in subsequent data reconstruction, but rather are used in conjunction with the dual-criteria only in the procedure "selecting" the pertinent ICs. Unselected columns, i.e., task-unrelated components, of mixing matrix **U** (Eq. (3)) are zeroed to produce a matrix ˆ U such that task-related rhythmic signals are reconstructed by multiplying ˆ U and **S** (Fig. 3). The reconstructed data in each trial are then filtered within a trial-specific frequency band to extract reactive beta activities.

**2.3.3 Creation of a spatial template using topographical distribution of event-related** 

Individual spatial templates were first generated from the topographical distributions of event-related beta modulation values (see above). The five templates from the model generation group were then averaged to generate a common spatial template. In order to optimize conditions for spatial averaging, subjects' heads were carefully positioned before actual measurements to keep head positioning and orientation as similar as possible. Distances between head centers of the five subjects and the reference point (the origin of the MEG sensor array) in the horizontal plane were less than 4mm, and angles between the vertical axis of the helmet and that of the head (the normal vector of the plane constituted by the three landmark points, i.e., nasion, and both pre-auricular points) remained within 5.5

Only the left half of the spatial map (unshaded in Fig. 2b) was used as the spatial template because this study focused on beta event-related activities in the hemisphere contralateral to the side of finger lifting; however, the other half can be generated analogously to extract activities in the ipsilateral hemisphere. Correlations among individual spatial templates ranged from 0.92 to 0.68. Respective correlations between the common spatial template and the individual spatial templates were 0.973, 0.811, 0.881, 0.904, and 0.915. These high correlation values support the use of the spatial template in component selection for each

**2.3.4 Selection of pertinent independent components for the reconstruction of** 

A spatial map (Eq. (4)) and corresponding VAMWs of each IC were generated for the selection of task-related ICs. Since the original signals may be decomposed into multiple ICs, the spectrum of each IC may vary from the one in the original signal due to the decomposition process. When settings for band-pass filtering for VAMW computation cannot be optimally determined using two-spectrum comparison for the generation of a VAMW*template* (Pfurtscheller & Lopes da Silva, 1999), three standard beta bands, 12-16, 16-20 and 20-24 Hz (Pfurtscheller G., 1981), enclosing the event-related beta activities in motor task, were used to band-pass filter (zero-phase, tenth-order, IIR Butterworth filter) for each single-trial IC such that the three frequency-laden resultant VAMW*IC*s (the VAMWs bandpass filtered in three frequency bands of each IC) retained all task-related information. These VAMW*IC*s were subsequently used in the selection of task-related ICs, which must fulfill the following dual criteria: 1) at least one of three corresponding VAMW*IC*s has a correlation with the VAMW*template* higher than 95% (Z>1.63, P<0.05) among VAMW*IC*s of all the ICs for that single epoch, and 2) correlation between the spatial map and spatial template is above 95% (Z>1.63, P<0.05) for the spatial maps of all ICs. Data processed via 3 standard band filtering are not used in subsequent data reconstruction, but rather are used in conjunction with the dual-criteria only in the procedure "selecting" the pertinent ICs. Unselected columns, i.e., task-unrelated components, of mixing matrix **U** (Eq. (3)) are zeroed

U such that task-related rhythmic signals are reconstructed by

U and **S** (Fig. 3). The reconstructed data in each trial are then filtered within a

**beta modulation values** 

(maximum deviation 1.5) between subjects.

individual's magnetic signals.

**reactive beta activities** 

to produce a matrix ˆ

trial-specific frequency band to extract reactive beta activities.

multiplying ˆ

Fig. 2. Creation of common temporal and spatial templates. (a) The common temporal template, VAMW*template*, is created by averaging VAMWs (500 trials, 100 trials for each subject, 5 subjects pooled). Event-related beta modulation is defined as the amplitude difference between the mean amplitude of baseline activity (-2.5 to -2 s) and maximum amplitude in the post-movement interval (0.8 to 1.8 s). (b) The common spatial template is the average of the topographical distributions of event-related beta modulations of five subjects from model generation group. Only the half the spatial map (unshaded) contralateral to the side of finger lifting is used as the spatial template.

Extraction of Single-Trial Post-Movement MEG Beta

(arrows and trace in blue; IC9 included for reconstruction).

subjected to source estimation and beta rebound (BR) computations.

**selection using a nonparametric sign test** 

point as 1 1 ( ) ( ( ) )/( ) 2 2

P<0.05) using 1 1 ( ) ( ( ) )/( ) 2 2

**2.3.6 Calculation of VAMWrecon of reactive beta activities and single-trial epoch** 

Movement-related beta rebound (BR) can be quantified from single-epoch reactive beta activities and VAMW*recon* (VAMW of reconstructed data) for reactive beta activity at each sensor site computed. The VAMW*recon* of highest event-related beta modulation (see creation of temporal template) among the nine sensor sites vicinal to SMI is designated as VAMW*recon\_max* and is used in turn for single-trial epoch selection and BR computation, as the sensor site expressing VAMW*recon\_max* did not change throughout the experiment in our observations. A deterministic procedure, modified from Brovelli's et al. (2002) approach, is used to select the significant trial. A nonparametric sign test is applied to the VAMW*recon\_max* designated for BR calculation in each single trial by computing the Z-score at each time

*Zt N t N N* , in which *N t*( ) denotes the number of trials whose

*Zi N i N N IOI IOI IOI IOI* , in which ( ) *Z i IOI* is the Z value of the

magnitudes are larger than the median value of their baseline activities at time point t, and N the total number of trials. Time points with Z values greater than 3.09 (P<0.01) are defined as the time interval-of-interest (IOI). After the determination of IOI for each subject, another sign test is then applied to find epochs showing significant increases in amplitude (Z>1.63,

**beta activities** 

Synchronization in Normal and Parkinson's Patient Using ICA-Based Spatiotemporal Approach 95

reconstruction. For example, IC 3 meets the dual criteria (underscored in red): i) correlation value between spatial map and spatial template is 0.84 (rank= 97%, Z= 1.89, P=0.03); ii) correlation value between16-20Hz VAMW*IC* and 20-24 Hz VAMW*IC* with VAMW*template* is 0.8 (rank= 99%, Z=3.08, P=0.01) and 0.78 (rank= 97.8%, Z= 2.85, P=0.022), respectively. (b) Noise identification and removal. The deselected IC 2 in Fig. 2a may emanate from background noise since it resembles the IC 1 extracted from empty room measurement. (c) The impact of including task-unrelated IC into signal reconstruction. This figure illustrates (a different trial

from Fig. 3a) that inclusion of task-unrelated IC (IC 9) with a high spatial correlation (correlation value=0.61, rank= 95.2%, Z=1.67, P=0.048) but poor temporal correlation (correlation value=0.28, rank=13%, Z=0.34, P=0.87) causes deterioration in the beta BR from 28.9 fT/cm (arrows and trace in red; IC9 eliminated from reconstruction) to 18.6 fT/cm

**2.3.5 Detection of task-laden trial-specific frequency band and extraction of reactive** 

The trial-specific frequency band detected in each trial is used to confine the reconstructed data within the most reactive beta band for further BR computation and source estimation. This frequency band is defined by the reactive beta band of the sensor site showing highest event-related beta modulation value (see creation of temporal template) over the nine SMI vicinal sensor sites (identified by SEF) and is identified using the aforementioned twospectrum procedure which has been suggested as the best approach for the determination of reactive frequencies (Pfurtscheller & Lopes da Silva, 1999). Following data filtering with a trial-specific frequency band (zero-phase, tenth-order, IIR Butterworth filter), reactive beta activities in each single epoch can be extracted. The extracted reactive beta activities are then

single epoch by ICA. Only ICs fulfilling the dual criteria are selected for signal

Fig. 3. Examples of IC-selection and signal reconstruction procedure. (a) Spatial maps, IC waveforms, Fourier spectra of IC waveforms and VAMW*IC*s of five ICs obtained from one

**Spectrum VAMW***IC*

**12-16 Hz 16-20 Hz 20-24 Hz**

**12-16 Hz 16-20 Hz 20-24 Hz**

**12-16 Hz 16-20 Hz 20-24 Hz**

**12-16 Hz 16-20 Hz 20-24 Hz**

**12-16 Hz 16-20 Hz 20-24 Hz**

**Fourier Spectrum**

**0 50Hz**

**0 50Hz**

**20**

**0** (Arbitrary Unit)

**20**

(Arbitrary Unit)

**0 17**

**0 16**

(Arbitrary Unit)

**0 20**

(Arbitrary Unit)

**0**

**(empty room)**

**20**

(Arbitrary Unit)

**0**

**16**

(Arbitrary Unit)

**0**

**Time (sec)**

**-2.5 2.5**

**Time (sec) -2.5 2.5**

**-2.5 2.5**

**Temporal Waveform**

(Arbitrary Unit)

**0 50Hz**

**0 50Hz**

**0 50Hz**

**20**

(Arbitrary Unit)

**0**

**Waveform VAMW***IC*

**Fourier Spectrum**

**0 50**

Fig. 3. Examples of IC-selection and signal reconstruction procedure. (a) Spatial maps, IC waveforms, Fourier spectra of IC waveforms and VAMW*IC*s of five ICs obtained from one

**Hz**

**-2.5 2.5s**

**0.05**

**0**

**IC 2**

**(a)** 

**Coeff=0.05**

**Coeff=0.84**

**Coeff=0.3**

**Coeff=0.11**

**Coeff=0.06**

**IC** 1

Coeff=0.94

**IC 2**

**(c)** 

**IC 9**

**Spatial map**

**IC 3**

**IC 5**

**IC 9**

**IC 11**

**(b)** 

**0.02**

**0.1**

(Normalized)

**-0.07**

**Spatial Map Temporal** 

Amplitude

**0.07**

(Normalized)

**-0.1 0.1**

Amplitude

**-0.1 0.08**

(Normalized)

Amplitude

**-0.08 0.14**

Amplitude

(Normalized)

**-0.07**

**(empty room) Time (sec)**

**Spatial map Temporal** 

**0.05**

**0**

**0.06**

Coeff=0.61 **0.03**

**0**

**--0.03**

Amplitude

(Normalized)

**0**

Amplitude

(Normalized)

**0 0.03**

**-2.5 2.5s**

**-2.5 2.5s**

**waveform**

**-2.5 2.5s**

**-2.5 2.5s**

**-2.5 2.5s**

**0.1**

**-0.1**

(Normalized)

Amplitude

**0.1**

(Normalized)

**-0.1**

Amplitude

**0 0.03**

> **0 0.03**

**0**

**0.02**

**0.05**

**Spatial Weight** 

**0.07**

**Spatial Weight** 

**0**

**(Normalized)**

**0**

**(Normalized)**

**-2.5 2.5s**

**Left Right**

**spatial template**

**0-2.5 2.5s VAMW***template*

> **Coeff=0.06 Coeff=0.31 Coeff=0.05**

**Time (sec) -2.5 2.5**

**VAMWrecon**

**50**

**0**

**(fT/cm)**

**50**

**X**

**X**

**Coeff=0.29 Coeff=0.21 Coeff=0.02**

**Coeff=0.25 Coeff=0.8 Coeff=0.78**

**Coeff=0.73 Coeff=0.49 Coeff=0.25**

**Coeff=0.45 Coeff=0.27 Coeff=0.59**

**Coeff=0.05 Coeff=0.17 Coeff=0.03** **Magnetic Field (fT/cm)**

**-50**

**0.04**

**(Normalized)**

**VAMW***IC*

**Time (sec)**

**Coeff=0.66 Coeff=0.89 Coeff=0.22**

**Coeff=0.28 Coeff=0.12 Coeff=0.20**

**X**

**Amplitude** 

**X**

Normalized amplitude

**0.03 0**

**-2.5 2.5 0 50 Hz**

**Time (sec)**

**0 50 Hz -2.5 2.5**

**16-20 Hz 20-24 Hz**

**12-16 Hz**

**12-16 Hz 16-20 Hz 20-24 Hz**

**Time (sec) -2.5 2.5**

**12-16 Hz 16-20 Hz 20-24 Hz** **X**

single epoch by ICA. Only ICs fulfilling the dual criteria are selected for signal reconstruction. For example, IC 3 meets the dual criteria (underscored in red): i) correlation value between spatial map and spatial template is 0.84 (rank= 97%, Z= 1.89, P=0.03); ii) correlation value between16-20Hz VAMW*IC* and 20-24 Hz VAMW*IC* with VAMW*template* is 0.8 (rank= 99%, Z=3.08, P=0.01) and 0.78 (rank= 97.8%, Z= 2.85, P=0.022), respectively. (b) Noise identification and removal. The deselected IC 2 in Fig. 2a may emanate from background noise since it resembles the IC 1 extracted from empty room measurement. (c) The impact of including task-unrelated IC into signal reconstruction. This figure illustrates (a different trial from Fig. 3a) that inclusion of task-unrelated IC (IC 9) with a high spatial correlation (correlation value=0.61, rank= 95.2%, Z=1.67, P=0.048) but poor temporal correlation (correlation value=0.28, rank=13%, Z=0.34, P=0.87) causes deterioration in the beta BR from 28.9 fT/cm (arrows and trace in red; IC9 eliminated from reconstruction) to 18.6 fT/cm (arrows and trace in blue; IC9 included for reconstruction).

#### **2.3.5 Detection of task-laden trial-specific frequency band and extraction of reactive beta activities**

The trial-specific frequency band detected in each trial is used to confine the reconstructed data within the most reactive beta band for further BR computation and source estimation. This frequency band is defined by the reactive beta band of the sensor site showing highest event-related beta modulation value (see creation of temporal template) over the nine SMI vicinal sensor sites (identified by SEF) and is identified using the aforementioned twospectrum procedure which has been suggested as the best approach for the determination of reactive frequencies (Pfurtscheller & Lopes da Silva, 1999). Following data filtering with a trial-specific frequency band (zero-phase, tenth-order, IIR Butterworth filter), reactive beta activities in each single epoch can be extracted. The extracted reactive beta activities are then subjected to source estimation and beta rebound (BR) computations.

#### **2.3.6 Calculation of VAMWrecon of reactive beta activities and single-trial epoch selection using a nonparametric sign test**

Movement-related beta rebound (BR) can be quantified from single-epoch reactive beta activities and VAMW*recon* (VAMW of reconstructed data) for reactive beta activity at each sensor site computed. The VAMW*recon* of highest event-related beta modulation (see creation of temporal template) among the nine sensor sites vicinal to SMI is designated as VAMW*recon\_max* and is used in turn for single-trial epoch selection and BR computation, as the sensor site expressing VAMW*recon\_max* did not change throughout the experiment in our observations. A deterministic procedure, modified from Brovelli's et al. (2002) approach, is used to select the significant trial. A nonparametric sign test is applied to the VAMW*recon\_max* designated for BR calculation in each single trial by computing the Z-score at each time

point as 1 1 ( ) ( ( ) )/( ) 2 2 *Zt N t N N* , in which *N t*( ) denotes the number of trials whose

magnitudes are larger than the median value of their baseline activities at time point t, and N the total number of trials. Time points with Z values greater than 3.09 (P<0.01) are defined as the time interval-of-interest (IOI). After the determination of IOI for each subject, another sign test is then applied to find epochs showing significant increases in amplitude (Z>1.63,

P<0.05) using 1 1 ( ) ( ( ) )/( ) 2 2 *Zi N i N N IOI IOI IOI IOI* , in which ( ) *Z i IOI* is the Z value of the

Extraction of Single-Trial Post-Movement MEG Beta

**3. Results** 

**(a)** 

signals was further validated on one additional subject.

Synchronization in Normal and Parkinson's Patient Using ICA-Based Spatiotemporal Approach 97

spatial and temporal templates for the extraction of individuals' neuromagnetic single-trial

Based on the known spatial location and temporal expression in terms of spatial and temporal templates, reactive beta activities were successfully extracted. Figure 3a shows that IC 3 meets the dual criteria: i) the correlation values between spatial map and spatial template is 0.84 (rank= 97%, *Z*= 1.89, *P*=0.03); ii) correlation values of 16-20Hz VAMW*IC* and 20-24 Hz VAMW*IC* vs. VAMW*template* are 0.8 (rank= 99%, Z=3.08, P=0.01) and 0.78 (rank= 97.8%, Z= 2.85, P=0.022), respectively. Fig. 3a illustrates that noise could also be identified and removed. IC2 in Fig. 3a correlates highly (=0.88) in spatial distribution with the IC1

Measured MEG data

Determine the **task-specific frequency band** by comparing two short-time spectra. One spectrum is computed over reference time interval (-4s – -3s) and the other is computed over

Filter the MEG epochs measured from each gradiometer with the task-specific frequency band and rectify the results using amplitude modulation method to obtain the **AM waveforms:**

Compute the vector norm (**VAMW**) on each sensor site with a

*y 2 <sup>x</sup> timtimtiV* ),(),(),( .

**VAMWs.**

model generation group.

For each subject in model generation group, construct an **individual spatial template** which is defined as the topographical distribution of **event-related beta modulation** of averaged

The **common spatial template** is created by averaging the individual spatial templates in

**Step1**:Average the **VAMWs** over many trials for each subject. **Step2**:Compute the **event-related beta modulation** of each sensor site, which is the difference between the maximum amplitude of VAMW in the post-movement (0.8s to 1.8s) interval and the mean amplitude of **VAMW** during -2.5s to

pair of gradiometers: *<sup>2</sup>*

Fig. 5a. Flow chart for creation of common spatial and temporal templates.

extracted from empty room measurements (Fig. 3b), and is therefore rejected.

post-movement time interval (0.8s – 1.8s).

<sup>2</sup> <sup>2</sup> (()()( *tMHtMtm* )) *BP BP* .


Find the sensor site displaying maximum **event-related modulation** in each of five subjects in model generation group, and average them to create the **common temporal template**,

**VAMW***template*.

*th <sup>i</sup>* trial, ( ) *N i IOI* is the number of data points in post-movement IOI with values larger than the median of baseline activities of the *th <sup>i</sup>* trial, and *NIOI* is the total number of time points in post-movement *IOI* (Brovelli et al., 2002). An example of single-trial epoch selection is given in Fig. 4 (Subject I). The first trial in Fig. 4 with a *ZIOI* score equal to -4.53 is marked as an insignificant epoch and eliminated from further analysis.

#### **2.3.7 Source estimation of the reactive beta activities**

Source estimation of the MEG reactive beta activities was done using equivalent current dipole (ECD) analysis and minimum current estimation (MCE, Uutela et al., 1999; toolbox provided by Neuromag Ltd, Helsinki, Finland). A single dipole model was applied to explain the field every 1ms, and only dipoles showing goodness-of-fit (Jensen and Vanni, 2002) values higher than 80% were used for data explanation. In MCE, the lattice constant of the triangular grid was 10mm and locations closer than 30mm to the center of the conductor were excluded from current estimates. Both analyses used a realistic head model for each subject. Template generation and single-trial data processing procedure are schematized in Figs. 5a and 5b respectively. Epochs achieving significance in the increase of beta activities were chosen for subsequent BR calculation and dipole/source analysis.

Fig. 4. Example of single-trial epoch selection based on a nonparametric sign test. Single-trial VAMW*recon\_max*s of reconstructed data are examined through a nonparametric sign test. ( ) *Z i IOI*

is the Z value of the *th <sup>i</sup>* trial, ( ) *N i IOI* is the number of data points in post-movement *IOI* with

values larger than the median of baseline activities of the *th <sup>i</sup>* trial, and *NIOI* is the total

number of time points in post-movement *IOI*. Only epochs showing significant increase of beta activities are chosen for further analysis. The first trial with a *ZIOI* score equal to -4.53 is marked as an insignificant epoch and eliminated from further analysis.

#### **2.3.8 Validation of coupled common spatial and temporal templates for single-trial analysis**

Since there are inevitably differences in head size and variations in head positions inside the MEG scanner among subjects, BR amplitude differences were compared using both individual spatial templates and the common spatial template. The use of a pair of common spatial and temporal templates for the extraction of individuals' neuromagnetic single-trial signals was further validated on one additional subject.

## **3. Results**

96 Diagnostics and Rehabilitation of Parkinson's Disease

*<sup>i</sup>* trial, ( ) *N i IOI* is the number of data points in post-movement IOI with values larger

points in post-movement *IOI* (Brovelli et al., 2002). An example of single-trial epoch selection is given in Fig. 4 (Subject I). The first trial in Fig. 4 with a *ZIOI* score equal to -4.53 is

Source estimation of the MEG reactive beta activities was done using equivalent current dipole (ECD) analysis and minimum current estimation (MCE, Uutela et al., 1999; toolbox provided by Neuromag Ltd, Helsinki, Finland). A single dipole model was applied to explain the field every 1ms, and only dipoles showing goodness-of-fit (Jensen and Vanni, 2002) values higher than 80% were used for data explanation. In MCE, the lattice constant of the triangular grid was 10mm and locations closer than 30mm to the center of the conductor were excluded from current estimates. Both analyses used a realistic head model for each subject. Template generation and single-trial data processing procedure are schematized in Figs. 5a and 5b respectively. Epochs achieving significance in the increase of beta activities

marked as an insignificant epoch and eliminated from further analysis.

were chosen for subsequent BR calculation and dipole/source analysis.

**-2.5 2.5s** 

**-2.5 2.5s** 

**-2.5 2.5s** 

values larger than the median of baseline activities of the *th*

as an insignificant epoch and eliminated from further analysis.

Fig. 4. Example of single-trial epoch selection based on a nonparametric sign test. Single-trial VAMW*recon\_max*s of reconstructed data are examined through a nonparametric sign test. ( ) *Z i IOI*

number of time points in post-movement *IOI*. Only epochs showing significant increase of beta activities are chosen for further analysis. The first trial with a *ZIOI* score equal to -4.53 is marked

**2.3.8 Validation of coupled common spatial and temporal templates for single-trial** 

Since there are inevitably differences in head size and variations in head positions inside the MEG scanner among subjects, BR amplitude differences were compared using both individual spatial templates and the common spatial template. The use of a pair of common

*i* trial, and *NIOI* is the total number of time

*NIO I***<sup>+</sup> = 126**  *Z IOI***= -4.53** 

*IOI* = 0.76s ~2.1s *NIOI***=336**

*<sup>i</sup>* trial, ( ) *N i IOI* is the number of data points in post-movement *IOI* with

*NIO I+* **= 336**  *ZIOI* **= 18.41**  *i* trial, and *NIOI* is the total

×

*NIO I+* **= 331**  *ZIOI* **= 17.86**

*th*

than the median of baseline activities of the *th*

**70**

**(fT/cm)** 

**Amplitude** 

**0**

**60**

**(fT/cm)** 

**Amplitude** 

**0**

**60**

**(fT/cm)** 

**Amplitude** 

is the Z value of the *th*

**analysis** 

**0**

**2.3.7 Source estimation of the reactive beta activities** 

Based on the known spatial location and temporal expression in terms of spatial and temporal templates, reactive beta activities were successfully extracted. Figure 3a shows that IC 3 meets the dual criteria: i) the correlation values between spatial map and spatial template is 0.84 (rank= 97%, *Z*= 1.89, *P*=0.03); ii) correlation values of 16-20Hz VAMW*IC* and 20-24 Hz VAMW*IC* vs. VAMW*template* are 0.8 (rank= 99%, Z=3.08, P=0.01) and 0.78 (rank= 97.8%, Z= 2.85, P=0.022), respectively. Fig. 3a illustrates that noise could also be identified and removed. IC2 in Fig. 3a correlates highly (=0.88) in spatial distribution with the IC1 extracted from empty room measurements (Fig. 3b), and is therefore rejected.

Fig. 5a. Flow chart for creation of common spatial and temporal templates.

Extraction of Single-Trial Post-Movement MEG Beta

**fT/cm 60 0-2.5 2.5s**

**Left** 

**(a)** 

**(b)** 

**(b)** 

activation (one trial) in the left hemisphere (Fig. 6a and 6c).

**1500**

**Amplitude spectrum** 

**(fT/cm)** 

**0**

Synchronization in Normal and Parkinson's Patient Using ICA-Based Spatiotemporal Approach 99

Figure 6a depicts the single-trial VAMW*recon*s of subject I filtered within the trial-specific frequency band (Fig. 6b). The conventional AM method on the average of 100 epochs reveals a bilateral post-movement rebound pattern with contralateral (left hemisphere) dominance, whereas the current ICA-based single-trial analysis (one hemisphere template) yields only

Epoch acceptance rates were 84% (65/78), 89% (83/91), 71% (60/85), 73% (68/93), and 87% (76/87), respectively for the model generation group and 81% (71/88) for the validation subject; the average for all six was 80.8%. The *IOI*s of significance were 0.76s - 2.1s, 0.66s - 1.5s, 0.8s - 1.75s, 0.46s - 1.49s, and 0.71s – 1.28s for the five subjects in the model generation group, and 0.88s – 1.67s for the validation subject. Averaged magnitude of BR was calculated from the reconstructed data on trials that survived the epoch-selection procedure.

**Front** 

**A B**

**0 50 Hz** 

 **0.8s~1.8s -4s~-3s**

Trial-specific frequency band (16~18Hz)

**60**

**(fT/cm)** 

**Amplitude** 

**0**

**60**

**0**

 **(fT/cm)** 

**Amplitude** 

**-2.5 2.5s**

**B**

**-2.5 2.5s**

**A**

**Right** 

1. Compute **BR** of the **VAMW***recon\_max*.

2. Estimate source locations of **reactive beta activities**.

Fig. 5b. Flow chart for ICA-based single-trial analysis method.

**Determine the trial-specific frequency band:**

**Reconstructed data** 

during -2.5s to -2s.

**frequency band**.

**frequency band** to extract **the reactive beta activity**. **Trial-specific frequency band** 

No

**Step1:**Determine the **reactive frequency bands** at the nine sensor sites in the vicinity of SMI using the two-spectrum comparison method. **Step2:**Filter the reconstructed data of the nine SMI vicinal sensor sites using the

**Step3:**For each of nine sensor sites, compute the **event-related beta modulation** which is the difference between the maximum amplitude of **VAMW** in the post-movement (0.8s to 1.8s) interval and the mean amplitude of **VAMW**

**Step4:**Find the maximum **event-related beta modulation** among nine sensor sites and designate its reactive frequency band as the **trial-specific** 

**reactive frequency bands** and compute their **VAMWs**.

Input MEG epoch (**B**mxn with m channels and n time points)

**Step2:**Compute **VAMWs** of each **IC** with three standard beta bands, i.e. 12-16Hz, 16-20Hz and 20-24 Hz, denoted as **VAMW***IC***s.** The **VAMW***IC***s** are subsequently in conjunction with spatial map for

**Step3:**Select **ICs** for data reconstruction. Dual criteria should be fulfilled: 1) the correlation between the temporal template and at least one of three **VAMW***IC***s** should be higher than 95% (*Z*>1.63, *P*<0.05) among all ICs and 2) correlation between the spatial template and spatial map also

Apply the **nonparametric sign test** for **VAMW***recon\_max* to detect the increase of beta activity emerging from baseline activity. Epochs with significant increase (Z>1.63, P<0.05) in post-movement beta activity are selected for current source estimation and **BR** amplitude quantification.

Compute the **VAMW** of **reactive beta activity** in each sensor site and denote it by **VAMW***recon*. The largest **VAMW***recon* among the nine SMI vicinal sensor sites is denoted by **VAMW***recon\_max* and is subjected to

On each sensor site, filter the reconstructed data with the **trial-specific** 

1. Compute **BR** of the **VAMW***recon\_max*. 2. Estimate source locations of **reactive beta activities**.

Fig. 5b. Flow chart for ICA-based single-trial analysis method.

Yes

Good epoch?

should be higher than 95%(*Z*>1.63, *P*<0.05) among all ICs.

**Preprocessing:** Bandpass filtering (6 Hz~50 Hz)

**Select IC component for data reconstruction: Step1:**Compute the **IC**s using **FastICA**: *SUB*

task-related IC selection.

**(b)** 

*Single-trial epoch selection:*

*Compute* **VAMW***recon* and **VAMW***recon\_max* **:**

**single-trial epoch selection** and **BR** computation.

Figure 6a depicts the single-trial VAMW*recon*s of subject I filtered within the trial-specific frequency band (Fig. 6b). The conventional AM method on the average of 100 epochs reveals a bilateral post-movement rebound pattern with contralateral (left hemisphere) dominance, whereas the current ICA-based single-trial analysis (one hemisphere template) yields only activation (one trial) in the left hemisphere (Fig. 6a and 6c).

Epoch acceptance rates were 84% (65/78), 89% (83/91), 71% (60/85), 73% (68/93), and 87% (76/87), respectively for the model generation group and 81% (71/88) for the validation subject; the average for all six was 80.8%. The *IOI*s of significance were 0.76s - 2.1s, 0.66s - 1.5s, 0.8s - 1.75s, 0.46s - 1.49s, and 0.71s – 1.28s for the five subjects in the model generation group, and 0.88s – 1.67s for the validation subject. Averaged magnitude of BR was calculated from the reconstructed data on trials that survived the epoch-selection procedure.

Extraction of Single-Trial Post-Movement MEG Beta

**65** 

**(a)** 

**Trial Number** 

**1** 

**-2.5**

**0**

 **(fT/cm)**

*Single epoch*

*15 epochs averaged*

*25 epochs averaged*

**70**

**Magnetic Amplitude** 

**70**

**Magnetic Amplitude** 

**(b)** 

**-2.5**

**0**

**(fT/cm)** 

**70**

**Magnetic Amplitude** 

averaging method.

**-2.5**

**0**

**(fT/cm)**

Synchronization in Normal and Parkinson's Patient Using ICA-Based Spatiotemporal Approach 101

The ICA-based single-trial approach shows remarkable latency jittering and inter-trial variability throughout the whole measurement process. Both factors can result in attenuation and smearing of averaged movement-related MEG responses. Figure 7a shows the raster plot of sixty-five normalized single-trial VAMW*recon\_max*s which survived the

> **-2.5 2.5**  *Movement onset*

> > **2.5**

**2.5**

**2.5**

**Time (sec)** 

**-2.5**

**0**

**70**

**70**

*35 epochs averaged*

*45 epochs averaged*

*65 epochs averaged*

**(fT/cm)**

**Time (sec) Time (sec)**

**Magnetic Amplitude** 

**Magnetic Amplitude** 

**-2.5**

**0**

**(fT/cm)**

**70**

**Magnetic Amplitude** 

**Time (sec) Time (sec)**

**-2.5**

**0**

Fig. 7. Smearing of MEG profile and decrease of BR magnitude due to latency jittering. (a) Raster plot of normalized VAMW*recon\_max*s as sorted by the latency measured between the time of peak beta rebound and the movement onset. Black dashed line indicates movement onset time. (b) Latency jittering resulting in a smearing of the MEG profile and a decrease of BR magnitude when more VAMW*recon\_max*s are averaged, as is common in the conventional

**(fT/cm)**

**Time (sec) Time (sec)**

**1** 

Normalized Amplitude

**0** 

**2.5**

**2.5**

**2.5**

Fig. 6. Sensor-array display of VAMW*recon*s and VAMWs. (a) One example of ICA single-trial VAMW*recon*s of all sensor sites in subject I. The single-trial result shows only left sensorimotor area dominance of event-related activities, as the present study focuses on the area contralateral to movement side and only the left spatial template is used. The dashed trapezoid marks the nine SMI vicinal sensor sites and the VAMW*recon\_max* is marked with the red circle. (b) Trial-specific frequency band used for VAMW*recon*s calculation in Fig. 6a. (c) VAMWs obtained from the conventional averaging method over 100 trials in subject I. This figure shows a bilateral beta rebound pattern with contralateral (left hemisphere) dominance.

The BR amplitudes computed from individual spatial templates were 20.9±7.1 (mean±sd), 18.1±10.3, 16.2±6.2, 23.2±10.89, and 6.2±2.7 for the first 5 subjects, respectively, and 27.6±11.1 fT/cm for the 6th subject (Table 1). Using the common spatial template, BR amplitudes were 21.1±7.97, 19.02±9.7, 15.5±5.3, 19.75±8.75, 5.91±3.2, and 27.1±10.2 fT/cm, respectively (Table 1). There was no significant difference between the results obtained with two approaches (*p*=0.88; unpaired two-tailed t test). BR amplitudes obtained with the conventional method of averaging on 100 trials were 18.2, 7.254, 12.92, 16.4, 2.9, and 23.12 fT/cm, respectively. Means for single-trial ICA-derived BRs, using either individual or common spatial templates, were significantly higher than those obtained using the conventional method of averaging (*p*<0.005; Matched-pair Wilcoxon test; Table 1). The comparisons of BR amplitude and task-specific frequency band between ICA-based single-trial and conventional methods are given in Table 1.

**(c)** 

**Front**

**A B**

Fig. 6. Sensor-array display of VAMW*recon*s and VAMWs. (a) One example of ICA single-trial

sensorimotor area dominance of event-related activities, as the present study focuses on the area contralateral to movement side and only the left spatial template is used. The dashed trapezoid marks the nine SMI vicinal sensor sites and the VAMW*recon\_max* is marked with the red circle. (b) Trial-specific frequency band used for VAMW*recon*s calculation in Fig. 6a. (c) VAMWs obtained from the conventional averaging method over 100 trials in subject I. This

The BR amplitudes computed from individual spatial templates were 20.9±7.1 (mean±sd), 18.1±10.3, 16.2±6.2, 23.2±10.89, and 6.2±2.7 for the first 5 subjects, respectively, and 27.6±11.1 fT/cm for the 6th subject (Table 1). Using the common spatial template, BR amplitudes were 21.1±7.97, 19.02±9.7, 15.5±5.3, 19.75±8.75, 5.91±3.2, and 27.1±10.2 fT/cm, respectively (Table 1). There was no significant difference between the results obtained with two approaches (*p*=0.88; unpaired two-tailed t test). BR amplitudes obtained with the conventional method of averaging on 100 trials were 18.2, 7.254, 12.92, 16.4, 2.9, and 23.12 fT/cm, respectively. Means for single-trial ICA-derived BRs, using either individual or common spatial templates, were significantly higher than those obtained using the conventional method of averaging (*p*<0.005; Matched-pair Wilcoxon test; Table 1). The comparisons of BR amplitude and task-specific frequency band between ICA-based single-trial and conventional methods

VAMW*recon*s of all sensor sites in subject I. The single-trial result shows only left

figure shows a bilateral beta rebound pattern with contralateral (left hemisphere)

**45**

**(fT/cm)** 

**Amplitude** 

**0**

**45**

 **(fT/cm)** 

**Amplitude** 

**0**

**-2.5 2.5s** 

**B**

**-2.5 2.5s**

**A**

**Right** 

**Left** 

**(c)** 

dominance.

are given in Table 1.

**45 0 -2.5 2.5s**

**fT/cm**

The ICA-based single-trial approach shows remarkable latency jittering and inter-trial variability throughout the whole measurement process. Both factors can result in attenuation and smearing of averaged movement-related MEG responses. Figure 7a shows the raster plot of sixty-five normalized single-trial VAMW*recon\_max*s which survived the

Fig. 7. Smearing of MEG profile and decrease of BR magnitude due to latency jittering. (a) Raster plot of normalized VAMW*recon\_max*s as sorted by the latency measured between the time of peak beta rebound and the movement onset. Black dashed line indicates movement onset time. (b) Latency jittering resulting in a smearing of the MEG profile and a decrease of BR magnitude when more VAMW*recon\_max*s are averaged, as is common in the conventional averaging method.

Extraction of Single-Trial Post-Movement MEG Beta

**(a)** 

**(a)** 

**Left** 

**(b)** 

**(b)** 

**Top**

**Front** 

**Front Step =0.02** 

**Top** 

**Front** 

1

*k recon j j x x* 

**Sp atia l weigh t** 

**(normalized)** 

**Spatial weight** 

**Right** 

**Right** 

Fig. 8. Overlay of extracted reactive beta activities on MR image. (a) Spatial map

**Left** 

superposition of the isocontour spatial map on the segmented MRI brain. (d) Representative trace of reconstructed reactive beta activities in the vicinity of SMI. (e) Upper panels are isocontour maps of reconstructed neuromagnetic signals at 1202 ms post movement. Lower panels show that all dipoles (from 1202 to 1302 ms after movement onset as box-framed in

**Front** 

**(no rmalized)**

**Step= 0.02** 

**(c)** 

**(c)** 

**Left** 

**Left** 

**(f)** 

**Left** 

reconstructed using

**Front** 

Synchronization in Normal and Parkinson's Patient Using ICA-Based Spatiotemporal Approach 103

**Magnetic Amplitude** 

**(d)** 

**(d)** 

**Spatial Weight** 

**(normalized)** 

**Right** 

**(e)** 

**(e)** 

**Front**

**Front** 

**(fT/cm) 70** 

**Top**

**-70 -2.5 2.5 Time (sec)**

**Left** 

**Left** 

**Top Front**

**Top** 

**Front** 

(see Method Section). (b) and (c): Different views of

**Back** 

**12**

**0** 

**nAm** 

**Right** 

**Front**

1202 ~ 1302 ms

**Right** 

**Right** 

selection procedure for subject I, sorted by VAMW*recon\_max* peak latency as indexed to movement onset. The mean latency of peak beta rebound for the 65 trials was 1.410.43 s (meansd). With more epochs (random selection) averaged as with the conventional method of averaging, the averaged BR was attenuated (25.3, 24.6, 22.3, 21.5, 20.3 and 21.1 fT/cm for 1, 15, 25, 35, 45 and 65 trials averaged, respectively; values taken from the averaged VAMW*recon\_max*s using common spatial template) and the time-activity plots smeared (Fig. 7b).


Table 1. The comparison of BR amplitude and specific frequency bands for ICA-based single-trial and conventional methods.

Source estimation using ECD and MCE both showed a cluster of current sources centered (mean coordinates) in the anterior bank of the central sulcus (see Fig. 8e and 8f) on data points around the rebound peak of extracted reactive beta activities (see Fig. 8d, time interval between 1202ms – 1302ms of one single epoch of subject I). The ECD-located dipoles oscillate and span a sector. Furthermore, the center of MCE-estimated current sources (yellow dots) lies less than 2mm from the center of ECD-estimated dipoles (red dots) (see Fig. 8f). These results cross-verify the validity of the ICA-based single-trial method.

Figure 10 depicts the time-frequency plot of a normal subject and a Parkinson's disease patient at an MEG channel in the vicinity of left sensorimotor area. Clear suppression of post-movement ERS (red circle) and an attenuated ERD (yellow circle) are observed in the Parkinson's disease patient. The VAMWs was significantly larger both in alpha band and beta band in the normal subject than in the Parkinson's patient. These imply the slowed and reduced recovery after motor act may impede cortical preparation of the next movement.

selection procedure for subject I, sorted by VAMW*recon\_max* peak latency as indexed to movement onset. The mean latency of peak beta rebound for the 65 trials was 1.410.43 s (meansd). With more epochs (random selection) averaged as with the conventional method of averaging, the averaged BR was attenuated (25.3, 24.6, 22.3, 21.5, 20.3 and 21.1 fT/cm for 1, 15, 25, 35, 45 and 65 trials averaged, respectively; values taken from the averaged VAMW*recon\_max*s using common spatial template) and the time-activity plots

ICA based single-trial method Conventional AM method

(fT/cm)

Common spatial template

15.57±3.21~ 22.17±3.3

17.92±2.3~ 21.9±2 .7 2

16.8±2 .3 ~ 20.91±2.22

15.5±3 .3 ~ 19.2±2 .7 7

16.8±3 .1 ~ 21.2±2 .9

16.81±2.72~ 20.14±3.1

Task-specific frequency band

(Hz)

18.2 15 ~21

7.25 17 ~20

12.92 15 ~19

16.4 14 ~17

2.9 17 ~20

23.12 16 ~20

BR amplitude (fT/cm) Trial-specific frequency band (Hz) BR amplitude

21.22±2.44

22.18±3.12

20.49±2.3

20.7±3.08

20.77±3.67

19.94±2.68

Source estimation using ECD and MCE both showed a cluster of current sources centered (mean coordinates) in the anterior bank of the central sulcus (see Fig. 8e and 8f) on data points around the rebound peak of extracted reactive beta activities (see Fig. 8d, time interval between 1202ms – 1302ms of one single epoch of subject I). The ECD-located dipoles oscillate and span a sector. Furthermore, the center of MCE-estimated current sources (yellow dots) lies less than 2mm from the center of ECD-estimated dipoles (red dots) (see Fig. 8f). These results cross-verify the validity of the ICA-based single-trial

Figure 10 depicts the time-frequency plot of a normal subject and a Parkinson's disease patient at an MEG channel in the vicinity of left sensorimotor area. Clear suppression of post-movement ERS (red circle) and an attenuated ERD (yellow circle) are observed in the Parkinson's disease patient. The VAMWs was significantly larger both in alpha band and beta band in the normal subject than in the Parkinson's patient. These imply the slowed and reduced recovery after motor act may impede cortical preparation of the next

Table 1. The comparison of BR amplitude and specific frequency bands for ICA-based

Individual spatial template

Common spatial template

I 20.9±7.1 21.1±7.97 16.67±2.77 ~

II 18.1±10.3 19.02±9.7 18.04±2.62 ~

III 16.2±6.2 15.5±5.3 16.2±1.89 ~

IV 23.2±10.89 19.75±8.75 16.1±2.37 ~

V 6.2±2.7 5.91±3.2 17.31±3.23 ~

VI (validation) 27.6±11.1 27.1±10.2 16.32±2.83 ~

single-trial and conventional methods.

method.

movement.

smeared (Fig. 7b).

Subject index Individual spatial

template

Fig. 8. Overlay of extracted reactive beta activities on MR image. (a) Spatial map

reconstructed using 1 *k recon j j x x* (see Method Section). (b) and (c): Different views of

superposition of the isocontour spatial map on the segmented MRI brain. (d) Representative trace of reconstructed reactive beta activities in the vicinity of SMI. (e) Upper panels are isocontour maps of reconstructed neuromagnetic signals at 1202 ms post movement. Lower panels show that all dipoles (from 1202 to 1302 ms after movement onset as box-framed in

Extraction of Single-Trial Post-Movement MEG Beta

**ICA & Trial-specific frequency band Bandpass** 

**Task-specific frequency band Bandpass** 

> **ICA & Trial-specific frequency band Bandpass**

**Task-specific frequency band Bandpass** 

Synchronization in Normal and Parkinson's Patient Using ICA-Based Spatiotemporal Approach 105

**Subject I** 

Fig. 9. Comparisons of magnetic fields and source locations preprocessed with ICAbandpass trial specific (upper panel) and task-specific bandpass filtering (lower panel). (a) Neuromagnetic field maps. Data preprocessed with ICA-trial specific bandpass filter (15.57±3.21~22.17±3.3 Hz) gives a much less noisy neuromagnetic field pattern than that processed with the task-specific bandpass filtering method (15~21 Hz) (Pfurtscheller et al. 1999). Black vertical lines in the tracings of the left column denote time points of the

corresponding field maps in the right column. (b) Source localizations by ECD model. Only dipoles in post-movement IOI (interval-of-interest) with goodness-of-fit higher than 80% are accepted. The one with highest goodness-of-fit value out of each trial is rendered onto the subjects' 3D MRI surfaces. The estimated source positions preprocessed by ICA-bandpass filtering (upper panel) are (x, y, z)= (-454.45, -3.96.33, 80.73.63mm; goodness-of-fit= 97.53.7%) in subject I (65 trials) and (x, y, z)=(-35.33.5, 5.76.02, 88.75.61mm; goodnessof-fit= 96.93.7%) in subject VI (71 trials), whereas task specific bandpass filtering (lower panel) yields (x, y, z)=(-46.311.6, -9.9910.3, 84.516.7; goodness-of-fit=89.73.4%) in subject I (65 trials) and (x, y, z)=(-31.88.13, 0.514.01, 87.912.54mm; goodness-of-fit=87.24.5%) in subject VI (71 trials), respectively. The ICA-trial specific bandpass procedure yields better

**Subject VI**

**Subject I Subject VI**

(d)) are located in the primary motor area and oscillate accordingly. (f) The center of the MCE-estimated current sources (yellow dot) overlays the source location determined using the equivalent current dipole method (ECD) (red dot). Upper-left panel: coronal view. Upper-right panel: sagittal view. Lower-left panel: axial view. Lower-right panel: distribution of MCE estimated current sources.

Examining the single-trial variability using the proposed ICA-based method (Fig. 11), subtle dynamics of the beta rhythmic activities can be further studied. Figure 11 shows the ongoing trial-by-trial variabilities in amplitudes and latencies over 60 ICA de-noised post-movement ERS trials. With the utilization of ICA-based single-trial analysis, it is possible to investigate the reasoning of ERS vanishment is due to suppression of post-movement beta rebound in each single-trial, rather than the cause of temporal jittering and/or loss of synchronization. Even though the patient could perform lifting behavior well, his neuron activities show distinct sensorimotor patterns from normal subject, regardless of movement performance.

## **4. Discussion**

The movement-related oscillatory modulations (ERD/ERS of alpha, beta and gamma) have been reported to be spatially extended (Babiloni et al., 1999; Crone et al., 1998a; Crone et al., 1998b; Leocani et al., 1997; Neuper and Pfurtscheller, 2001; Salmelin and Hari, 1994a; Taniguchi et al., 2000; van Burik et al., 1998). Source localizations using conventional filtering have also been reported to disperse among several regions (Salmelin & Hari, 1994a). However, our results strongly indicate that proper treatment when trial-by-trial dynamics can be accounted for yields clustered localizations congruent to neuroanatomical representations.

The present ICA-based spatiotemporal approach for single-trial analysis study is dedicated to the extraction of neuromagnetic measurements of event-related beta oscillatory activities. One distinct feature of the current ICA-based method as compared with other single-trial approaches (Guger et al., 2000; Ioannides et al., 1993; Jung et al., 2001) is the simultaneous use of a spatial template and a temporal template for component selection. The spatial template provides a priori spatial information for brain signals, while the temporal template contains temporal characteristics of event-related responses. Using the paired criteria for component selection, identification specificity of task-related components for signal reconstruction is significantly improved. As shown in Fig. 3c, the inclusion of IC 9 with high spatial correlation (correlation value=0.61, rank= 95.2%, Z=1.67, P=0.048) but devoid of temporal congruence (correlation value =0.28, rank=13%, Z=0.34, P=0.87) causes beta BR to deteriorate from 28.9 fT/cm (red curve) to 18.6 fT/cm (blue curve). The ICA-preprocessed dataset yields cleaner field maps (Fig. 9a), which result in circumscribed localizations (Figs. 9b-9c and 9e-9f., Salmelin and Hari, 1994a).

Significantly, the current method also makes possible the analysis of the reactive frequency band for every single trial once task-related rhythmic activities are extracted. The conventional method discounts this subtle but potentially important information. Notwithstanding, the idea of using a fixed window for signal filtering is neurophysiologically not optimal. We emphasize the precise identification of reactive trial-specific frequencies for BR calculation, since taskrelated frequency modulation might exist in one or multiple bands (Pfurtscheller & Lopes da Silva, 1999). The three-standard frequency band procedure is used for generation of VAMW*IC*s to recover all possible task-related information and is followed by a two short-time spectra

Examining the single-trial variability using the proposed ICA-based method (Fig. 11), subtle dynamics of the beta rhythmic activities can be further studied. Figure 11 shows the ongoing trial-by-trial variabilities in amplitudes and latencies over 60 ICA de-noised post-movement ERS trials. With the utilization of ICA-based single-trial analysis, it is possible to investigate the reasoning of ERS vanishment is due to suppression of post-movement beta rebound in each single-trial, rather than the cause of temporal jittering and/or loss of synchronization. Even though the patient could perform lifting behavior well, his neuron activities show distinct sensorimotor patterns from normal subject, regardless of movement performance.

The movement-related oscillatory modulations (ERD/ERS of alpha, beta and gamma) have been reported to be spatially extended (Babiloni et al., 1999; Crone et al., 1998a; Crone et al., 1998b; Leocani et al., 1997; Neuper and Pfurtscheller, 2001; Salmelin and Hari, 1994a; Taniguchi et al., 2000; van Burik et al., 1998). Source localizations using conventional filtering have also been reported to disperse among several regions (Salmelin & Hari, 1994a). However, our results strongly indicate that proper treatment when trial-by-trial dynamics can be accounted for yields clustered localizations congruent to neuroanatomical

The present ICA-based spatiotemporal approach for single-trial analysis study is dedicated to the extraction of neuromagnetic measurements of event-related beta oscillatory activities. One distinct feature of the current ICA-based method as compared with other single-trial approaches (Guger et al., 2000; Ioannides et al., 1993; Jung et al., 2001) is the simultaneous use of a spatial template and a temporal template for component selection. The spatial template provides a priori spatial information for brain signals, while the temporal template contains temporal characteristics of event-related responses. Using the paired criteria for component selection, identification specificity of task-related components for signal reconstruction is significantly improved. As shown in Fig. 3c, the inclusion of IC 9 with high spatial correlation (correlation value=0.61, rank= 95.2%, Z=1.67, P=0.048) but devoid of temporal congruence (correlation value =0.28, rank=13%, Z=0.34, P=0.87) causes beta BR to deteriorate from 28.9 fT/cm (red curve) to 18.6 fT/cm (blue curve). The ICA-preprocessed dataset yields cleaner field maps (Fig. 9a), which result in circumscribed localizations (Figs.

Significantly, the current method also makes possible the analysis of the reactive frequency band for every single trial once task-related rhythmic activities are extracted. The conventional method discounts this subtle but potentially important information. Notwithstanding, the idea of using a fixed window for signal filtering is neurophysiologically not optimal. We emphasize the precise identification of reactive trial-specific frequencies for BR calculation, since taskrelated frequency modulation might exist in one or multiple bands (Pfurtscheller & Lopes da Silva, 1999). The three-standard frequency band procedure is used for generation of VAMW*IC*s to recover all possible task-related information and is followed by a two short-time spectra

(d)) are located in the primary motor area and oscillate accordingly. (f) The center of the MCE-estimated current sources (yellow dot) overlays the source location determined using the equivalent current dipole method (ECD) (red dot). Upper-left panel: coronal view. Upper-right panel: sagittal view. Lower-left panel: axial view. Lower-right panel:

distribution of MCE estimated current sources.

9b-9c and 9e-9f., Salmelin and Hari, 1994a).

**4. Discussion** 

representations.

Fig. 9. Comparisons of magnetic fields and source locations preprocessed with ICAbandpass trial specific (upper panel) and task-specific bandpass filtering (lower panel). (a) Neuromagnetic field maps. Data preprocessed with ICA-trial specific bandpass filter (15.57±3.21~22.17±3.3 Hz) gives a much less noisy neuromagnetic field pattern than that processed with the task-specific bandpass filtering method (15~21 Hz) (Pfurtscheller et al. 1999). Black vertical lines in the tracings of the left column denote time points of the corresponding field maps in the right column. (b) Source localizations by ECD model. Only dipoles in post-movement IOI (interval-of-interest) with goodness-of-fit higher than 80% are accepted. The one with highest goodness-of-fit value out of each trial is rendered onto the subjects' 3D MRI surfaces. The estimated source positions preprocessed by ICA-bandpass filtering (upper panel) are (x, y, z)= (-454.45, -3.96.33, 80.73.63mm; goodness-of-fit= 97.53.7%) in subject I (65 trials) and (x, y, z)=(-35.33.5, 5.76.02, 88.75.61mm; goodnessof-fit= 96.93.7%) in subject VI (71 trials), whereas task specific bandpass filtering (lower panel) yields (x, y, z)=(-46.311.6, -9.9910.3, 84.516.7; goodness-of-fit=89.73.4%) in subject I (65 trials) and (x, y, z)=(-31.88.13, 0.514.01, 87.912.54mm; goodness-of-fit=87.24.5%) in subject VI (71 trials), respectively. The ICA-trial specific bandpass procedure yields better

Extraction of Single-Trial Post-Movement MEG Beta

possibility for future profound study of subtle brain dynamics.

disclose hitherto unexplored mechanisms underlying these phenomena.

A concern with any data driven method is that prominent artifacts or noise can be intermingled with task-specific information (Ermer et al., 2000; Lins et al., 1993a; Lins et al., 1993b). However, previous ICA reports (Makeig et al., 2002; Mckeown & Radtke, 2001) indicate that brain rhythmic signals generated from different sources usually have their own oscillatory frequencies with distinct phases and are located in specific brain regions with patterns that are distinct from artifacts or noise (see also Fig. 3). This endorses the feasibility of using ICA to separate targeted rhythmic signals from irrelevant ones. The high epochacceptance rate (~80%) can be attributed to an improved SNR as compared to other studies on single-trial approaches to sensorimotor oscillatory activities (Brovelli et al., 2002; Wolpaw and McFarland, 1994). For instance, the spatial map of IC2 in Fig. 2a correlates highly (0.88) with the spatial map of IC1 from empty room measurement as shown in Fig. 3b; this

Synchronization in Normal and Parkinson's Patient Using ICA-Based Spatiotemporal Approach 107

comparison procedure (Pfurtscheller & Lopes da Silva, 1999) for the identification of the optimal reactive trial-specific frequency band in the reconstructed epochs. The present approach not only extracts the specific reactive frequencies but also retains phase information on a trial-by-trial basis. The trial-specific frequency band of post-movement beta modulation anchors mainly (~85% of all trials) in the lower beta band (16Hz~20Hz) and less frequently (~15%) in the higher beta band (20~24Hz). Great variation of BR values is also seen, as reflected in large SD (Table 1). The revealed trial-by-trial dynamics provide a

It is noteworthy that not all the data reconstructed from the selected ICs survives the statistical threshold. We have carefully monitored online and thoroughly checked offline the EMG measurements in terms of EMG onset (*p*=0.61, unpaired two-tailed t- test), termination (*p*=0.53, unpaired two-tailed *t*-test) and the EMG duration (*p*=0.573, unpaired two-tailed ttest) during finger lifting between significant and insignificant trials as indexed to the movement registration by the optic pad (Abbink et al., 1998). The data indicate an absence of prominent behavioral difference commensurate to the differential neuromagnetic responses. Some epochs with a fluctuating baseline, e.g., non-task-related spontaneous bursts of beta oscillatory activities, may manifest high baseline activity, which in turn results in a decrease in BR readout leading to exclusion after statistical manipulation (Fig 4). It has been suggested that baseline spontaneous activities may carry important information relevant to attention level, wakefulness, task difficulty, etc. (Buser & Rougeul-Buser, 1999; Sterman, 1999). The jittering of the neuromagnetic beta ERS is likewise interesting and may be also physiological. A zero-phase Butterworth filter was used to bandpass filter the raw data. The symmetric property of the zero-phase filter means that processed signals have precisely zero phase distortion and therefore no time shift of peak beta rebound was introduced. Hence, fluctuations of significance level and the jittering of central processing despite similar behavioral performance may be ascribed to the subject's variant cognitive states or the degree of training (Buser & Rougeul-Buser, 1999; Sterman, 1999; Flotzinger et al., 1992; Wolpaw et al., 1994; Bastiaansen et al., 2001; Bastiaansen et al., 1999; Earle, 1988; Haig et al., 1995; Hoffman et al., 1991; Yabe et al., 1993). The exploration of underlying mechanisms mandates more meticulous designs in the future. Using the conventional method of averaging, certain diseases, such as Parkinson's and Unverricht-Lundborg myoclonic epilepsy, have been observed to show either attenuated, prolonged or abolished ERS responses (Silen et al., 2000; Tamas et al., 2003). Such cases can be further examined using the current ICA-based single-trial method for the time course and trial-by-trial dynamics to

results in terms of much focused source locations and higher goodness-of-fit. x, y, and z denote the dipole location in the head coordinate system as anchored by the HPI (head position indicator) coils. The x-axis passes through the preauricular points, pointing to the right; the positive y-axis traverses the nasion and is normal to the x-axis; the positive z-axis points upward and normal to the xy-plane.

Fig. 10. The comparion of neural activity via time-frequency, VAMW and spectrum analysis obtained from one normal subject and a parkinson's disease patient. The ensemble averaging results reflect the ERD attenuation (yellow circle) and the ERS disappearance (red circle) in the pakinson's patient.

Fig. 11. Trial-by-trial comparison of VAMWs could be performed via the proposed ICAbased approach between the normal subject and the Parkinson's disease patient.

Fig. 10. The comparion of neural activity via time-frequency, VAMW and spectrum analysis

averaging results reflect the ERD attenuation (yellow circle) and the ERS disappearance (red

**100 fT/c m** 

**Trial number** 

*Movemen t onset Movemen t onset* 

**Time (sec) Time (sec)** 

**Magnetic Amplitude** 

Fig. 11. Trial-by-trial comparison of VAMWs could be performed via the proposed ICAbased approach between the normal subject and the Parkinson's disease patient.

obtained from one normal subject and a parkinson's disease patient. The ensemble

**Normal subject HT PD patient SH** 

results in terms of much focused source locations and higher goodness-of-fit. x, y, and z denote the dipole location in the head coordinate system as anchored by the HPI (head position indicator) coils. The x-axis passes through the preauricular points, pointing to the right; the positive y-axis traverses the nasion and is normal to the x-axis; the positive z-axis

points upward and normal to the xy-plane.

circle) in the pakinson's patient.

**100 fT/cm** 

**Trial number** 

**Magnetic Amplitude** 

comparison procedure (Pfurtscheller & Lopes da Silva, 1999) for the identification of the optimal reactive trial-specific frequency band in the reconstructed epochs. The present approach not only extracts the specific reactive frequencies but also retains phase information on a trial-by-trial basis. The trial-specific frequency band of post-movement beta modulation anchors mainly (~85% of all trials) in the lower beta band (16Hz~20Hz) and less frequently (~15%) in the higher beta band (20~24Hz). Great variation of BR values is also seen, as reflected in large SD (Table 1). The revealed trial-by-trial dynamics provide a possibility for future profound study of subtle brain dynamics.

It is noteworthy that not all the data reconstructed from the selected ICs survives the statistical threshold. We have carefully monitored online and thoroughly checked offline the EMG measurements in terms of EMG onset (*p*=0.61, unpaired two-tailed t- test), termination (*p*=0.53, unpaired two-tailed *t*-test) and the EMG duration (*p*=0.573, unpaired two-tailed ttest) during finger lifting between significant and insignificant trials as indexed to the movement registration by the optic pad (Abbink et al., 1998). The data indicate an absence of prominent behavioral difference commensurate to the differential neuromagnetic responses. Some epochs with a fluctuating baseline, e.g., non-task-related spontaneous bursts of beta oscillatory activities, may manifest high baseline activity, which in turn results in a decrease in BR readout leading to exclusion after statistical manipulation (Fig 4). It has been suggested that baseline spontaneous activities may carry important information relevant to attention level, wakefulness, task difficulty, etc. (Buser & Rougeul-Buser, 1999; Sterman, 1999). The jittering of the neuromagnetic beta ERS is likewise interesting and may be also physiological. A zero-phase Butterworth filter was used to bandpass filter the raw data. The symmetric property of the zero-phase filter means that processed signals have precisely zero phase distortion and therefore no time shift of peak beta rebound was introduced. Hence, fluctuations of significance level and the jittering of central processing despite similar behavioral performance may be ascribed to the subject's variant cognitive states or the degree of training (Buser & Rougeul-Buser, 1999; Sterman, 1999; Flotzinger et al., 1992; Wolpaw et al., 1994; Bastiaansen et al., 2001; Bastiaansen et al., 1999; Earle, 1988; Haig et al., 1995; Hoffman et al., 1991; Yabe et al., 1993). The exploration of underlying mechanisms mandates more meticulous designs in the future. Using the conventional method of averaging, certain diseases, such as Parkinson's and Unverricht-Lundborg myoclonic epilepsy, have been observed to show either attenuated, prolonged or abolished ERS responses (Silen et al., 2000; Tamas et al., 2003). Such cases can be further examined using the current ICA-based single-trial method for the time course and trial-by-trial dynamics to disclose hitherto unexplored mechanisms underlying these phenomena.

A concern with any data driven method is that prominent artifacts or noise can be intermingled with task-specific information (Ermer et al., 2000; Lins et al., 1993a; Lins et al., 1993b). However, previous ICA reports (Makeig et al., 2002; Mckeown & Radtke, 2001) indicate that brain rhythmic signals generated from different sources usually have their own oscillatory frequencies with distinct phases and are located in specific brain regions with patterns that are distinct from artifacts or noise (see also Fig. 3). This endorses the feasibility of using ICA to separate targeted rhythmic signals from irrelevant ones. The high epochacceptance rate (~80%) can be attributed to an improved SNR as compared to other studies on single-trial approaches to sensorimotor oscillatory activities (Brovelli et al., 2002; Wolpaw and McFarland, 1994). For instance, the spatial map of IC2 in Fig. 2a correlates highly (0.88) with the spatial map of IC1 from empty room measurement as shown in Fig. 3b; this

Extraction of Single-Trial Post-Movement MEG Beta

Synchronization in Normal and Parkinson's Patient Using ICA-Based Spatiotemporal Approach 109

method blindly decomposes the MEG epochs (**B**) into a spatially distributed map (**U**) multiplied by temporal signals (**S**), i.e. **B**=**US**, on the basis of independency among sources (Vigario & Oja, 2000), whereas SSP mandates a pre-defined spatial filter (**U\_sf**) for recovering signals (**S**), i.e. **S**=**U\_sf**+**B**, where + denotes pseudo inverse, based on orthogonal projection. When ambient noise and the spatial filter are not mutually orthogonal, the SSP has difficulty in resolving the two. Subsequent application of ICA following SSP does not ensure finer signal extraction or further noise removal since the data recovered from SSP are already linear mixtures of components out of a pre-defined spatial filter, which is a constraint drag

Left and right sensorimotor rhythms can be decomposed into two distinct ICs (IC3 and IC5 in Fig. 3), implying possible independent modulatory mechanisms between the two hemispheres. This view is corroborated by an event-related coherence study (Andrew and Pfurtscheller, 1999) that reports a lack of interhemispheric coherence in human postmovement beta activities. Movement-related beta oscillatory activities of the right hemisphere can be extracted in the same way using spatial and temporal templates for right sensorimotor rhythm. The source locations for extracted right hemispheric beta activities were mainly in the right premotor area (data not shown), which agrees with previous studies (Brovelli et al., 2002; Ilmoniemi, R. J., 1991). Event-related beta activities in SMA and posterior parietal cortical areas (Brovelli et al., 2002; Joliot et al., 1999) are not observed in our data, possibly due to the fact that the contributing sources here are radial in orientation

The agreement between the values of BR amplitude obtained with the common spatial/temporal templates and the individually generated ones (Table 1) promises a flexibility in both experimental design and analytical strategy. The proposed ICA-based spatiotemporal approach for single trial analysis can also be applied on fewer trials (Fig 7b), which is a great advantage over conventional methods. Given meticulous head positioning (see above the Method Section), common spatial and temporal templates can be used to extract pertinent movement-related neuromagnetic signals from subjects, which may shorten the overall time needed to run an experiment. We have no preference for the use of a grand averaged template over individual ones. On the contrary, the use of an individual template is suggested for any profound individual–based ERD/ERS study. However, the feasibility of using a grand averaged template provides an effective alternative in cases where lengthy procedures cannot be endured by the participants. This is particularly true for clinical settings where patients have attention problems or are incapable of sustaining long experiments so that individual templates cannot be optimally obtained. Nevertheless, caution should be exercised when applying the current ICA-based single-trial method for clinical studies. For patients whose heads cannot be properly positioned in the center of the MEG helmet, the use of a common spatial template may fail, making a customized individual spatial template mandatory for IC selection. For patients whose motor performance deviates significantly from normal, e.g., victims of motor stroke or severe movement disorders, the use of the common temporal template might not be justified since the time courses of event-related brain activities may be significantly altered due to primary deficit or secondary plasticity. Accordingly, in such situations, an individual spatial template can be applied without a temporal template as an aid to component selection. Our future investigations will combine the current dual-template approach with a source estimation method so that a spatial filter of better precision and higher dimensions can be

on the optimal performance of ICA designed for blind decomposition.

and thus could not be optimally detected by MEG (Salmelin and Hari, 1994b).

suggests that the neuromagnetic signal IC2, deselected for subsequent processing, can be accounted for by background noise in the shielding room. IC11 in Fig. 3a has a stationary cycle around 1.2 Hz, and its spatial map has higher weights at the outer rim of the MEG sensor array, which suggests a plausible connection with cardiac cycles. It was also observed (Fig. 3) that rhythmic activities in left and right SMIs as well as the occipital areas could be extracted into separate ICs that can be reminiscent of various mechanisms and time courses of different brain oscillatory activities (Pfurtscheller & Lopes da Silva, 1999; Pfurtscheller et al., 1997; Pfurtscheller et al., 1998b; Stancak and Pfurtscheller, 1996a; Stancak & Pfurtscheller, 1996b; Andrew & Pfurtscheller, 1999).

Since most task irrelevant signals, e.g., internal and external noises, can be removed by proper de-selection of ICs, it is possible to reconstitute the representative spatial map of all

contributing ICs using 1 *k recon j j x x* , in which *recon <sup>x</sup>* is the reconstructed spatial map, *k* is the

number of selected ICs and *<sup>i</sup> x* is the spatial map of the *i*th selected IC in Fig. (4) (Fig. 8a). This spatial map of reconstructed signals, which is a topographical distribution of weighting factors on the sensor array, can be overlaid with the segmented MRI brain (Fig. 8b & 8c; ASA program, ANT Software, Dutch). The highest weight is shown to project over the SMI area, which demonstrates that the high SNR of the ICA-extracted rhythmic activities of each trial has made possible the use on single-trial data of source estimation methods that require high SNR on input data for processing, e.g., the equivalent current dipole technique (ECD), minimum current estimation (MCE), and minimum norm estimation (MNE) (Delorme et al., 2001; Delorme et al., 2002; Jung et al., 2001; Makeig et al., 1997; Mckeown et al., 2001). Conventionally, these estimation methods exploit averaged data out of a large amount of trials.

Another reason why the intricate phase-unlocked signal can be preserved is the fact that no averaging procedure is needed; such a procedure would otherwise inherently distort the embedded information. Accordingly, as shown in Fig. 8d, source modeling with a moving dipole on a msec by msec basis on the reconstructed oscillatory beta signals during the rebound period (Brovelli et al., 2002) of a single-trial epoch results in a focused clustering of dipole foci at the pre-central area, i.e., the primary motor cortex (Fig. 8e). Figure 8 shows the result of MCE modeling (Uutela et al, 1999), where the center of MCE-estimated current sources (yellow dot) is very close (< 2mm distance) to the dipole location as estimated using the ECD approach (red dot).

It can be argued that one can first localize the generator area and then build a spatial filter for extracting single-trial data so that the subsequent analysis can be conducted on the source level instead of the sensor level. One premise and justification of using a source-areagenerated spatial filter is that the source area can be precisely localized for the generation of a spatial filter (Tesche et al., 1995). The very first step is to filter the signals to obtain a presupposed reactive frequency band. However, using conventional simple filtering techniques, ambient noise with ~20Hz components cannot be optimally removed, and this will cause localization uncertainty for the probed sources (Fig 9). However, ICA preprocessing decomposes the compound neuromagnetic signals into various independent task-related and task-unrelated/noise components so that ~20Hz activities not related to the a priori spatiotemporal profile will not confound the selected ones. Furthermore, our ICAbased method differs from other spatial filtering techniques, e.g., signal space projection (SSP) which is a fixed spatial filter for signal extraction (Tesche et al., 1995). The ICA-based

suggests that the neuromagnetic signal IC2, deselected for subsequent processing, can be accounted for by background noise in the shielding room. IC11 in Fig. 3a has a stationary cycle around 1.2 Hz, and its spatial map has higher weights at the outer rim of the MEG sensor array, which suggests a plausible connection with cardiac cycles. It was also observed (Fig. 3) that rhythmic activities in left and right SMIs as well as the occipital areas could be extracted into separate ICs that can be reminiscent of various mechanisms and time courses of different brain oscillatory activities (Pfurtscheller & Lopes da Silva, 1999; Pfurtscheller et al., 1997; Pfurtscheller et al., 1998b; Stancak and Pfurtscheller, 1996a; Stancak &

Since most task irrelevant signals, e.g., internal and external noises, can be removed by proper de-selection of ICs, it is possible to reconstitute the representative spatial map of all

This spatial map of reconstructed signals, which is a topographical distribution of weighting factors on the sensor array, can be overlaid with the segmented MRI brain (Fig. 8b & 8c; ASA program, ANT Software, Dutch). The highest weight is shown to project over the SMI area, which demonstrates that the high SNR of the ICA-extracted rhythmic activities of each trial has made possible the use on single-trial data of source estimation methods that require high SNR on input data for processing, e.g., the equivalent current dipole technique (ECD), minimum current estimation (MCE), and minimum norm estimation (MNE) (Delorme et al., 2001; Delorme et al., 2002; Jung et al., 2001; Makeig et al., 1997; Mckeown et al., 2001). Conventionally, these estimation methods exploit averaged data out of a large amount of

Another reason why the intricate phase-unlocked signal can be preserved is the fact that no averaging procedure is needed; such a procedure would otherwise inherently distort the embedded information. Accordingly, as shown in Fig. 8d, source modeling with a moving dipole on a msec by msec basis on the reconstructed oscillatory beta signals during the rebound period (Brovelli et al., 2002) of a single-trial epoch results in a focused clustering of dipole foci at the pre-central area, i.e., the primary motor cortex (Fig. 8e). Figure 8 shows the result of MCE modeling (Uutela et al, 1999), where the center of MCE-estimated current sources (yellow dot) is very close (< 2mm distance) to the dipole location as estimated using

It can be argued that one can first localize the generator area and then build a spatial filter for extracting single-trial data so that the subsequent analysis can be conducted on the source level instead of the sensor level. One premise and justification of using a source-areagenerated spatial filter is that the source area can be precisely localized for the generation of a spatial filter (Tesche et al., 1995). The very first step is to filter the signals to obtain a presupposed reactive frequency band. However, using conventional simple filtering techniques, ambient noise with ~20Hz components cannot be optimally removed, and this will cause localization uncertainty for the probed sources (Fig 9). However, ICA preprocessing decomposes the compound neuromagnetic signals into various independent task-related and task-unrelated/noise components so that ~20Hz activities not related to the a priori spatiotemporal profile will not confound the selected ones. Furthermore, our ICAbased method differs from other spatial filtering techniques, e.g., signal space projection (SSP) which is a fixed spatial filter for signal extraction (Tesche et al., 1995). The ICA-based

is the reconstructed spatial map, *k* is the

is the spatial map of the *i*th selected IC in Fig. (4) (Fig. 8a).

Pfurtscheller, 1996b; Andrew & Pfurtscheller, 1999).

1

, in which *recon <sup>x</sup>*

*k recon j j x x* 

contributing ICs using

trials.

number of selected ICs and *<sup>i</sup> x*

the ECD approach (red dot).

method blindly decomposes the MEG epochs (**B**) into a spatially distributed map (**U**) multiplied by temporal signals (**S**), i.e. **B**=**US**, on the basis of independency among sources (Vigario & Oja, 2000), whereas SSP mandates a pre-defined spatial filter (**U\_sf**) for recovering signals (**S**), i.e. **S**=**U\_sf**+**B**, where + denotes pseudo inverse, based on orthogonal projection. When ambient noise and the spatial filter are not mutually orthogonal, the SSP has difficulty in resolving the two. Subsequent application of ICA following SSP does not ensure finer signal extraction or further noise removal since the data recovered from SSP are already linear mixtures of components out of a pre-defined spatial filter, which is a constraint drag on the optimal performance of ICA designed for blind decomposition.

Left and right sensorimotor rhythms can be decomposed into two distinct ICs (IC3 and IC5 in Fig. 3), implying possible independent modulatory mechanisms between the two hemispheres. This view is corroborated by an event-related coherence study (Andrew and Pfurtscheller, 1999) that reports a lack of interhemispheric coherence in human postmovement beta activities. Movement-related beta oscillatory activities of the right hemisphere can be extracted in the same way using spatial and temporal templates for right sensorimotor rhythm. The source locations for extracted right hemispheric beta activities were mainly in the right premotor area (data not shown), which agrees with previous studies (Brovelli et al., 2002; Ilmoniemi, R. J., 1991). Event-related beta activities in SMA and posterior parietal cortical areas (Brovelli et al., 2002; Joliot et al., 1999) are not observed in our data, possibly due to the fact that the contributing sources here are radial in orientation and thus could not be optimally detected by MEG (Salmelin and Hari, 1994b).

The agreement between the values of BR amplitude obtained with the common spatial/temporal templates and the individually generated ones (Table 1) promises a flexibility in both experimental design and analytical strategy. The proposed ICA-based spatiotemporal approach for single trial analysis can also be applied on fewer trials (Fig 7b), which is a great advantage over conventional methods. Given meticulous head positioning (see above the Method Section), common spatial and temporal templates can be used to extract pertinent movement-related neuromagnetic signals from subjects, which may shorten the overall time needed to run an experiment. We have no preference for the use of a grand averaged template over individual ones. On the contrary, the use of an individual template is suggested for any profound individual–based ERD/ERS study. However, the feasibility of using a grand averaged template provides an effective alternative in cases where lengthy procedures cannot be endured by the participants. This is particularly true for clinical settings where patients have attention problems or are incapable of sustaining long experiments so that individual templates cannot be optimally obtained. Nevertheless, caution should be exercised when applying the current ICA-based single-trial method for clinical studies. For patients whose heads cannot be properly positioned in the center of the MEG helmet, the use of a common spatial template may fail, making a customized individual spatial template mandatory for IC selection. For patients whose motor performance deviates significantly from normal, e.g., victims of motor stroke or severe movement disorders, the use of the common temporal template might not be justified since the time courses of event-related brain activities may be significantly altered due to primary deficit or secondary plasticity. Accordingly, in such situations, an individual spatial template can be applied without a temporal template as an aid to component selection. Our future investigations will combine the current dual-template approach with a source estimation method so that a spatial filter of better precision and higher dimensions can be

Extraction of Single-Trial Post-Movement MEG Beta

Silva, 13-32, Elsevier, Amsterdam.

130-141.

Synchronization in Normal and Parkinson's Patient Using ICA-Based Spatiotemporal Approach 111

Bastiaansen, M. C. M.; Bocker, K. B. E. & Brunia, C. H. M. (2001). Event-related

Buser, P. & Rougeul-Buser A. (1999). EEG synchronization in cat, monkey and human

Clochon, P.; Fontbonne, J. M. & Etevenon, P. (1996). A new method for quantifying EEG

Delorme, A. & Makeig, S. (2003). EEG changes accompanying learned regulation of 12-Hz EEG activity, *IEEE transactions on rehabilitation engineering*, Vol. 11, pp. 133-137. Earle, J. B. (1988). Task difficulty and EEG alpha asymmetry: an amplitude and frequency

Ermer, J. J.; Mosher, J. C.; Huang, M. & Leahy, R. M. (2000). Paired MEG data set source

Florian, G. & Pfurtscheller, G. (1995). Dynamic spectral analysis of event-related EEG data, *Electroencephalography and Clinical Neurophysiology*, Vol. 95, pp. 393-396. Flotzinger, D.; Kalcher, J. & Pfurtscheller, G. (1992). EEG classification by learning vector

Guger, C.; Ramoser, H. & Pfurtscheller, G. (2000). Real-time EEG analysis with subject-

Haig, A. R.; Gordon, E.; Rogers, G., & Anderson, J. (1995). Classification of single-trial ERP

Hyvarinen, A.; Karhunen, J. & Oja, E. (2001). *Independent component analysis.* John Wiley and

Hyvarinen, A & Oja, E. (1997). A fast fixed-point algorithm for independent component

Ilmoniemi, R. J. (1991). Estimates of neuronal current distributions, *Acta Otolaryngol*, Vol.

Ioannides, A. A.; Singh, K. D.; Hasson, R.; Baumann, S. B.; Rogers, R. L.; Guinto, F. C. &

Papanicolaou, A. C. (1993). Comparison of current dipole and magnetic field tomography analyses of the cortical response to auditory stimuli, *Brain Topography*,

localization using recursively applied and projected (RAP) MUSIC, *IEEE* 

specific spatial patterns for a brain-computer interface (BCI). *IEEE Transactions on* 

sub-types: application of globally optimal vector quantization using simulated annealing, *Electroencephalography and Clinical Neurophysiology*, Vol. 94, pp. 288-297. Hoffman, R. E.; Buchsbaum, M. S.; Escobar, M. D.; Makuch, R. W.; Nuechterlein, K. H. &

Guich, S. M. (1991). EEG coherence of prefrontal areas in normal and schizophrenic males during perceptual activation, *Journal of Neuropsychiatry and Clinical* 

*Electroencephalography and Clinical Neurophysiology*, Vol. 98, pp. 126-129. Delorme, A.; Makeig, S.; Fabre-Thorpe, M. & Sejnowske, T. 2001. From single-trial EEG to

Brain Area dynamics, *Neurocomputing*, Vol. 44, pp. 1057-1064.

*Transactions on Biomedical Engineering*, Vol. 47, pp. 1248-1260.

quantization, *Biomedizinische Technik*, Vol. 37, pp. 303-309.

analysis, *Neuropsychobiology*, Vol. 20, pp. 95-112.

*Rehabilitation Engineering*, Vol. 8, pp. 447-456.

analysis, *Neural Computation*, Vol. 9, pp. 1483-1492.

*Neurosciences*, Vol. 3, pp. 169-175.

Sons, New York.

Vol. 6, pp. 27-34.

49(Suppl.), pp. 80–87.

desynchronization during anticipatory attention for an upcoming stimulus: a comparative EEG/MEG study, *Clinical Neurophysiology*, Vol. 112, pp. 393-403. Brovelli, A.; Battaglini, P. P.; Naranjo, J. R. & Budai, R. (2002). Medium-range oscillatory

network and the 20-Hz sensorimotor induced potential, *Neuroimage*, Vol. 16, pp.

during attentive states: a brief survey, In: *Event-related desynchronization. Handbook of electroencephalography and clinical neurophysiology*, G. Pfurtscheller & F. H. Lopes da

event-related desynchronization: amplitude envelope analysis,

designed, which will make possible sophisticated analysis on the source level instead of the sensor level, eliminating the positioning problem.

Degeneration of the dopaminergic neurons in substania nigra pars compacta (SNc) in Parkinson's patients result in abnormal projection in thalamo-cortical pathway which causes an abnormal projection from thalamus to supplementary motor area (SMA). Pfurtscheller et al. (1998) also have demonstrated that Parkinson's patients have delayed ERD and abolished post-movement ERS and speculated there is dysfunction in subcortico-cortical connections in Parkinson's patients. In this study, we analyzed post-movement ERS in one Parkinson's patient. The present ICA-based approach may be helpful for disclosing the mechanism of movement-related brain rhythms which could be used as a clinical index for diagnosing Parkinson's patients.

## **5. Conclusions**

The present novel ICA-based spatiotemporal approach for single trial analysis features a paired-template matching for stringent component selection. The spatial template provides a priori spatial information for targeted brain signals while the temporal template contains temporal characteristics of event-related responses. The method promises not only a high extraction rate of post-movement beta synchronization but also better localization of the corresponding sources. Various source modeling methods commanding high SNR can now be applied to single trial data as extracted using the ICA-spatiotemporal procedure. Our method takes into account subtle trial-by-trial dynamics. The reconstructed MEG brain signals per trial unravel the temporal information and inter-trial variations of reactive oscillatory activities, which in turn may shed light on the subtle dynamics of brain processing. The embodied common template approach permits an effective alternative in cases where lengthy procedures cannot be endured by the participants or in clinical settings where patients have attention problems or are incapable of sustaining long experiments.

## **6. Acknowledgment**

This study was funded by the National Central University, Center for Dynamical Biomarkers and Translational Medicine (99-2911-1-008-100), National Science Council (99- 2628- E-008-003, 99-2628- E-008-012, 100-2628-E-008-013, 100-2628-E-008-001, 100-2613-E-008- 006-D), and Veterans General Hospital University System of Taiwan Joint Research Program (VGHUST96-P4-15, VGHUST97-P3-11, VGHUST98-98- P3-09).

## **7. References**


designed, which will make possible sophisticated analysis on the source level instead of the

Degeneration of the dopaminergic neurons in substania nigra pars compacta (SNc) in Parkinson's patients result in abnormal projection in thalamo-cortical pathway which causes an abnormal projection from thalamus to supplementary motor area (SMA). Pfurtscheller et al. (1998) also have demonstrated that Parkinson's patients have delayed ERD and abolished post-movement ERS and speculated there is dysfunction in subcortico-cortical connections in Parkinson's patients. In this study, we analyzed post-movement ERS in one Parkinson's patient. The present ICA-based approach may be helpful for disclosing the mechanism of movement-related brain rhythms which could be used as a clinical index for diagnosing

The present novel ICA-based spatiotemporal approach for single trial analysis features a paired-template matching for stringent component selection. The spatial template provides a priori spatial information for targeted brain signals while the temporal template contains temporal characteristics of event-related responses. The method promises not only a high extraction rate of post-movement beta synchronization but also better localization of the corresponding sources. Various source modeling methods commanding high SNR can now be applied to single trial data as extracted using the ICA-spatiotemporal procedure. Our method takes into account subtle trial-by-trial dynamics. The reconstructed MEG brain signals per trial unravel the temporal information and inter-trial variations of reactive oscillatory activities, which in turn may shed light on the subtle dynamics of brain processing. The embodied common template approach permits an effective alternative in cases where lengthy procedures cannot be endured by the participants or in clinical settings where patients have attention problems or are incapable of sustaining long experiments.

This study was funded by the National Central University, Center for Dynamical Biomarkers and Translational Medicine (99-2911-1-008-100), National Science Council (99- 2628- E-008-003, 99-2628- E-008-012, 100-2628-E-008-013, 100-2628-E-008-001, 100-2613-E-008- 006-D), and Veterans General Hospital University System of Taiwan Joint Research Program

Andrew, C. & Pfurtscheller G. (1995). Event-related coherence during finger movement: a

Andrew, C. & Pfurtscheller, G. (1999). Lack of bilateral coherence of post-movement central

Bastiaansen, M. C. M.; Bocker, K. B. E.; Cluitmans, P. J. M. & Brunia, C. H. M. (1999). Event-

knowledge of results, *Clinical Neurophysiology*, Vol. 110, pp. 250-260.

beta oscillations in the human electroencephalography, *Neuroscience letters*, Vol.

related desynchronization related to the anticipation of a stimulus providing

(VGHUST96-P4-15, VGHUST97-P3-11, VGHUST98-98- P3-09).

pilot study, *Biomedizinische Technik*, Vol. 40, pp. 326-332.

sensor level, eliminating the positioning problem.

Parkinson's patients.

**6. Acknowledgment** 

**7. References** 

273, pp. 82-89.

**5. Conclusions** 


Clochon, P.; Fontbonne, J. M. & Etevenon, P. (1996). A new method for quantifying EEG event-related desynchronization: amplitude envelope analysis, *Electroencephalography and Clinical Neurophysiology*, Vol. 98, pp. 126-129.


Extraction of Single-Trial Post-Movement MEG Beta

*Neurophysiology*, Vol. 42, pp. 817-826.

*Neurophysiology*, Vol. 98, pp. 281-293.

*Clinical Neurophysiology*, Vol. 103, pp. 642-651.

*Clinical Neurophysiology*, Vol. 15, pp. 243-250.

*Neurophysiology*, Vol. 109, pp. 154-160.

*Measurement,* Vol. 22, pp. 121-130.

39, pp. 541-557.

110, pp. 787-798.

pp. 250-258.

43-46.

Synchronization in Normal and Parkinson's Patient Using ICA-Based Spatiotemporal Approach 113

Mosher, J. C.; Lewis, P. S. & Leahy, R. M. (1992). Multiple dipole modeling and localization

Muller-Gerking, J.; Pfurtscheller, G. & Flyvbjerg, H. (1999). Designing optimal spatial filters

Pfurtscheller, G. & Aranibar A. 1(977). Event-related cortical desynchronization detected by

Pfurtscheller, G. & Aranibar A. (1979a). Evaluation of event-related desynchronization

*Electroencephalography and Clinical Neurophysiology*, Vol. 46, pp. 138-146. Pfurtscheller, G. & Aranibar A. (1979b). Evaluation of event-related desynchronization

*Electroencephalography and Clinical Neurophysiology*, Vol. 46, pp. 138-146. Pfurtscheller, G. (1981). Central beta rhythm during sensorimotor activities in man, *Electroencephalography and Clinical Neurophysiology*, Vol. 51, pp. 253-264. Pfurtscheller, G. & Berghold, A. (1989). Patterns of cortical activation during planning of

Pfurtscheller, G.; Pregenzer M. & Neuper C. (1994). Visualization of sensorimotor areas

Pfurtscheller, G.; Stancak A., Jr. & Neuper C. (1996). Post-movement beta synchronization: a

Pfurtscheller, G.; Neuper C.; Flotzinger D. & Pregenzer M. (1997). EEG-based discrimination

Pfurtscheller, G.; Pichler-Zalaudek K.; Ortmayr B.; Diez J. & Reisecker F. (1998a).

Pfurtscheller, G.; Zalaudek K. & Neuper C. (1998b). Event-related beta synchronization after

Pfurtscheller, G. & Lopes da Silva F. H. (1999). Event-related EEG/MEG synchronization and desynchronization: basic principles, *Clinical Neurophysiology*, Vol. 110, pp. 1842-1857. Rosell, J.; Casanas, R. & Scharfetter H. (2001). Sensitivity maps and system requirements for

Salmelin, R. & Hari, R. (1994a). Characterization of spontaneous MEG rhythms in healthy adults, *Electroencephalography and Clinical Neurophysiology*, Vol. 91, pp. 237-248. Salmelin, R. & Hari, R. (1994b). Spatiotemporal characteristics of sensorimotor neuromagnetic rhythms related to thumb movement, *Neuroscience*, Vol. 60, pp. 537-550.

from spatio-temporal MEG data, *IEEE Transactions on Biomedical Engineering*, Vol.

for single-trial EEG classification in a movement task, *Clinical Neurophysiology*, Vol.

power measurements of scalp EEG, *Electroencephalography and Clinical* 

(ERD) preceding and following voluntary self-paced movement,

(ERD) preceding and following voluntary self-paced movements,

voluntary movement, *Electroencephalography and Clinical Neurophysiology*, Vol. 72,

involved in preparation for hand movement based on classification of alpha and central beta rhythms in single EEG trials in man, *Neuroscience Letters*, Vol. 181, pp.

correlate of an idling motor area?, *Electroencephalography and Clinical* 

between imagination of right and left hand movement, *Electroencephalography and* 

Postmovement beta synchronization in patients with Parkinson's disease, *Journal of* 

wrist, finger and thumb movement, *Electroencephalography and Clinical* 

magnetic induction tomography using a plannar gradiometer, *Physiological* 


Jensen, O. & Vanni, S. (2002). A new method to identify multiple sources of oscillatory activity from magnetoencephalographic data, *Neuroimage*, Vol. 15, pp. 568-574. Joliot, M.; Papathanassiou, D.; Mellet, E.; Quinton, O.; Mazoyer, N.; Courtheoux, P. &

intersubject stereotaxic averaged data, *Neuroimage*, Vol. 10, pp. 430-447. Jung, T. P.; Makeig, S.; Westerfield, M.; Townsend, J.; Courchesne,E. & Sejnowski, J. T.

Kajola, M.; Ahonen, A.; Hamalainen, M. S.; Knuutila, J.; Lounasmaa, O. V.; Simola, J. &

in Finland, *Clinical Physics & Physiological Measurement*, Vol. 12, pp. 39-44. Kao, Y. H.; Guo, W. Y.; Wu, Y. T.; Liu, K. C.; Chai, W. Y.; Lin,C. Y.; Hwang, Y. H.; Liou, A. J.

Kalcher, J. & Pfurtscheller, G. (1995). Discrimination between phase-locked and non-phase-

Klimesch, W.; Russegger, H.; Doppelmayr, M. & Pachinger, T. (1998). A method for the

Leocani, L.; Manganotti, C. T.; Zhuang, P. & Hallett, M. (1997). Event-related coherence and

Lins, O. G.; Picton, T. W.; Berg, P. & Scherg, M. (1993a). Ocular artifacts in EEG and eventreated potentials. I. Scalp topography, *Brain Topography*, Vol. 6, pp. 51-63. Lins, O. G.; Picton, T. W.; Berg,P. & Scherg, M. (1993b). Ocular artifacts in recording EEGs

Lopes da Silva, F. H. (1991). Neural mechanisms underlying brain waves: from neural

Lopes da Silva, F. H. (1996). The generation of electric and magnetic signals of the brain by

Makeig, S.; Jung, T. P.; Bell, A. J.; Ghahremani, D. & Sejnowski, T. (1997). Blind separation of

McKeown, M. J.; Makeig, S.; Brown. G. G.; Jung. T. P.; Kindermann. S. S.; Bell. A. J &

Mckeown, M. & Radtke, R. (2001). Phasic and tonic coupling between EEG and EMG

local networks, *Comprehensive human physiology*, Vol. 1, pp. 509-531.

spatial components, *Human Brain Mapping*, Vol. 6, pp. 160-188.

*Brain Mapping*, Vol. 14, pp. 166-185.

*Neurophysiology*, Vol. 94, pp. 381-384.

Vol. 103, pp. 199-206.

pp. 81-93.

10979-10984.

*Topography*, Vol. 6, pp. 65-78.

*neurophysiology*, Vol. 18, pp. 45-57.

885-894.

Mazoyer, B. (1999). FMRI and PET of self-paced finger movement: comparison of

(2001). Analysis and Visualization of Single-Trial Event-Related Potentials, *Human* 

Vilkman, V. (1991). Development of multichannel neuromagnetic instrumentation

K.; Cheng, H. C.; Yeh, T. C.; Hsieh, J. C. & Teng, M. M. H. (2003). Hemodynamic segmentation of MR brain perfusion images using independent component, thresholding and Bayesian estimation, *Magnetic Resonance in Medicine*, Vol. 49, pp.

locked event-related EEG activity, *Electroencephalography and Clinical* 

calculation of induced band power: implications for the significance of brain oscillation, *Electroencephalography and Clinical Neurophysiology*, Vol. 108, pp. 123-130.

event-related desynchronization/synchronization in the 10 Hz and 20 Hz EEG during self-paced movements, *Electroencephalography and clinical neurophysiology*,

and event-related potentials. II. Source dipoles and source components, *Brain* 

membranes to networks, *Electroencephalography and Clinical Neurophysiology*, Vol. 79,

auditory event-related brain responses into independent components, Proceedings of the National Academy of Sciences of the United States of America, Vol. 94, pp.

Sejnowski, T. J. (1998). Analysis of fMRI data by blind separation into independent

demonstrated with independent component analysis, *Journal of clinical* 


**5** 

*1,2USA 3PR of China* 

*Research Center of Excellence,* 

*Medical Center, Lexington,* 

**Developing an MRI-Based Biomarker for** 

**Early Diagnosis of Parkinson's Disease** 

*1Department of Anatomy and Neurobiology, Morris K. Udall Parkinson's Disease* 

*3Department of Physiology, Key Laboratory for Neurodegenerative Disorders of the* 

*2Magnetic Resonance Imaging and Spectroscopy Center, University of Kentucky Chandler* 

Parkinson's disease (PD) is a relentlessly progressive disorder causing disability in most individuals and cannot be controlled with available medication. PD is currently considered a systemic disease with complex motor disorders and non-motor deficits which appear before or in parallel with motor deficits and then worsen with disease progression (Chaudhuri *et al.*, 2006; Ferrer *et al.*, 2010). In a recent survey, the projected number of individuals with PD will dramatically increase in 20 years especially in the most populated countries like China, India, Brazil and the United States (Dorsey *et al.*, 2007). Current causative theories for PD include complex interactions between genetic susceptibility and environmental factors. These and possibly other mechanisms lead to a progressive and variable degree of dopamine (DA) neuron loss in the substantia nigra compacta (SNc) resulting in DA depletion in the striatum (Hornykiewicz & Kish, 1987; Marsden & Obeso, 1994) that then leads to the clinical manifestation of PD. Studies have demonstrated that PD is characterized by a presymptomatic phase, likely lasting years, or even decades, during which neuronal degeneration is occurring but before clinical symptoms appear (Hubble, 2000; DeKosky & Marek, 2003; Katzenschlager & Lees, 2004). In addition, studies have demonstrated that most patients when diagnosed with PD have already lost a significant amount of SNc DA neurons in the range of 50% cell loss. Based on detailed pathological studies, Fearnley and Lees (1991) have proposed the notion that the loss of nigral neurons would occur exponentially, with greater loss occurring within the first decade in the disease

While our understanding of PD has grown over the course of the last two centuries and PD is one of the best understood neurodegenerative diseases, our ability to treat PD remains limited. Given the progressive nature of the disease, the question becomes is it possible to divert or change the rate of the progression? Inherent to this question is our ability to identify where an individual is along the path of this disease. Thus it would behoove us to

process, and then reaching over 90% loss at the time of death.

**1. Introduction** 

Jorge E. Quintero1, Xiaomin Wang3 and Zhiming Zhang1,2,\*

*Ministry Education, Capital Medical University, Beijing* 


## **Developing an MRI-Based Biomarker for Early Diagnosis of Parkinson's Disease**

Jorge E. Quintero1, Xiaomin Wang3 and Zhiming Zhang1,2,\* *1Department of Anatomy and Neurobiology, Morris K. Udall Parkinson's Disease Research Center of Excellence, 2Magnetic Resonance Imaging and Spectroscopy Center, University of Kentucky Chandler Medical Center, Lexington, 3Department of Physiology, Key Laboratory for Neurodegenerative Disorders of the Ministry Education, Capital Medical University, Beijing 1,2USA 3PR of China* 

### **1. Introduction**

114 Diagnostics and Rehabilitation of Parkinson's Disease

Salmelin, R.; Hamalainen, M.; Kajola, M. & Hari, R. (1995). Functional segregation of

Silen T.; Forss, N.; Jensen, O. & Hari, R. (2000). Abnormal Reactivity of the ~20-Hz Motor

Stancak, A. Jr. & Pfurtscheller G. (1996a). The effects of handedness and type of movement

Sterman, M. B. (1999). Event-related EEG response correlates of task difficulty, sleep

Tamas, G.; Szirmai, I.; Palvolgyi, L.; Takats, A. & Kamondi, A. (2003). Impairment of post-

Tang, A. C.; Pearlmutter, B. A.; Malaszenko, N. A. & Phung, D. B. (2002). Independent

Taniguchi, M.; Kato, A.; Fujita, N.; Hirata, M.; Tanaka, H.; Kihara, T.; Ninomiya, H.;

Tesche, C. D.; Unsitalo M. A.; Ilmoniemi R. J.; Huotilainen M.; Kajola M. & Salonen O.

Uutela, K.; Hamalainen, M. & Somersalo, E. (1999). Visualization of

Wolpaw, J. R. & McFarland, D. J. (1994). Multichannel EEG-based brain-computer

Wu, Y. T.; Lee, P. L.; Chen, L. F.; Yeh, T. C. & Hsieh, J. C. (2002). Single-trial quantification of

Wu, Y. T.; Lee, P. L.; Chen, L. F.; Yeh, T. C. & Hsieh, J. C. (2003). Quantification of

Yabe, H.; Satio, F. & Fukushima, Y. (1993). Median method for detecting endogenous event-

filtered magnetoencephalograpy, *Neuroimage*, Vol. 12, pp. 298-306.

*Electroencephalography and Clinical Neurophysiology*, Vol. 99, pp. 174-182. Stancak, A. Jr. & Pfurtscheller G. (1996b). Mu-rhythm changes in brisk and slow self-paced

finger movements, *Neuroreport*, Vol. 7, pp. 1161-1164.

tremor, *Clinical Neurophysiology*, Vol. 114, pp. 614-623.

237-243.

*Neuroimage*, Vol. 12, pp. 707-712.

Silva, 233-242, Elsevier, Amsterdam.

*Neuroimage*, Vol. 17, pp. 1773-1789.

Neurophysiology, Vol. 95, pp. 189-200.

10, pp. 173-180.

449.

1045-1047.

pp. 396-398.

pp. 403-407.

movement-related rhythmic activity in the human brain, *Neuroimage*, Vol. 2, pp.

Cortex Rhythm in Unverricht Lundborg Type Progressive Myoclonus Epilepsy,

on the contralateral preponderance of mu-rhythm desynchronization,

deprivation and sensory distraction, In: *Event-related desynchronization. Handbook of electroencephalography and clinical neurophysiology*. G. Pfurtscheller & F. H. Lopes da

movement beta synchronization in parkinson's disease is related to laterality of

components of magnetoencephalography: single-trial response onset times,

Hirabuki, N.; Nakamura, H.; Robinson, S. E.; Cheyne, D. & Yoshimine, T. (2000). Movement-related desynchronization of the cerebral cortex studied with spatially

(1995). Signal-space projections of MEG data characterize both distributed and well-localized neuroal sources, Electroencephalography and Clinical

magnetoencephalography data using minimum current estimates, *Neuroimage*, Vol.

communication, *Electroencephalography and clinical Neurophysiology*, Vol. 90, pp. 444-

imagery beta-band Mu rhythm in finger lifting task using independent component analysis (ICA). *Proceeding of BioMag 13th international conference on biomagnetism*, pp.

movement-related modulation on beta activity of single-trial magnetoencephalography measuring using independent component analysis (ICA). *Proceeding of the 1st international IEEE EMBS conference on neural engineering*,

related brain potentials, *Electroencephalography and clinical Neurophysiology*, Vol. 87,

Parkinson's disease (PD) is a relentlessly progressive disorder causing disability in most individuals and cannot be controlled with available medication. PD is currently considered a systemic disease with complex motor disorders and non-motor deficits which appear before or in parallel with motor deficits and then worsen with disease progression (Chaudhuri *et al.*, 2006; Ferrer *et al.*, 2010). In a recent survey, the projected number of individuals with PD will dramatically increase in 20 years especially in the most populated countries like China, India, Brazil and the United States (Dorsey *et al.*, 2007). Current causative theories for PD include complex interactions between genetic susceptibility and environmental factors. These and possibly other mechanisms lead to a progressive and variable degree of dopamine (DA) neuron loss in the substantia nigra compacta (SNc) resulting in DA depletion in the striatum (Hornykiewicz & Kish, 1987; Marsden & Obeso, 1994) that then leads to the clinical manifestation of PD. Studies have demonstrated that PD is characterized by a presymptomatic phase, likely lasting years, or even decades, during which neuronal degeneration is occurring but before clinical symptoms appear (Hubble, 2000; DeKosky & Marek, 2003; Katzenschlager & Lees, 2004). In addition, studies have demonstrated that most patients when diagnosed with PD have already lost a significant amount of SNc DA neurons in the range of 50% cell loss. Based on detailed pathological studies, Fearnley and Lees (1991) have proposed the notion that the loss of nigral neurons would occur exponentially, with greater loss occurring within the first decade in the disease process, and then reaching over 90% loss at the time of death.

While our understanding of PD has grown over the course of the last two centuries and PD is one of the best understood neurodegenerative diseases, our ability to treat PD remains limited. Given the progressive nature of the disease, the question becomes is it possible to divert or change the rate of the progression? Inherent to this question is our ability to identify where an individual is along the path of this disease. Thus it would behoove us to

Developing an MRI-Based Biomarker for Early Diagnosis of Parkinson's Disease 117

the basal ganglia, much controversy still exists. Recent problems have been encountered in clinical trials that have used radioligand imaging to quantify medication response. For example, there appears to be a discrepancy between current imaging protocols and clinical outcomes. In National Institutes of Health (NIH) sponsored randomized double-blind studies on PD patients receiving either fetal tissue transplants or sham surgery, a 40% increase in 18F-dopa uptake in the putamen contrasted with a modest (non-significant) 18% improvement in the mean UPDRS in one study involving 40 patients (Freed *et al.*, 2001). In the second study involving 34 patients, a 20-30% increase of 18F-dopa uptake was seen in the striatum, but clinical changes failed to reach statistical significance (Brooks, 2004). Most recently, a significant increase was found in 18F-dopa uptake in the putamen of PD patients receiving trophic therapy, while clinical improvements did not differ significantly from the

In general, biomarkers must be biologically and clinically relevant, analytically sound, operationally practical, timely, interpretable and cost effective. On the other hand, biomarkers must be objectively measured indicators of biological and pathobiological process or pharmacologic responses to treatment. The biomarkers should be used to substitute for a clinical endpoint (predict benefit or harm) based on epidemiologic, therapeutic, pathophysiologic or other scientific evidence (Biomarkers definition working group, 2001). Specifically for PD, the biomarkers must be indicators of biological processes that change with the progression of the nigrostriatal system. The biomarkers should 1) correlate to some extents with severity of PD assessed by behavior and with pathophysiological changes such as the number of surviving neurons in the SNc; 2) reflect true disease status or predict clinical outcomes; 3) be used to assess efficacy and/or

Our preliminary studies have shown evidence that blood-oxygenation-level-dependent (BOLD)-phMRI can be used as a non-invasive imaging modality to detect functional changes of the dopamine system in parkinsonian monkeys. More importantly, the studies were conducted in a conventional clinical MRI scanner without the injection of contrast agents. Using this imaging method, a significant correlation was found between the amphetamine-evoked BOLD response and the number of surviving dopamine neurons in the substantia nigra, which was also significantly correlated with bradykinesia scores on the nonhuman primate parkinsonian rating scale (Ovadia *et al.*, 1995), suggesting that phMRI may be used as a biomarker to assess dopamine neuronal loss in PD. Recently, fMRI has become a popular tool for imaging of functionally active brain regions in healthy and diseased brains. The use of fMRI is promoting the emergence of a new area of research, one that is complementary to more invasive techniques for measuring neural activity in animal models while better understanding the function and dysfunction of the human brain. The most common method of fMRI is the BOLD imaging technique. fMRI takes advantage of the coupling between neural activity and hemodynamics (the local control of blood flow and oxygenation) in the brain to allow the non-invasive localization and measurement of brain

control group (Lang *et al.*, 2006).

**1.4 What is BOLD-phMRI?** 

activity (Fig. 1).

**1.3 What are imaging biomarkers for PD?** 

responsiveness of treatment, and 4) be used as surrogate endpoints.

be able to establish indicators of the disease stage while intervention remains a possibility. Here we describe the development of using non-invasive functional imaging as a biomarker for the early diagnosis of PD**.** 

## **1.1 Difficulty of early detection of PD**

The diagnosis and treatment of PD is fraught with problems: 1) so far, no objective measures are available for the diagnosis of PD (Wu *et al.*, 2011); 2) it is unknown whether a linear relationship exists between a worsening in the Unified Parkinson's Disease Rating Scale (UPDRS), or other clinical scales, and the progressive degeneration of the nigrostriatal system; 3) no objective measures are available for testing responsiveness of therapy. Therefore, biomarkers of disease progression before the appearance of symptoms would be of onsiderable value; thus, neuroimaging techniques may be good candidates for meeting the challenges. In the past decade, radiotracer imaging of the nigrostriatal dopaminergic system has been extensively explored with positron emission tomography (PET) and single photon emission computed tomography (SPECT) imaging protocols and has become a prominent biomarker in PD although these techniques are still controversial in some aspects such as the interpretation of imaging data and disconnection with clinical outcomes (Brooks *et al.*, 2003; Ravina *et al.*, 2005). However, the spatial resolution of these techniques is relatively poor, thus reducing their utility in mapping subtle changes in neuroanatomy and neurochemistry with PD progression (Snow *et al.*, 2000). Furthermore, PET imaging is not widely available and is expensive (~US\$3,000-\$6,000) because of the need to generate and use radioactive nucleotides onsite. Clearly there is a need for imaging techniques that do not require radioactive isotopes but ones that would still be sensitive enough to usefully and longitudinally monitor the development, progression, and treatment of PD. The ideal technique would 1) permit high-resolution imaging of brain sites affected by PD processes, 2) provide valid assessment of the underlying neuroanatomical state, and 3) be safe to allow repeated tests. A hypothesis for PD is that the disease severity corresponds to the magnitude and pattern of histological and neuroimaging abnormalities (DeKosky & Marek, 2003; Eckert & Eidelberg, 2004; Seibyl *et al.*, 2005). Based on our own previous studies, and those of others in rodents, nonhuman primates, and humans, pharmacological MRI (phMRI; or functional MRI with specific pharmacological stimulation) would be a good candidate because of its high resolution, sensitivity, reproducibility, wide availability, and low cost (Nguyen *et al.*, 2000; Tracey, 2001; Honey & Bullmore, 2004; Jenkins *et al.*, 2004; Chin *et al.*, 2008; Thiel, 2009; Rasmussen Jr, 2010).

#### **1.2 Why is a new imaging protocol needed for PD?**

In the past decade, PET and SPECT have become the most widely used and accepted imaging methods for PD research (de la Fuente-Fernandez & Stoessl, 2002; Eckert & Eidelberg, 2004). Worsening motor disability along with 18F-dopa uptake decreases in the putamen (Brooks *et al.*, 1990) correlate with the storage of DA within vesicles (Hoshi *et al.*, 1993) and with the number of functioning DA terminals in the striatum (Snow *et al.*, 1993). Currently, *in vivo* measurements can be conducted using SPECT with ligands for the DA transporter (DAT) such as [(123)I]N-omega-fluoropropyl-2beta-carbomethoxy-3beta-{4 iodophenyl}nortropane (FP-CIT) that provide a measure of DA terminal integrity (DeKosky & Marek, 2003; Andringa *et al.*, 2005). Although the aforementioned studies have shown that these neuroimaging techniques are capable of mapping changes in dopaminergic function in the basal ganglia, much controversy still exists. Recent problems have been encountered in clinical trials that have used radioligand imaging to quantify medication response. For example, there appears to be a discrepancy between current imaging protocols and clinical outcomes. In National Institutes of Health (NIH) sponsored randomized double-blind studies on PD patients receiving either fetal tissue transplants or sham surgery, a 40% increase in 18F-dopa uptake in the putamen contrasted with a modest (non-significant) 18% improvement in the mean UPDRS in one study involving 40 patients (Freed *et al.*, 2001). In the second study involving 34 patients, a 20-30% increase of 18F-dopa uptake was seen in the striatum, but clinical changes failed to reach statistical significance (Brooks, 2004). Most recently, a significant increase was found in 18F-dopa uptake in the putamen of PD patients receiving trophic therapy, while clinical improvements did not differ significantly from the control group (Lang *et al.*, 2006).

#### **1.3 What are imaging biomarkers for PD?**

116 Diagnostics and Rehabilitation of Parkinson's Disease

be able to establish indicators of the disease stage while intervention remains a possibility. Here we describe the development of using non-invasive functional imaging as a biomarker

The diagnosis and treatment of PD is fraught with problems: 1) so far, no objective measures are available for the diagnosis of PD (Wu *et al.*, 2011); 2) it is unknown whether a linear relationship exists between a worsening in the Unified Parkinson's Disease Rating Scale (UPDRS), or other clinical scales, and the progressive degeneration of the nigrostriatal system; 3) no objective measures are available for testing responsiveness of therapy. Therefore, biomarkers of disease progression before the appearance of symptoms would be of onsiderable value; thus, neuroimaging techniques may be good candidates for meeting the challenges. In the past decade, radiotracer imaging of the nigrostriatal dopaminergic system has been extensively explored with positron emission tomography (PET) and single photon emission computed tomography (SPECT) imaging protocols and has become a prominent biomarker in PD although these techniques are still controversial in some aspects such as the interpretation of imaging data and disconnection with clinical outcomes (Brooks *et al.*, 2003; Ravina *et al.*, 2005). However, the spatial resolution of these techniques is relatively poor, thus reducing their utility in mapping subtle changes in neuroanatomy and neurochemistry with PD progression (Snow *et al.*, 2000). Furthermore, PET imaging is not widely available and is expensive (~US\$3,000-\$6,000) because of the need to generate and use radioactive nucleotides onsite. Clearly there is a need for imaging techniques that do not require radioactive isotopes but ones that would still be sensitive enough to usefully and longitudinally monitor the development, progression, and treatment of PD. The ideal technique would 1) permit high-resolution imaging of brain sites affected by PD processes, 2) provide valid assessment of the underlying neuroanatomical state, and 3) be safe to allow repeated tests. A hypothesis for PD is that the disease severity corresponds to the magnitude and pattern of histological and neuroimaging abnormalities (DeKosky & Marek, 2003; Eckert & Eidelberg, 2004; Seibyl *et al.*, 2005). Based on our own previous studies, and those of others in rodents, nonhuman primates, and humans, pharmacological MRI (phMRI; or functional MRI with specific pharmacological stimulation) would be a good candidate because of its high resolution, sensitivity, reproducibility, wide availability, and low cost (Nguyen *et al.*, 2000; Tracey, 2001; Honey & Bullmore, 2004; Jenkins *et al.*, 2004; Chin *et al.*,

In the past decade, PET and SPECT have become the most widely used and accepted imaging methods for PD research (de la Fuente-Fernandez & Stoessl, 2002; Eckert & Eidelberg, 2004). Worsening motor disability along with 18F-dopa uptake decreases in the putamen (Brooks *et al.*, 1990) correlate with the storage of DA within vesicles (Hoshi *et al.*, 1993) and with the number of functioning DA terminals in the striatum (Snow *et al.*, 1993). Currently, *in vivo* measurements can be conducted using SPECT with ligands for the DA transporter (DAT) such as [(123)I]N-omega-fluoropropyl-2beta-carbomethoxy-3beta-{4 iodophenyl}nortropane (FP-CIT) that provide a measure of DA terminal integrity (DeKosky & Marek, 2003; Andringa *et al.*, 2005). Although the aforementioned studies have shown that these neuroimaging techniques are capable of mapping changes in dopaminergic function in

for the early diagnosis of PD**.** 

**1.1 Difficulty of early detection of PD** 

2008; Thiel, 2009; Rasmussen Jr, 2010).

**1.2 Why is a new imaging protocol needed for PD?** 

In general, biomarkers must be biologically and clinically relevant, analytically sound, operationally practical, timely, interpretable and cost effective. On the other hand, biomarkers must be objectively measured indicators of biological and pathobiological process or pharmacologic responses to treatment. The biomarkers should be used to substitute for a clinical endpoint (predict benefit or harm) based on epidemiologic, therapeutic, pathophysiologic or other scientific evidence (Biomarkers definition working group, 2001). Specifically for PD, the biomarkers must be indicators of biological processes that change with the progression of the nigrostriatal system. The biomarkers should 1) correlate to some extents with severity of PD assessed by behavior and with pathophysiological changes such as the number of surviving neurons in the SNc; 2) reflect true disease status or predict clinical outcomes; 3) be used to assess efficacy and/or responsiveness of treatment, and 4) be used as surrogate endpoints.

#### **1.4 What is BOLD-phMRI?**

Our preliminary studies have shown evidence that blood-oxygenation-level-dependent (BOLD)-phMRI can be used as a non-invasive imaging modality to detect functional changes of the dopamine system in parkinsonian monkeys. More importantly, the studies were conducted in a conventional clinical MRI scanner without the injection of contrast agents. Using this imaging method, a significant correlation was found between the amphetamine-evoked BOLD response and the number of surviving dopamine neurons in the substantia nigra, which was also significantly correlated with bradykinesia scores on the nonhuman primate parkinsonian rating scale (Ovadia *et al.*, 1995), suggesting that phMRI may be used as a biomarker to assess dopamine neuronal loss in PD. Recently, fMRI has become a popular tool for imaging of functionally active brain regions in healthy and diseased brains. The use of fMRI is promoting the emergence of a new area of research, one that is complementary to more invasive techniques for measuring neural activity in animal models while better understanding the function and dysfunction of the human brain. The most common method of fMRI is the BOLD imaging technique. fMRI takes advantage of the coupling between neural activity and hemodynamics (the local control of blood flow and oxygenation) in the brain to allow the non-invasive localization and measurement of brain activity (Fig. 1).

Developing an MRI-Based Biomarker for Early Diagnosis of Parkinson's Disease 119

We hypothesize that the BOLD-fMRI response to a specific DA stimulation could serve as a potential biomarker for PD because of its unique features which are different from other neuroimaging technologies as follows: 1) High sensitivity and reproducibility, and relatively high specificity, 2) Minimal invasiveness or patient discomfort ("subject friendly"), 3) Low per-usage cost (this is especially important if widespread screening is contemplated), and 4)

A reliable and reproducible model of dopamine deficiency/a model simulating human PD is developed by unilateral administration of neurotoxin 1-methyl-4-phenyl-1,2,3,6 etrahydropyridine (MPTP) through the carotid artery. The specific neurotoxic actions of MPTP are produced when it is metabolized by monoamine oxidase B into 1-methyl-4 phenylpyridinium (MPP +), a complex I mitochondrial neurotoxin with relative specificity for dopamine neurons in the substantia nigra (Langston & Ballard, 1983; Nicklas *et al.*, 1987; Richardson *et al.*, 2007). Ding and colleagues (2008) described the key features modeled by MPTP toxicity including 1) all animals show parkinsonian features often seen in idiopathic PD such as bradykinesia, rigidity, postural, and balance instability, 2) these PD features can be partially normalized by levodopa treatment, which is the most efficacious drug to treat PD motor symptoms and is widely considered the "gold standard" treatment for the disease, 3) massive neuronal loss of dopaminergic neurons in the SNc and dopaminergic

1) Mapping of MPTP-induced functional changes with d-amphetamine stimulations (from a pre-synaptic perspective) and 2) Mapping of MPTP-induced functional changes with APO challenge (from a post-synaptic perspective). In early studies, the scans were conducted on a Siemens VISION 1.5 T MRI scanner using the body coil to transmit radio frequency and an 8 cm diameter surface coil placed above the monkey's head for RF signal reception. The anatomical structures of interest were visualized using a 3D FLASH sequence with 1 mm isotropic resolution (TR/TE=21/6 ms, flip angle = 30°, image matrix size = 128x128x90, field of view = 128 mm). The functional MR images from pharmacological challenges were acquired continuously using a FLASH 2D multiple gradient-recalled-echo (MGRE) navigator sequence (Chen *et al.*, 1996). The ROI dimensions were 3x3x3 mm, each representing a 27 mm3 volume. ROIs were manually selected in both hemispheres of MPTPlesioned and normal control animals based on the co-registered 3D anatomical images acquired from the FLASH sequence. Because of variability in the inherent noise level due to differences in positioning animals for each scan and the movements during scanning, the replicate scans were treated as independent observations in the analysis. For later studies, images were acquired on a Siemens 3T Trio clinical MRI system using a dedicated receiveonly coil for reception, which was designed and developed by our group. The BOLD-effect weighted MR images used to measure the phMRI response were acquired in an anatomically coronal plane. The image planes of the acquisition were arranged to cover the motor cortex and the basal ganglia. A segmented gradient-echo EPI sequence with TE=28 ms and a turbo factor of 7 was used to reduce echo train length and minimize magnetic susceptibility-related artifacts. The EPI sequence acquisition parameters are FOV=112x98 mm and image matrix 64x56 for an in-plane resolution of 1.75 mm. A total of 15 contiguous

**2. phMRI detects dopamine deficiency in parkinsonian monkeys** 

fibers in the striatum, and 4) remarkable declines in DA and DA metabolites.

**2.1 An animal model of dopamine deficiency in rhesus monkeys** 

wide availability.

**2.2 phMRI procedures** 

Fig. 2. Using phMRI for the early diagnosis of PD. As part of a multi-factor approach, phMRI provides a possible means of screening for the underlying neurological changes in parkinsonism or PD.

We hypothesize that the BOLD-fMRI response to a specific DA stimulation could serve as a potential biomarker for PD because of its unique features which are different from other neuroimaging technologies as follows: 1) High sensitivity and reproducibility, and relatively high specificity, 2) Minimal invasiveness or patient discomfort ("subject friendly"), 3) Low per-usage cost (this is especially important if widespread screening is contemplated), and 4) wide availability.

## **2. phMRI detects dopamine deficiency in parkinsonian monkeys**

#### **2.1 An animal model of dopamine deficiency in rhesus monkeys**

A reliable and reproducible model of dopamine deficiency/a model simulating human PD is developed by unilateral administration of neurotoxin 1-methyl-4-phenyl-1,2,3,6 etrahydropyridine (MPTP) through the carotid artery. The specific neurotoxic actions of MPTP are produced when it is metabolized by monoamine oxidase B into 1-methyl-4 phenylpyridinium (MPP +), a complex I mitochondrial neurotoxin with relative specificity for dopamine neurons in the substantia nigra (Langston & Ballard, 1983; Nicklas *et al.*, 1987; Richardson *et al.*, 2007). Ding and colleagues (2008) described the key features modeled by MPTP toxicity including 1) all animals show parkinsonian features often seen in idiopathic PD such as bradykinesia, rigidity, postural, and balance instability, 2) these PD features can be partially normalized by levodopa treatment, which is the most efficacious drug to treat PD motor symptoms and is widely considered the "gold standard" treatment for the disease, 3) massive neuronal loss of dopaminergic neurons in the SNc and dopaminergic fibers in the striatum, and 4) remarkable declines in DA and DA metabolites.

## **2.2 phMRI procedures**

118 Diagnostics and Rehabilitation of Parkinson's Disease

Fig. 1. fMRI provides an insight into neural activity. The BOLD signal has several constituents: (1) the neuronal response to a stimulus or background modulation; (2) the complex relationship between neuronal activity and triggering a haemodynamic response (termed neurovascular coupling); (3) the haemodynamic response itself; and (4) the detection of the response by an MRI scanner (from Arthurs & Boniface, 2002).

Fig. 2. Using phMRI for the early diagnosis of PD. As part of a multi-factor approach, phMRI provides a possible means of screening for the underlying neurological changes in

parkinsonism or PD.

1) Mapping of MPTP-induced functional changes with d-amphetamine stimulations (from a pre-synaptic perspective) and 2) Mapping of MPTP-induced functional changes with APO challenge (from a post-synaptic perspective). In early studies, the scans were conducted on a Siemens VISION 1.5 T MRI scanner using the body coil to transmit radio frequency and an 8 cm diameter surface coil placed above the monkey's head for RF signal reception. The anatomical structures of interest were visualized using a 3D FLASH sequence with 1 mm isotropic resolution (TR/TE=21/6 ms, flip angle = 30°, image matrix size = 128x128x90, field of view = 128 mm). The functional MR images from pharmacological challenges were acquired continuously using a FLASH 2D multiple gradient-recalled-echo (MGRE) navigator sequence (Chen *et al.*, 1996). The ROI dimensions were 3x3x3 mm, each representing a 27 mm3 volume. ROIs were manually selected in both hemispheres of MPTPlesioned and normal control animals based on the co-registered 3D anatomical images acquired from the FLASH sequence. Because of variability in the inherent noise level due to differences in positioning animals for each scan and the movements during scanning, the replicate scans were treated as independent observations in the analysis. For later studies, images were acquired on a Siemens 3T Trio clinical MRI system using a dedicated receiveonly coil for reception, which was designed and developed by our group. The BOLD-effect weighted MR images used to measure the phMRI response were acquired in an anatomically coronal plane. The image planes of the acquisition were arranged to cover the motor cortex and the basal ganglia. A segmented gradient-echo EPI sequence with TE=28 ms and a turbo factor of 7 was used to reduce echo train length and minimize magnetic susceptibility-related artifacts. The EPI sequence acquisition parameters are FOV=112x98 mm and image matrix 64x56 for an in-plane resolution of 1.75 mm. A total of 15 contiguous

Developing an MRI-Based Biomarker for Early Diagnosis of Parkinson's Disease 121

Apomorphine administration strongly activated the MPTP-denervated putamen (Figs. 3A and 4C) and substantia nigra (Fig.4D). An opposite response (a positive ΔR2\* value) was evident in the contralateral putamen (Fig. 4G) and substantia nigra (Fig. 4H). The differences between the intact and lesioned substantia nigra and between the intact and lesioned putamen were highly significant, P < 0.01(*t*-test), in both cases. In contrast, ΔR2\* responses in the caudate nucleus and in the corpus callosum were not significant, nor were there significant hemispheric differences in activation or deactivation with the contralateral

The phMRI responses to amphetamine treatment in the putamen (Figs. 3B and 4G) and substantia nigra (Fig. 4H) were the inverse of those seen with apomorphine. Amphetamineinduced decreases (positive ΔR2\* values) in the lesioned putamen and substantia nigra suggested diminished neuronal activity in both sites. In contrast, amphetamine induced the opposite ΔR2\* response in the intact left side, tending to increase activation in the putamen and substantia nigra. The responses in the intact putamen and intact substantia nigra were significantly different from their lesioned counterparts. Again, the corpus callosum and the caudate nucleus displayed only small, insignificant changes in response to amphetamine

Fig. 4. phMRI responses in the nigrostriatal system. Depending on the means of stimulation, phMRI reveals a differential activations and deactivations in the nigrostriatal system. After APO stimulation (A-D) or *d*-amphetamine stimulation (E-H). \*\**P*<0.01; \**P*<0.05; unpaired *t*-

In a later study, post-mortem histopathological evaluation revealed that the unilateral MPTP administration (received 5 years before the analysis) produced a massive (85%) loss of the rate-limiting enzyme for DA formation, tyrosine hydroxylase, (TH+) cells in the midbrain on the ipsilateral side receiving the infusion of the neurotoxin. TH+ cell numbers were significantly higher on the un-lesioned side compared to the MPTP-lesioned side. More importantly, the number of TH+ cells was strongly correlated with the phMRI responses in

caudate or with a comparable region in the contralateral callosum (Figs. 4A and 4E).

**2.4 phMRI-response and MPTP-induced dopamine deficiency** 

**2.4.1 phMRI responses in MPTP-lesioned structures** 

stimulation (Figs. 4E and 4F).

test (from Zhang *et al.*, 2006).

**2.4.2 phMRI-responses and loss of DA neurons in the SN** 

slices, each 2 mm-thick, were acquired at a rate of 15 s per EPI volume. The overall scan duration was 80 minutes with 128 volumes acquired prior to apomorphine (APO) administration as a baseline and 192 after APO to track the response. Images were motion corrected and spatially smoothed using a Gaussian kernel of width 3.5 mm. phMRI response was calculated as the fractional signal change in % of the average of the post-APO image data relative to the pre-APO baseline. A co-registered high-resolution (0.67×0.67×1 mm) T1 weighted anatomical MRI scan was acquired in each session for spatial localization of the activation response. Prior to the administration of d-amphetamine (2.0mg/kg) or APO (0.1 mg/kg), a total of 40 image frames were collected over 20 min to determine the baseline state. Following injection of d-amphetamine or APO, an additional 40 frames were collected to track the dynamic response (Zhang *et al.*, 2001; Andersen *et al.*, 2002). The change in R2\*, i.e. R2\* which represents the phMRI activation response to drug, was determined as the difference between the mean R2\* across 20 images post drug administration during the period of peak response (5-15 min) and the mean R2 \* within the 40 baseline images. A reduction ("negative" change) in R2 \* associated with a local decrease of paramagnetic deoxyhemoglobin is an indicator of BOLD-effect activation (Chen *et al.*, 1996).

#### **2.3 phMRI-responses correlate with severity of PD**

Six out of six animals responded positively to APO treatment represented by 44% improvements in parkinsonian symptoms. The same dose of APO also evoked phMRI responses by increasing the phMRI signal intensity. The typical phMRI (BOLD effect) responses to APO were gradually increased after APO administration only in the structure on the ipsilateral side receiving MPTP administration. Interestingly, but not surprising, APOinduced behavioral changes (PD features) were significantly correlated with APO-induced phMRI responses in the putamen, premotor cortex, and cingulate gyrus. When compared with standard but objective measures, there was a significant negative correlation between the phMRI responses in the putamen and distance travelled and movement speed. Similar relationships were also seen between phMRI responses in the motor cortex and daytime homecage activity and between phMRI responses in the caudate nucleus and movement speed.

Fig. 3. phMRI reveals nigrostriatal system responsiveness to dopamine stimulation. Coronal MRI scans depicting areas of activation and deactivation (represented by the psedudocolor) in the brain after an APO or amphetamine challenge in unitalteral MPTP-lesioned nonhuman primates (from Zhang *et al.*, 2006).

#### **2.4 phMRI-response and MPTP-induced dopamine deficiency 2.4.1 phMRI responses in MPTP-lesioned structures**

120 Diagnostics and Rehabilitation of Parkinson's Disease

slices, each 2 mm-thick, were acquired at a rate of 15 s per EPI volume. The overall scan duration was 80 minutes with 128 volumes acquired prior to apomorphine (APO) administration as a baseline and 192 after APO to track the response. Images were motion corrected and spatially smoothed using a Gaussian kernel of width 3.5 mm. phMRI response was calculated as the fractional signal change in % of the average of the post-APO image data relative to the pre-APO baseline. A co-registered high-resolution (0.67×0.67×1 mm) T1 weighted anatomical MRI scan was acquired in each session for spatial localization of the activation response. Prior to the administration of d-amphetamine (2.0mg/kg) or APO (0.1 mg/kg), a total of 40 image frames were collected over 20 min to determine the baseline state. Following injection of d-amphetamine or APO, an additional 40 frames were collected to track the dynamic response (Zhang *et al.*, 2001; Andersen *et al.*, 2002). The change in R2

i.e. R2\* which represents the phMRI activation response to drug, was determined as the difference between the mean R2\* across 20 images post drug administration during the period of peak response (5-15 min) and the mean R2\* within the 40 baseline images. A reduction ("negative" change) in R2\* associated with a local decrease of paramagnetic

Six out of six animals responded positively to APO treatment represented by 44% improvements in parkinsonian symptoms. The same dose of APO also evoked phMRI responses by increasing the phMRI signal intensity. The typical phMRI (BOLD effect) responses to APO were gradually increased after APO administration only in the structure on the ipsilateral side receiving MPTP administration. Interestingly, but not surprising, APOinduced behavioral changes (PD features) were significantly correlated with APO-induced phMRI responses in the putamen, premotor cortex, and cingulate gyrus. When compared with standard but objective measures, there was a significant negative correlation between the phMRI responses in the putamen and distance travelled and movement speed. Similar relationships were also seen between phMRI responses in the motor cortex and daytime homecage activity and between phMRI responses in the caudate nucleus and movement speed.

Fig. 3. phMRI reveals nigrostriatal system responsiveness to dopamine stimulation. Coronal MRI scans depicting areas of activation and deactivation (represented by the psedudocolor)

in the brain after an APO or amphetamine challenge in unitalteral MPTP-lesioned

nonhuman primates (from Zhang *et al.*, 2006).

deoxyhemoglobin is an indicator of BOLD-effect activation (Chen *et al.*, 1996).

**2.3 phMRI-responses correlate with severity of PD** 

\*,

Apomorphine administration strongly activated the MPTP-denervated putamen (Figs. 3A and 4C) and substantia nigra (Fig.4D). An opposite response (a positive ΔR2\* value) was evident in the contralateral putamen (Fig. 4G) and substantia nigra (Fig. 4H). The differences between the intact and lesioned substantia nigra and between the intact and lesioned putamen were highly significant, P < 0.01(*t*-test), in both cases. In contrast, ΔR2\* responses in the caudate nucleus and in the corpus callosum were not significant, nor were there significant hemispheric differences in activation or deactivation with the contralateral caudate or with a comparable region in the contralateral callosum (Figs. 4A and 4E).

The phMRI responses to amphetamine treatment in the putamen (Figs. 3B and 4G) and substantia nigra (Fig. 4H) were the inverse of those seen with apomorphine. Amphetamineinduced decreases (positive ΔR2\* values) in the lesioned putamen and substantia nigra suggested diminished neuronal activity in both sites. In contrast, amphetamine induced the opposite ΔR2\* response in the intact left side, tending to increase activation in the putamen and substantia nigra. The responses in the intact putamen and intact substantia nigra were significantly different from their lesioned counterparts. Again, the corpus callosum and the caudate nucleus displayed only small, insignificant changes in response to amphetamine stimulation (Figs. 4E and 4F).

Fig. 4. phMRI responses in the nigrostriatal system. Depending on the means of stimulation, phMRI reveals a differential activations and deactivations in the nigrostriatal system. After APO stimulation (A-D) or *d*-amphetamine stimulation (E-H). \*\**P*<0.01; \**P*<0.05; unpaired *t*test (from Zhang *et al.*, 2006).

#### **2.4.2 phMRI-responses and loss of DA neurons in the SN**

In a later study, post-mortem histopathological evaluation revealed that the unilateral MPTP administration (received 5 years before the analysis) produced a massive (85%) loss of the rate-limiting enzyme for DA formation, tyrosine hydroxylase, (TH+) cells in the midbrain on the ipsilateral side receiving the infusion of the neurotoxin. TH+ cell numbers were significantly higher on the un-lesioned side compared to the MPTP-lesioned side. More importantly, the number of TH+ cells was strongly correlated with the phMRI responses in

Developing an MRI-Based Biomarker for Early Diagnosis of Parkinson's Disease 123

!

Fig. 6. Hemiparkinsonian nonhuman primates have markedly dimished dopaminergic function. K+ (100 mM)- and amphetamine (250 µM)-evoked DA release was significantly attenuated in the ipsilesional A) putamen (Put) and B) SNc; \*\*\*: P < 0.0001 (paired *t*-test).

the putamen (each measured for a single time point, 30 minutes after stimulus administration) had significant correlations with phMRI responses in the putamen. DA levels in the putamen were also significantly correlated with phMRI responses in the premotor cortex and cingulate gyrus, as well as in the caudate nucleus. Finally, damphetamine-evoked DA release in the SNc was found to have a significant, but negatively



There is a great need for the development of noninvasive, highly sensitive, and widely available imaging methods which can potentially be used to longitudinally monitor treatment of PD. We reported the monitoring of glial-cell-line-derived neurotrophic factor (GDNF) induced functional changes of the basal ganglia in hemiparkinsonian monkeys via phMRI measuring the BOLD response to a direct dopamine agonist, APO, (Luan *et al.*, 2008). The effectiveness of GDNF to protect and restore the nigrostriatal dopaminergic system in rodent and nonhuman primate models of PD has been extensively documented (Beck *et al.*,

Fig. 7. DA levels in the right SNc correlate with the BOLD responses in the right motor cortex. In animals with lower DA levels in the right SNc, less activation was observed in the

**3. Using phMRI to monitor therapeutic effects in parkinsonian monkeys** 

 

  \$ ---

\$ ---

#" #" #" #"

correlated relationship with the motor cortex (Fig. 7).

 

 -

\$ ---

right motor cortex

!

 

\$ ---

 

the caudate nucleus and in the cingulate gyrus. When comparing d-amphetamine-induced DA release in the putamen and DA neuron counts in the SNc, a significant correlation was also seen. In an earlier study (Zhang *et al.*, 2006), amphetamine administration evoked a BOLD response in the SN that correlated with the number of TH+ dopamine neurons in the same structure. These data support that there is a strong relationship between BOLDresponses to dopaminergic challenge and the number of dopaminergic neurons in the midbrain.

#### **2.4.3 phMRI-responses and loss of DA fibers in the striatum**

Similar to the effect on dopaminergic neurons, the MPTP administration also produced a remarkable reduction of TH+ fibers on the ipsilateral side of the lesion. A comparison of the fiber density in the putamen on the MPTP-lesioned side with other elements of the corticobasal ganglia-cortical circuit (Braak & Del Tredici, 2008) such as ipsilateral phMRI responses in the motor cortex (Fig. 5A) and caudate nucleus (Fig. 5B) showed strong correlations. In addition, the fiber density in the MPTP-lesioned caudate nucleus was strongly correlated with phMRI responses in the premotor cortex, caudate nucleus, and cingulate gyrus. Those changes in TH+ fiber density were also correlated with behavior and DA levels in the striatum and with the number of DA neurons in the SNc.

Fig. 5. Lower TH+ fiber density in the ipsilesional putamen corresponds with higher phMRI activation. TH+ fiber density in the right putamen (R-Put) is inversely correlated with phMRI activation in A) the right motor cortex (R-MC) and B) the right caudate nucleus (R-CD).

#### **2.4.4 phMRI-responses correlate with dopamine overflow**

The microdialysis experiments were conducted months after the parkinsonian symptoms had been fully developed and stabilized. First, the single administration of MPTP produced significant reduction in both potassium- and d-amphetamine-evoked overflow of DA in the putamen (Fig. 6A) and SNc (Fig.6B) on the ipsilateral side of the lesion. Second, there were several important correlations between DA levels in the putamen and SNc and the phMRI responses. For example, both potassium- and d-amphetamine-evoked overflow of DA in

the caudate nucleus and in the cingulate gyrus. When comparing d-amphetamine-induced DA release in the putamen and DA neuron counts in the SNc, a significant correlation was also seen. In an earlier study (Zhang *et al.*, 2006), amphetamine administration evoked a BOLD response in the SN that correlated with the number of TH+ dopamine neurons in the same structure. These data support that there is a strong relationship between BOLDresponses to dopaminergic challenge and the number of dopaminergic neurons in the

Similar to the effect on dopaminergic neurons, the MPTP administration also produced a remarkable reduction of TH+ fibers on the ipsilateral side of the lesion. A comparison of the fiber density in the putamen on the MPTP-lesioned side with other elements of the corticobasal ganglia-cortical circuit (Braak & Del Tredici, 2008) such as ipsilateral phMRI responses in the motor cortex (Fig. 5A) and caudate nucleus (Fig. 5B) showed strong correlations. In addition, the fiber density in the MPTP-lesioned caudate nucleus was strongly correlated with phMRI responses in the premotor cortex, caudate nucleus, and cingulate gyrus. Those changes in TH+ fiber density were also correlated with behavior and DA levels in the

> # ' \$(#),#\$\*)+( ' (&%\$( #\$

> >

!# ' \$(#),-

 

Fig. 5. Lower TH+ fiber density in the ipsilesional putamen corresponds with higher phMRI activation. TH+ fiber density in the right putamen (R-Put) is inversely correlated with phMRI activation in A) the right motor cortex (R-MC) and B) the right caudate nucleus (R-

The microdialysis experiments were conducted months after the parkinsonian symptoms had been fully developed and stabilized. First, the single administration of MPTP produced significant reduction in both potassium- and d-amphetamine-evoked overflow of DA in the putamen (Fig. 6A) and SNc (Fig.6B) on the ipsilateral side of the lesion. Second, there were several important correlations between DA levels in the putamen and SNc and the phMRI responses. For example, both potassium- and d-amphetamine-evoked overflow of DA in

 

&")#+)#%\$-

 -

 

**2.4.3 phMRI-responses and loss of DA fibers in the striatum** 

striatum and with the number of DA neurons in the SNc.
