**2.1 Computed Tomography (CT)**

In the last two decades, the first important data about neuroanatomic abnormalities in BD were obtained by means of CT. More recently, the widespread use of MRI has brought several advantages over CT, particularly in terms of higher resolution of images of subcortical regions (Steffens, 1998). Although a typical pattern of abnormality has not been identified yet (Supprian, 2004), several brain structures were found to be affected in patients with BD according to imaging studies.

CT provides excellent imaging data and rapid image acquisition at relatively low cost, it is widely available and more easily tolerated by patients, remaining the imaging modality of first choice in many clinical situations (Dougherty et al., 2004).

CT consists of a series of slices or tomograms. Its measurements are performed at the periphery of the body. The image of each slice is acquired by means of an X-ray source and detectors positioned at 180 degrees on the other side of the body. By spinning the source and the detectors on one plane of the head, data are collected from multiple angles. A computer then processes X-ray attenuation measured from different points and uses specific algorithms to create a structural image within the plane. Ionic and non-ionic intravenous

aforementioned core mood symptoms and clinical features of BD, over the last decade, neurocognitive dysfunction has been stressed as another nuclear dimension of BD and, possibly, a marker of its underlying pathophysiology (Lewandowski et al., 2010). There is accumulating evidence that individuals with BD have neurocognitive impairment that persists even during euthymia: the degree of impairment is more severe in patients with depressive symptoms, with functions associated with processing speed and attentional control being particularly implicated (Chaves et al., 2011; Van der Werf-Eldering et al., 2010). In addition, in older euthimic adults with BD, resting-state corticolimbic dysregulation was related to sustained attention deficits and inhibitory control, which could reflect the cumulative impact of repeated affective episodes upon cerebral metabolism and neurocognitive performance (Brooks et al., 2011). Cognitive impairment in BD is influenced

In addition, neuropsychological and imaging studies in BD suggested the presence of cognitive deficits and subtle magnetic resonance imaging (MRI) changes in limbic areas that may persist over euthymia. However, other studies are inconsistent with this claim. For example, a recent study did not identify any difference between BD patients and controls in levels of cognition over a two-year period, indicating that BD doesn't have a significant

Neuroimaging has recently gained an important role both in clinical practice and research of psychiatric disorders, including BD. Structural imaging techniques such as computed tomography (CT), magnetic resonance imaging (MRI) and magnetic resonance spectroscopy (MRS) have contributed to a deeper understanding of the structural changes in the brain in the context of psychiatric disorders. Positron Emission Tomography (PET), Single Photon Emission Computed Tomography (SPECT), Functional Magnetic Resonance Imaging (fMRI) and Diffusion Tensor Imaging (DTI) are techniques which measure changes in response to cognitive demand and/or connectivity between brain regions. As such, these approaches provide an opportunity for investigating the neural bases of behavioural and cognitive

In the last two decades, the first important data about neuroanatomic abnormalities in BD were obtained by means of CT. More recently, the widespread use of MRI has brought several advantages over CT, particularly in terms of higher resolution of images of subcortical regions (Steffens, 1998). Although a typical pattern of abnormality has not been identified yet (Supprian, 2004), several brain structures were found to be affected in patients

CT provides excellent imaging data and rapid image acquisition at relatively low cost, it is widely available and more easily tolerated by patients, remaining the imaging modality of

CT consists of a series of slices or tomograms. Its measurements are performed at the periphery of the body. The image of each slice is acquired by means of an X-ray source and detectors positioned at 180 degrees on the other side of the body. By spinning the source and the detectors on one plane of the head, data are collected from multiple angles. A computer then processes X-ray attenuation measured from different points and uses specific algorithms to create a structural image within the plane. Ionic and non-ionic intravenous

by the severity of illness (Yates et al., 2010).

adverse impact on cognition (Delaloye et al., 2011).

impairment in psychiatric populations, including BD.

first choice in many clinical situations (Dougherty et al., 2004).

**2. Structural neuroimaging 2.1 Computed Tomography (CT)** 

with BD according to imaging studies.

contrasts can be used to improve the visualization of certain normal or abnormal structures (Dougherty et al., 2004).

The measurement of total brain volume and ventricular volumes has been the aim of the first investigations using CT in psychiatric disorders. In this perspective, less consistent results have been found for affective disorders compared to schizophrenia and dementia (Beyer et al., 2002). The limited number of controlled CT studies focused on bipolar patients, in fact, showed heterogeneous findings. These include increased lateral ventricle size compared to controls (Andreasen et al., 1990; Nasrallah et al., 1982; Pearlson et al., 1984) or, in contrast, non significant differences between patients and controls (Dewan et al., 1988; Schlegel et al., 1987; Young et al., 1999). A larger third ventricle has been reported as well (Dewan et al., 1988; Schlegel et al., 1987). Studies on cortical alterations in BD revealed that there was no significant difference between patients and controls with respect to the level of cortical atrophy (Iacono et al., 1988; Rieder et al., 1983; Schlegel et al., 1987). However, a positive correlation between increased cortical sulcal widening and age of onset/age of first manic episode has been observed in bipolar patients in a subsequent study (Young et al., 1999). Volumetric changes in the cerebellum have been also reported, including higher rates of atrophy in bipolar patients (Nasrallah et al., 1982), even though the research of these abnormalities is limited.

In synthesis, some studies using CT in bipolar patients found an increased lateral ventricles size. In addition, cortical atrophy (which was not statistically different from controls), atrophy in the cerebellum as well as a larger third ventricle have also been reported.

#### **2.2 Magnetic Resonance Imaging (MRI)**

MRI takes advantage of the magnetic properties of the atomic constituents of the tissues in order to create an image of the different parts of the body. Every MRI scanner has a static magnet; its strength usually ranges from 1.5 to 3 Tesla. A steady magnetic field is generated as an electric current passes through the coils. In order to have a nuclear magnetic resonance signal, only atomic nuclei with unpaired protons and/or neutrons can be used. Medical MRI uses essentially hydrogen (¹H) as it is widely diffused in the human body and it has only one proton in its nucleus. Each proton has its own magnetic field or dipole moment, induced by the rotation around its axis. When an externally magnetic field is applied, protons' magnetic dipoles tend to align and to oscillate around the longitudinal axis of the applied field (this phenomenon is called precession) (Dougherty et al., 2004).

An horizontal radio frequency (RF) pulse is applied perpendicularly to the longitudinal axis of the external magnetic field with the aim to create a transverse component to the magnetization vector. This induces the generation of an electric current which is transduced into an MRI image. T1 is the "longitudinal" relaxation time and it indicates the time required to regain longitudinal magnetization following RF pulse. T2 is the "transverse" relaxation time that measures how long the resonating protons precess "in phase" following a 90° RF pulse. Due to the T1 and T2 relaxation properties in MRI, differentiation between various tissues in the body is possible (Jezzard et al., 2001).

Despite intensive research, to date no pathognomonic structural MRI finding has been correlated with affective disorders in general and to BD in particular. There are many heterogeneous data (Table 1) revealing a variety of structural alterations in bipolar patients (Dougherty et al., 2004). It must be considered, moreover, that some of these differences may be referred to the effects of medications (Van der Schot, 2009). For instance, chronic lithium treatment may prevent volume loss in treated patients because of its neuroprotective action

Neuroimaging Data in Bipolar Disorder: An Updated View 267

Frontal lobes Reduced or unchanged volume Temporal lobes Reduced or unchanged volume

Insular cortex Increased gray matter or no changes

Lateral ventricles Increased (association with number of

Amygdala Larger, smaller or unchanged volume

Caudate nucleus Increased or unchanged volume

Subcortical areas Increased hyperintensities

Third ventricle Increased

Corpus callosum Reduced volume

Magnetic Resonance Spectroscopy (MRS) is an MRI complement and serves as a noninvasive tool for tissue characterization. While MRI uses the signal from hydrogen protons to create a visual representation of the tissues, proton MRS (1H- MRS) uses this information to determine the concentration of brain metabolites such as N-acetyl aspartate, choline,

MRS has been principally used for the diagnosis of some metabolic disorders, especially those of the central nervous system. MRS has not an optimal specificity, but in association with MRI and clinical data can be very helpful. Indeed, the main purpose of this technique is to obtain biochemical information from any part of the body in a non invasive way, i.e. not

In psychiatry, MRS can be employed to assess the activity of different neurotransmitters, membrane and second messenger metabolism. The uniqueness of MRS is to provide an overview of the biochemical pathology of BD. Studies using proton MRS (1H- MRS) reported increased glutamate and GLX (glutamate, GABA and glutamine) levels in the dorsolateral prefrontal cortex, frontal lobes, basal ganglia and gray matter of medicationfree bipolar subjects and in patients with acute mania (Yildiz-Yesiloglu & Ankerst, 2006). Abnormal levels of N-acetyl aspartate, choline and myo-inositol have also been reported (Scherk et al., 2008b). N-acetyl aspartate seems to be reduced in the prefrontal cortex and hippocampus in bipolar individuals. Choline levels were found to be increased in the striatum and anterior cingulate cortex and can be normalized or decreased after treatment with antidepressants and lithium (Moore et al., 2000). Myo-inositol levels were increased in individuals with mania and euthymia and, on the contrary, reduced in bipolar depression.

by means of radioactive tracers or electromagnetic radiation (Dougherty et al., 2004).

Cerebellar posterior vermis Reduced

**Main MRI alterations in BD** 

Loss of the symmetry

episodes of mania)

Loss of normal symmetry

**Central nervous system structure involved** 

Table 1. Main MRI findings in BD.

**2.3 Magnetic Resonance Spectroscopy** 

creatine and lactate in the examined tissue (Gujar et al., 2005).

(Manji et al., 2000). Furthermore, genetic and/or environmental factors involved in BD may influence some brain abnormalities. In this perspective, decreases in white matter have been associated with the genetic risk of developing BD, whereas important environmental correlations have been found in relation to cortical gray matter volume (Van der Schot, 2009).

Brain abnormalities reported by fMRI studies in patients with BD include changes in cortical volumes, cerebral white matter, cortical and prefrontal gray matter. Enlargement of the ventricles, dimensional modifications of the amygdala, nuclei of the basal ganglia, corpus callosum and cerebellum have also been detected.

Main findings on lobar volumes concern frontal, temporal and insular cortex. Results on frontal lobes are quite discordant. In fact, they were found to be smaller (Coffman et al., 1990; Schlaepfer et al., 1994) or of the same size as controls (Strakowski et al., 1999). With respect to temporal lobes, no differences (Johnstone et al. 1989), bilateral reduction of volume (Altshuler et al., 1991) or loss of normal symmetry were found. Even in terms of loss of symmetry of the temporal lobes findings were sometimes discordant. In fact, a study reported a larger right temporal lobe than the left one in male bipolar patients (Swayze et al., 1992) and another study observed a larger left temporal lobe (Harvey et al., 1994). Voxelbased morphometric (VBM) MRI studies showed an increased gray matter in the insular cortex (Lochhead et al., 2004) or non significant differences in this region (McDonald et al., 2005; Nugent et al., 2006; Scherk et al., 2008a). An inverse correlation has been observed between the volume of the anterior insular cortex and the lifetime number of depressive episodes (Takahashy, 2010).

Bipolar patients, in particular those with late onset, were found to have a higher incidence of subcortical hypertensities (Dupont et al., 1990; Figiel et al., 1991; McDonald et al., 1991; Norris et al., 1997; Soares & Mann, 1997; Stoll et al., 2000; Swayze et al.,1990; Videbech, 1997). On the other hand, another study (Botteron & Figiel, 1997) identified an increased rate of white matter hyperintensity in relatively young individuals.

Lateral ventricular enlargement has been observed in BD and associated with multiple episodes of mania (Strakowski et al., 2002). A larger third ventricle was reported in elderly depressive patients and in cases of first manic episode (Strakowski et al., 1993). Likewise, correlations have been found between third ventricle volume and psychotic symptoms, advanced age, late onset of the disease, male gender and positive dexamethasone suppression test (Benabarre et al., 2002).

Studies on alterations of the amygdala in bipolar patients reported heterogeneous results, showing normal (Swayze et al., 1992), smaller (Pearlson et al., 1997) or larger volumes (Altshuler et al., 1998). More recent studies documented an increased volume in the right amygdala (Bremner et al., 2000), in bilateral amygdala in first episode subjects (Frodl et al., 2002) and loss of normal symmetry (Mervaala et al., 2000). The heterogeneity of the adult studies may be referred to the different age of subjects. It is still unclear, however, the positive correlation between increased amygdala volume and age (Usher, 2010).

A greater caudate volume as well as asymmetries among the structures of the basal ganglia were found in male bipolar patients (Aylward et al., 1994). Another study focused on the caudate volume in manic subjects in their first episode, reporting no significant differences vs healthy controls (Strakowski et al., 1999). The alterations may be attributed to a secondary effect of neuroleptic drugs (Benabarre et al., 2002). Studies examining alterations of the corpus callosum found volume reduction in bipolar patients, correlated with greater global neuropsychological dysfunction (Coffman et al., 1990). Finally, significant reduction of the cerebellar posterior vermis area was reported in patients with BD (DelBello et al., 1999).

(Manji et al., 2000). Furthermore, genetic and/or environmental factors involved in BD may influence some brain abnormalities. In this perspective, decreases in white matter have been associated with the genetic risk of developing BD, whereas important environmental correlations have been found in relation to cortical gray matter volume (Van der Schot, 2009). Brain abnormalities reported by fMRI studies in patients with BD include changes in cortical volumes, cerebral white matter, cortical and prefrontal gray matter. Enlargement of the ventricles, dimensional modifications of the amygdala, nuclei of the basal ganglia, corpus

Main findings on lobar volumes concern frontal, temporal and insular cortex. Results on frontal lobes are quite discordant. In fact, they were found to be smaller (Coffman et al., 1990; Schlaepfer et al., 1994) or of the same size as controls (Strakowski et al., 1999). With respect to temporal lobes, no differences (Johnstone et al. 1989), bilateral reduction of volume (Altshuler et al., 1991) or loss of normal symmetry were found. Even in terms of loss of symmetry of the temporal lobes findings were sometimes discordant. In fact, a study reported a larger right temporal lobe than the left one in male bipolar patients (Swayze et al., 1992) and another study observed a larger left temporal lobe (Harvey et al., 1994). Voxelbased morphometric (VBM) MRI studies showed an increased gray matter in the insular cortex (Lochhead et al., 2004) or non significant differences in this region (McDonald et al., 2005; Nugent et al., 2006; Scherk et al., 2008a). An inverse correlation has been observed between the volume of the anterior insular cortex and the lifetime number of depressive

Bipolar patients, in particular those with late onset, were found to have a higher incidence of subcortical hypertensities (Dupont et al., 1990; Figiel et al., 1991; McDonald et al., 1991; Norris et al., 1997; Soares & Mann, 1997; Stoll et al., 2000; Swayze et al.,1990; Videbech, 1997). On the other hand, another study (Botteron & Figiel, 1997) identified an increased rate

Lateral ventricular enlargement has been observed in BD and associated with multiple episodes of mania (Strakowski et al., 2002). A larger third ventricle was reported in elderly depressive patients and in cases of first manic episode (Strakowski et al., 1993). Likewise, correlations have been found between third ventricle volume and psychotic symptoms, advanced age, late onset of the disease, male gender and positive dexamethasone

Studies on alterations of the amygdala in bipolar patients reported heterogeneous results, showing normal (Swayze et al., 1992), smaller (Pearlson et al., 1997) or larger volumes (Altshuler et al., 1998). More recent studies documented an increased volume in the right amygdala (Bremner et al., 2000), in bilateral amygdala in first episode subjects (Frodl et al., 2002) and loss of normal symmetry (Mervaala et al., 2000). The heterogeneity of the adult studies may be referred to the different age of subjects. It is still unclear, however, the

A greater caudate volume as well as asymmetries among the structures of the basal ganglia were found in male bipolar patients (Aylward et al., 1994). Another study focused on the caudate volume in manic subjects in their first episode, reporting no significant differences vs healthy controls (Strakowski et al., 1999). The alterations may be attributed to a secondary effect of neuroleptic drugs (Benabarre et al., 2002). Studies examining alterations of the corpus callosum found volume reduction in bipolar patients, correlated with greater global neuropsychological dysfunction (Coffman et al., 1990). Finally, significant reduction of the cerebellar posterior vermis area was reported in patients with BD (DelBello et al., 1999).

positive correlation between increased amygdala volume and age (Usher, 2010).

callosum and cerebellum have also been detected.

of white matter hyperintensity in relatively young individuals.

episodes (Takahashy, 2010).

suppression test (Benabarre et al., 2002).


Table 1. Main MRI findings in BD.

#### **2.3 Magnetic Resonance Spectroscopy**

Magnetic Resonance Spectroscopy (MRS) is an MRI complement and serves as a noninvasive tool for tissue characterization. While MRI uses the signal from hydrogen protons to create a visual representation of the tissues, proton MRS (1H- MRS) uses this information to determine the concentration of brain metabolites such as N-acetyl aspartate, choline, creatine and lactate in the examined tissue (Gujar et al., 2005).

MRS has been principally used for the diagnosis of some metabolic disorders, especially those of the central nervous system. MRS has not an optimal specificity, but in association with MRI and clinical data can be very helpful. Indeed, the main purpose of this technique is to obtain biochemical information from any part of the body in a non invasive way, i.e. not by means of radioactive tracers or electromagnetic radiation (Dougherty et al., 2004).

In psychiatry, MRS can be employed to assess the activity of different neurotransmitters, membrane and second messenger metabolism. The uniqueness of MRS is to provide an overview of the biochemical pathology of BD. Studies using proton MRS (1H- MRS) reported increased glutamate and GLX (glutamate, GABA and glutamine) levels in the dorsolateral prefrontal cortex, frontal lobes, basal ganglia and gray matter of medicationfree bipolar subjects and in patients with acute mania (Yildiz-Yesiloglu & Ankerst, 2006). Abnormal levels of N-acetyl aspartate, choline and myo-inositol have also been reported (Scherk et al., 2008b). N-acetyl aspartate seems to be reduced in the prefrontal cortex and hippocampus in bipolar individuals. Choline levels were found to be increased in the striatum and anterior cingulate cortex and can be normalized or decreased after treatment with antidepressants and lithium (Moore et al., 2000). Myo-inositol levels were increased in individuals with mania and euthymia and, on the contrary, reduced in bipolar depression.

Neuroimaging Data in Bipolar Disorder: An Updated View 269

Functional neuroimaging attempts to explain psychiatric disorders by means of

In fact, elevated activity of the hippocampus or of the ventral prefrontal cortex as well as dorsolateral prefrontal cortex hypofunction are recurrent themes in literature (Savitz &

PET imaging is a direct measure of a radioactive decay due to cerebral metabolism of a radioactive substance or radionuclide. Different body tissues are characterized by different consumption rates of radionuclides (Ter-Pogossian et al., 1975; Vyas et al., 2011). Radionuclides used in clinical practice are usually major compounds of biologic molecules (18-Fluorine in the form of 18-Fluorodeoxyglucose or FDG for measuring glucose metabolism, 15-Oxygen for measuring blood flow, 11-Carbon or 13-Nitrogen common in diagnostic PET procedures). The nuclide is introduced in the patient and the radioactive decay is measured (Phelps et al., 1975): in particular, the positron emitted by nuclides has a collision with electrons producing a gamma photon which is measured by the PET camera (Roncali & Cherry, 2011). PET can measure both blood flow and glucose metabolism, often used as surrogate measures of neuronal synaptic activity. A first line comparison is between the neuroligand uptake in target regions and reference area while a more complex analysis can compare blood flow or glucose uptake in the same subject in different states, i.e. while resting or during a cognitive performance. Both ways provide useful data for research and clinical analysis; anyway, a major limitation is the use of a radioactive nuclide. Specifically, targeted PET radioligands are used to investigate neurotransmitter systems (Weisel, 1989). Cerebral PET has its major use in neurological disease: excluding primary or secondary oncologic lesions, evaluation of dementia, confirming epilepsy or assessing the state in cerebrovascular disease (Cavalcanti et al., 2011; Mazzuca et al., 2011; Person et al., 2010;

SPECT works capturing orbiting electrons without a positron-electron collision, but by means of an emission of a single photon by the SPECT nuclide. Main nuclides used in SPECT are 123-Iodine, 33-Technetium or 133-Xenon. Single photons are selected with the

PET and SPECT studies in depressive disorders have shown that blood rate and flow are increased both in BD and in unipolar depression in the frontal lobes during depressive episodes. However, they are increased during mania in the dorsal cingulate cortex, striatal regions, and the nucleus accumbens, as well as in limbic structures of the temporal lobes and reduced in dorsolateral prefrontal cortex, possibly reflecting its loss of modulatory

With respect to neurotransmitters, serotonin (5-HT) transporter was found to have an increased density in the thalamus (Laje et al., 2010), dorsal cingulate cortex, medial prefrontal cortex and insula of depressed BD patients. 5-HT has been implicated in mania as well: in particular, individuals with current mania had significantly lower 5-HT2 receptor binding potential in frontal, temporal, parietal and occipital cortical regions, with more prominent changes in the right cortical regions compared to controls (Yatham et al., 2010, 2002a, 2002b). With regards to 5-HT1A receptor, bipolar depressed patients were found to show higher 5-HT1A in raphe nuclei and forebrain (Sullivan et al., 2009). An interesting use of PET consists of assessing the role of serotonin in major depressive episodes comparing

degenerative or developmental model of illness and/or in terms of hypometabolism.

**3.1 Positron Emission Tomography (PET) and Single Photon Emission Computed** 

Drevets, 2009).

**Tomography (SPECT)** 

Quigley et al., 2010; Salas and Gonzales, 2011).

control over limbic structures (Gonul et al., 2009).

use of multiple collimators.

Studies using phosphorus MRS (31P- MRS) have found phase-specific alterations of phospholipid membranes, high energy phosphates and intracellular brain pH in BD. In particular, a number of investigations reported a reduced intracellular cerebral pH in bipolar subjects which has been associated with the increased levels of lactate observed in some 1H- MRS studies. Both conditions are indicative of a shift from oxidative phosphorylation to glycolysis. There is also a 31P- MRS based-report of decreased levels of phosphocreatine and of phosphomonoesters in BD (Kato et al., 1995).

Stork and Renshaw proposed a cohesive model that puts together the majority of MRS findings. They hypothesized that the impaired oxidative phopshorilation, the decreased cellular energy and the altered membrane metabolism could be due to an underlying altered mitochondrial metabolism in BD (Stork & Renshaw, 2005).

Main MRS findings in BD are synthetized in Table 2.


Table 2. Main MRS findings in BD.
