**2.1 X-radiography**

*Data Acquisition - Recent Advances and Applications in Biomedical Engineering*

smooth movements, pain, stiffness and reduction in joint function.

affecting over 18% of the population in England [5].

aspiration of future interventions for the disease.

disease and requiring joint replacement.

**2. Standard methods of detection**

GP visits and hospital admissions [6–8].

**1.2 Epidemiology and impact**

interactions between cartilage and bone, along with bone and bone, make for less

The most common joints affected by osteoarthritis include those of the knee, hip and hands with osteoarthritis of the knee the most commonly occurring form,

The underlying pathophysiology of osteoarthritis is unclear, with genetics, age, gender, obesity and previous injury all contributing to varying degrees in disease development and progression. The heterogeneous nature of the disease makes targeted treatment of cause and prevention of progression a challenge, with current best practice centring on patient education and lifestyle changes surrounding exercise, use of analgesics and anti-inflammatories to manage pain and inflammation and finally joint replacement at the severe end of the spectrum of disease [9]. However, this approach, with the exception of exercise targeting weight loss and strength, does not address an underlying cause or prevent progression of disease, an

Ranking the sixth most common cause of disability globally in 2010 [10], musculoskeletal conditions, including osteoarthritis, impact not only the healthcare system and patients but also their families [11]. Patients and their carers are at greater risk of being out of employment [12], with only 63% of those with a musculoskeletal condition in employment compared to 82% in those without a health condition [13]. With a predicted increase in the ageing population and an increase in obesity [14–16], the burden on health services and economic impact in terms of lost work time and disability is of growing concern. There is a real need for means of noninvasive early detection of osteoarthritis, sensitive means of monitoring progression and development of efficacious treatments to prevent and improve symptoms in order to improve quality of life and reduce the numbers progressing to severe

Osteoarthritis is a condition affecting a multitude of tissues within a joint, and as such, approaches which give information to the clinician on bone, muscle, cartilaginous tissue and the microenvironment within a joint are required to give a full picture of the condition of a joint. Imaging is currently the main diagnostic tool used to assess osteoarthritis. Dependent upon the form of imaging used, a variety of

In clinical practice, a combination of clinical presentation and X-radiography (X-ray) is used to diagnose osteoarthritis. When a patient presents as over 45 years of age, with typical symptoms of osteoarthritis including pain within the joint during activity and minimal stiffness within the joint in the morning lasting no more

However, X-ray is useful when differential diagnosis is possible, and in certain scenarios, magnetic resonance imaging is used to give additional information on

tissues can be examined as markers of disease state and progression.

than 30 min, then X-ray is not indicated for diagnosis [9, 17].

damage to tissues within the joint and inform treatment options.

With such a large proportion of the population affected, musculoskeletal conditions including osteoarthritis have considerable impact both medically and economically. Clinically, the pain and loss of function associated with osteoarthritis result in a lower quality of life reported by patients, who require a large number of

**32**

X-ray works upon the principle of differential absorbance of radiation by different tissues, with dense tissues such as the bone absorbing a large proportion of the radiation compared to soft tissues such as the muscle and connective tissue.

As a result, the bone appears bright white on images and can be studied for changes in morphology, whereas soft tissues show less differentiation and are not easily examined.

The current gold standard in the diagnosis of osteoarthritis from radiographic images involves the scoring of X-ray images using the Kellgren-Lawrence (KL) scale. The Kellgren-Lawrence is a five-point scale which categorises disease severity based upon the assessment of bony changes, appearances of osteophytes and joint space narrowing within the joint [18]. The description of the radiographic findings at different KL grades can be seen in **Table 1**.

The KL scale was first described in 1957 in response to an identified need to standardise the definition of changes within an osteoarthritic joint in order to improve inter-rater reliability when reporting the disease [18]. Thorough analysis of the performance of the scale at joints throughout the body revealed that whilst correlation between the defined changes and osteoarthritis were observed at all joints bars the wrist, the greatest inter-rater agreement was found within the knee joint. Intra-rater repeatability followed a similar trend with slightly better agreement between readings. This has subsequently been reflected in the most common use of the scale in the assessment of the knee joint.

More recent comparison of radiographic scoring systems has established that for the knee joint, the KL scale has stood the test of time, with no subsequently developed grading systems outperforming the inter-rater repeatability of this scale [19]. However, whilst the limit of inter and intra-observer reliability in assessing radiographic osteoarthritis may have been reached (correlation coefficients around 0.8), it is acknowledged that a more diverse manner of assessment of osteoarthritis may be warranted to improve sensitivity when assessing disease progression and specificity for aspects of the homogeneous pathophysiology underlying the disease.

In terms of sensitivity, KL scoring of radiographs does not perform well in the detection of early disease or in the monitoring of disease progression, where large time periods are required to observe a change in category during which time symptomatic progression may have occurred [20].

Alone, radiographic assessment using the Kellgren-Lawrence scale allows direct assessment of bony changes such as osteophyte formation, however, relies on indirect measures of joint space narrowing to assess cartilaginous change. The surrogate marker of joint space narrowing in place of direct measurement of cartilage, whilst


**Table 1.** *Kellgren-Lawrence scale description of radiographic findings.* important in the sensitivity of Kellgren-Lawrence scale to disease severity, does not perform well when compared with changes observed arthroscopically [19, 21].

This may go some way to explaining the disparity in patient symptom reporting in the form of self-reported osteoarthritis, clinically diagnosed osteoarthritis and disease severity suggested using the Kellgren-Lawrence scale [22]. In addition to indirect cartilage measurements, the Kellgren-Lawrence score is based solely on the femorotibial joint. As osteoarthritis can also affect the patellofemoral joint, this could account for further disparity between symptoms and radiographic severity of disease [20].

#### **2.2 Magnetic resonance imaging**

In contrast to X-radiography, magnetic resonance imaging (MRI) can directly image a number of tissues, including the cartilage, bone and fluids such as that found in the synovium. Several approaches have been taken to the assessment of joints with suspected osteoarthritis using MRI.

A number of joint-specific semi-quantitative scoring systems have been developed using features considered important in osteoarthritis disease manifestation, including bone marrow lesions, meniscal scores and scores of cartilage loss. For the knee, the scoring systems developed include the whole-organ MRI score (WORMS), the knee osteoarthritis scoring system (KOSS), the Boston-Leeds OA knee scoring (BLOKS) and the MRI osteoarthritis knee score (MOAKS), which brings together the strengths of the WORMS and BLOKS systems whilst standardising the definitions used [23].

Quantitative analysis of specific tissues has also been used to measure thickness, area and volume of cartilage, bone area and area of the bone that is denuded, as well as combining the two to assess cartilage thickness over areas of denuded bone. Whilst concentrating on a smaller region of the joint, this approach removes some of the subjectivity associated with the semi-quantitative scores detailed above, both for MRI and X-ray scoring [23–26].

The benefits of MRI for use both clinically and within research are a trade-off between increased sensitivity and specificity and protocols which are realistic for application in a given setting. Semi-quantitative MRI protocols can be performed using clinical MRI equipment, however, have the same caveats of KL scoring of X-rays in terms of inter and intra-rater reliability.

Quantitative measures of the cartilage and bone remove some of the subjective elements of semi-quantitative assessment. The changes of cartilage and bone measurements can be exceedingly small in magnitude, allowing assessment of much smaller anatomical change over shorter timeframes than those observed using X-ray. Making such small measurements presents its own challenges and is timeconsuming, whilst producing such small measurements of change that relationship to clinical outcomes can be weak [27]. However, being direct in nature, quantitative measures have shown promise in improving association of imaging techniques with disease symptoms and progression compared with KL scoring of X-rays. Denuded bone area has been shown to correlate with concurrent and incident knee pain [28], whilst changes in cartilage thickness have been linked to the likelihood of disease progression to the point of needing knee joint replacement surgery [29, 30].

In addition to semi-quantitative and quantitative measurements, the use of contrast and powerful MRI imaging protocols extend the means to assess tissue, enabling assessment of components of the ultrastructure of articular cartilage and the meniscus along with the synovial fluid via compositional and diffusion MRI, respectively. This makes MRI a potentially powerful tool in assessing the impact of osteoarthritis on the entirety of a joint, as well as in identifying factors driving disease and predicting disease progression.

**35**

*Acoustic Monitoring of Joint Health*

accepted in clinical practice [31].

*DOI: http://dx.doi.org/10.5772/intechopen.92868*

post-processing that is required limits use clinically.

**2.3 Other biomarkers of osteoarthritis**

to disease progression than X-ray or MRI.

High-resolution MRI protocols and high doses of contrast prove most useful in research aimed at understanding of the mechanisms of osteoarthritis and assessment of disease progression or slowing with intervention. However, these are time-consuming protocols and contrast doses can far outstrip recommended doses

The added power of MRI in the assessment of osteoarthritis is most likely to remain predominantly within the research field at this point in time, as access to advanced equipment, lack of uniform protocols and the time-consuming nature of

Whilst X-Ray and MRI are the two primary forms of imaging used to assess osteoarthritic joints, both computer tomography (CT) and ultrasound have also been employed for this purpose, generally in a research setting, where MRI is proving to provide greatest accuracy [32]. For CT, the use is limited due to CT scans delivering a high radiation dose without delivering significantly greater sensitivity

Whilst ultrasound allows direct imaging of the cartilage which is not obtained during X-ray, interpretation and observations made can vary between operators, especially at joints further from the surface of the skin. This is least marked in superficial joints, and assessment of inflammation and effusion has drawn parallels with disease severity and progression [33–35]. Therefore, ultrasound may be most useful in adding measures associated with inflammation when assessing joints of

Finally, biochemical markers associated with inflammation and degradation of the bone and cartilage are under investigation as additional biomarkers for osteoarthritis. This presents its own challenges as whilst these markers may well be sensitive to change in internal environment, their specificity to osteoarthritis and location of degeneration are proving more of obstacle, with generally weak associations seen between biochemical biomarkers of disease and measures of use in assessing disease severity and progression [36, 37]. That said, there is some evidence that markers may be able to offer additional strength in assessing osteoar-

Individually the current means to diagnose and assess progression of osteoarthritis are limited by one or more factors, namely, subjectivity of measures including high inter- and intra-rater repeatability in semi-quantitative imaging, low sensitivity for change in disease state or low specificity for disease tissue or location. This presents challenges when making informed clinical decisions, investigating new interventions and determining the effects of preventative measures on disease progression. The low sensitivity of current biomarkers also limits the application of stratified medicine in the approach to new treatments, an area that is of particular interest given the marked clinical and biological heterogeneity of this condition [39]. As the disease is driven by multiple pathogenic factors, it may be that a combination of multiple diagnostic measures is required to develop a sensitive biomarker for osteoarthritis. This concept is currently demonstrated through the development of computational risk factor tools based on a range of self-reported osteoarthritis risk factors, aimed at patient education and pre-emptive lifestyle intervention [40–42]. More recently, the tool for osteoarthritis risk prediction has proven inclusion of MRI measures in combination with KL scored radiographs provides a more powerful predictive

the hand rather than the knee and hip which are much deeper joints.

thritis severity and response to treatments with further research [38].

**2.4 Current challenges in diagnosis and treatment**

#### *Acoustic Monitoring of Joint Health DOI: http://dx.doi.org/10.5772/intechopen.92868*

*Data Acquisition - Recent Advances and Applications in Biomedical Engineering*

**2.2 Magnetic resonance imaging**

ing the definitions used [23].

for MRI and X-ray scoring [23–26].

X-rays in terms of inter and intra-rater reliability.

disease and predicting disease progression.

joints with suspected osteoarthritis using MRI.

important in the sensitivity of Kellgren-Lawrence scale to disease severity, does not perform well when compared with changes observed arthroscopically [19, 21].

This may go some way to explaining the disparity in patient symptom reporting in the form of self-reported osteoarthritis, clinically diagnosed osteoarthritis and disease severity suggested using the Kellgren-Lawrence scale [22]. In addition to indirect cartilage measurements, the Kellgren-Lawrence score is based solely on the femorotibial joint. As osteoarthritis can also affect the patellofemoral joint, this could account for further disparity between symptoms and radiographic severity of disease [20].

In contrast to X-radiography, magnetic resonance imaging (MRI) can directly image a number of tissues, including the cartilage, bone and fluids such as that found in the synovium. Several approaches have been taken to the assessment of

Quantitative analysis of specific tissues has also been used to measure thickness, area and volume of cartilage, bone area and area of the bone that is denuded, as well as combining the two to assess cartilage thickness over areas of denuded bone. Whilst concentrating on a smaller region of the joint, this approach removes some of the subjectivity associated with the semi-quantitative scores detailed above, both

The benefits of MRI for use both clinically and within research are a trade-off between increased sensitivity and specificity and protocols which are realistic for application in a given setting. Semi-quantitative MRI protocols can be performed using clinical MRI equipment, however, have the same caveats of KL scoring of

Quantitative measures of the cartilage and bone remove some of the subjective elements of semi-quantitative assessment. The changes of cartilage and bone measurements can be exceedingly small in magnitude, allowing assessment of much smaller anatomical change over shorter timeframes than those observed using X-ray. Making such small measurements presents its own challenges and is timeconsuming, whilst producing such small measurements of change that relationship to clinical outcomes can be weak [27]. However, being direct in nature, quantitative measures have shown promise in improving association of imaging techniques with disease symptoms and progression compared with KL scoring of X-rays. Denuded bone area has been shown to correlate with concurrent and incident knee pain [28], whilst changes in cartilage thickness have been linked to the likelihood of disease progression to the point of needing knee joint replacement surgery [29, 30]. In addition to semi-quantitative and quantitative measurements, the use of contrast and powerful MRI imaging protocols extend the means to assess tissue, enabling assessment of components of the ultrastructure of articular cartilage and the meniscus along with the synovial fluid via compositional and diffusion MRI, respectively. This makes MRI a potentially powerful tool in assessing the impact of osteoarthritis on the entirety of a joint, as well as in identifying factors driving

A number of joint-specific semi-quantitative scoring systems have been developed using features considered important in osteoarthritis disease manifestation, including bone marrow lesions, meniscal scores and scores of cartilage loss. For the knee, the scoring systems developed include the whole-organ MRI score (WORMS), the knee osteoarthritis scoring system (KOSS), the Boston-Leeds OA knee scoring (BLOKS) and the MRI osteoarthritis knee score (MOAKS), which brings together the strengths of the WORMS and BLOKS systems whilst standardis-

**34**

High-resolution MRI protocols and high doses of contrast prove most useful in research aimed at understanding of the mechanisms of osteoarthritis and assessment of disease progression or slowing with intervention. However, these are time-consuming protocols and contrast doses can far outstrip recommended doses accepted in clinical practice [31].

The added power of MRI in the assessment of osteoarthritis is most likely to remain predominantly within the research field at this point in time, as access to advanced equipment, lack of uniform protocols and the time-consuming nature of post-processing that is required limits use clinically.

### **2.3 Other biomarkers of osteoarthritis**

Whilst X-Ray and MRI are the two primary forms of imaging used to assess osteoarthritic joints, both computer tomography (CT) and ultrasound have also been employed for this purpose, generally in a research setting, where MRI is proving to provide greatest accuracy [32]. For CT, the use is limited due to CT scans delivering a high radiation dose without delivering significantly greater sensitivity to disease progression than X-ray or MRI.

Whilst ultrasound allows direct imaging of the cartilage which is not obtained during X-ray, interpretation and observations made can vary between operators, especially at joints further from the surface of the skin. This is least marked in superficial joints, and assessment of inflammation and effusion has drawn parallels with disease severity and progression [33–35]. Therefore, ultrasound may be most useful in adding measures associated with inflammation when assessing joints of the hand rather than the knee and hip which are much deeper joints.

Finally, biochemical markers associated with inflammation and degradation of the bone and cartilage are under investigation as additional biomarkers for osteoarthritis. This presents its own challenges as whilst these markers may well be sensitive to change in internal environment, their specificity to osteoarthritis and location of degeneration are proving more of obstacle, with generally weak associations seen between biochemical biomarkers of disease and measures of use in assessing disease severity and progression [36, 37]. That said, there is some evidence that markers may be able to offer additional strength in assessing osteoarthritis severity and response to treatments with further research [38].

#### **2.4 Current challenges in diagnosis and treatment**

Individually the current means to diagnose and assess progression of osteoarthritis are limited by one or more factors, namely, subjectivity of measures including high inter- and intra-rater repeatability in semi-quantitative imaging, low sensitivity for change in disease state or low specificity for disease tissue or location.

This presents challenges when making informed clinical decisions, investigating new interventions and determining the effects of preventative measures on disease progression. The low sensitivity of current biomarkers also limits the application of stratified medicine in the approach to new treatments, an area that is of particular interest given the marked clinical and biological heterogeneity of this condition [39].

As the disease is driven by multiple pathogenic factors, it may be that a combination of multiple diagnostic measures is required to develop a sensitive biomarker for osteoarthritis. This concept is currently demonstrated through the development of computational risk factor tools based on a range of self-reported osteoarthritis risk factors, aimed at patient education and pre-emptive lifestyle intervention [40–42]. More recently, the tool for osteoarthritis risk prediction has proven inclusion of MRI measures in combination with KL scored radiographs provides a more powerful predictive

tool for predicting disease progression [43]. Furthering this approach using other potential biomarkers for osteoarthritis, including imaging and biochemical markers of cartilage and bone change, may allow even greater sensitivity and specificity.

With this in mind, research has progressed in innovative approaches to develop biosensors that address aspects of osteoarthritis that are currently unmeasured. To date, all biomarkers for the disease consider circulating biochemicals or images of the knee in a static state. As the symptoms of osteoarthritis relate directly to movements of the joint, a novel approach to assessing changes in interactions between tissues during joint movement is being investigated using acoustics within the joint.

### **3. Acoustic medical technologies for joint health**

Due to its non-invasive nature, the use of sound or vibration has found many medical applications associated with the musculoskeletal system.

For instance, as discussed above, ultrasound imaging, or ultrasonography (US), can be a useful tool in rheumatology. It is increasingly used to image and evaluate the inflammatory aspects of rheumatic diseases as an assessment tool for tendons and soft tissue [44, 45]. It has been applied to osteoarthritis specifically, having been shown to be a sensitive tool for the evaluation of synovitis (joint inflammation) and joint effusion (the flow of blood and other fluids in joints), through direct imaging and the use of Doppler signal analysis, a form of flow velocimetry [44–48]. Whilst US can be used for imaging musculoskeletal changes in osteoarthritis, such as changes in cartilage thickness, it is limited. It has been noted that US may be limited in assessing cartilage in larger weight-bearing joints [49] because of the inherent inability of ultrasound to pass through denser bony structures and therefore penetrate to the deeper portions of the joint [50]. The central portion of thick joints cannot be visualised with US [51], but US can detect osteophytosis (bone spurs forming around joints) at greater rates than conventional radiography. Being non-ionising and able to image soft tissues, US is a good alternative to radiographic imaging. Magnetic resonance imaging (MRI) offers excellent tissue contrast and anatomical resolution compared to US [49]. MRI can detect changes in the volumes of cartilage, whereas US is only capable of quantifying changes in thicknesses. Therefore, whilst MRI is more expensive, US is primarily only used as an alternative for anatomical imaging when there is hardware present within the patient, i.e. implants and some older cardiac defibrillators and pacemakers, which precludes the use of MRI [52].

As well as for imaging, ultrasound can be utilised directly as a treatment for OA [53, 54]. The management of OA involves the relief of pain and the maintenance or improvement of joint function. The American College of Rheumatology (ACR) and the European League Against Rheumatism (EULAR) recommend a combination of pharmacological and nonpharmacological treatments [55]. Various nonpharmacological treatments, including exercise, physical therapy, hot packs and therapeutic ultrasound (TU) etc., exist with varying evidence of efficacy. In TU, mechanical energy in the form of pulsed or continuous high-frequency vibrations is applied directly to the joint [56]. This is reputed to reduce oedema or cysts [57], as well as reduce inflammation, relieve pain and accelerate tissue repair; however, results of clinical studies are conflicting [55, 56]. The applied ultrasonic vibrations cause atomic oscillations in the tissue; the amplitude of which depends on the intensity or power of the applied beam. When applied continuously, this can result in thermal effects in the tissue, which are reduced when the beam is pulsed [56]. When the ultrasonic beam has high intensity, the atoms in the attenuating medium no longer oscillate around their equilibrium position but have a net motion along the axis of the beam [53]. This can result in damage or micro-machining due to the

**37**

*Acoustic Monitoring of Joint Health*

facilitating local delivery.

**3.1 Acoustic detection**

*3.1.1 Phonoarthrography*

*DOI: http://dx.doi.org/10.5772/intechopen.92868*

ultrasound-induced forces, allowing TU to be used as a surgical tool [53]. Highintensity TU can also result in the movement of particles and fluid within the tissue. This phenomenon has been used to drive pharmaceuticals, such as non-steroidal anti-inflammatory drugs (NSAIDs) and corticosteroids, into the tissue [58, 59],

Spontaneous emission of acoustic waves and other vibrations has been recorded during the flexion and extension of joints, as well as the fracture and wear of bones and implants [60, 61]. Studies have shown that these vibrations are affected by musculoskeletal disorders in joints, making vibration monitoring a useful diagnostic tool [62]. However, joints are highly complex heterogeneous structures over a wide range of length scales. Parameters like wave velocity, dispersion and attenuation all affect how waves travel through tissues, making interpretation of the waveform complicated. The following techniques have been developed to resolve this issue:

The earliest studies on the monitoring of the spontaneous emission of acoustic waves were based on the use of stethoscopes to amplify audible sounds generated within joints [63, 64]. Early joint auscultation in this manner was initially a manual process and was inherently subjective. Still, these studies showed that whilst there are 'normal joint sounds', the sound produced is affected by different kinds of injury and arthritis [65]. That said, this method is not yet used in primary care and has only received modest attention in the literature since its first appearance in 1902 [63, 66]. Later studies attempted to reduce the subjectivity of this method by recording the sounds using microphones in conjunction with joint measurement technologies such as goniometers and video tracking [67]. Several of these studies note that pathological signals have major frequency components at low frequencies, that is, below 1000 Hz [64, 68]. The sensitivity range of the microphones used is usually in the range 50 Hz to 15 kHz; however, it has been suggested that standard acoustic recording microphones are not appropriate for the monitoring of joint signals, being too sensitive to background noise, with vibration transducers, or contact sensors, and accelerometers being preferred [61, 69]. Studies such as that by Chu et al. employed a differential microphone pair for noise cancellation and bandpass filters to minimise low-frequency movement artefacts and high-frequency transducer noise to mitigate this issue [61]. Conversely, other studies [70] suggest that as microphones are able to detect higher frequencies and no direct contact with the body is required, the combination of signals from both microphones and acceler-

Data analysis in early studies generally only used traditional stationary spectrum estimation methods using oscilloscopes or narrow-band spectrum analysers, with key measures being the frequency, wavelength, wave number and amplitude [64]. However, it is clear that the signals are nonstationary in nature, especially as different signals are generated at different joint positions [69]. As a result of this observation, more sophisticated spectral analysis methods were developed. One method is short-time Fourier analysis on segmented data where it is assumed that the data is stationary within each segment. This allows trends in the frequency component of the signal to be correlated with joint angle. The determination of the segments introduces subjectivity into the analysis. Therefore, techniques to track the nonstationarities in the signal, such as adaptive segmentation, linear prediction and autoregressive moving averages (ARMA), have been incorporated into the analysis [69].

ometers might perform better than anyone signal alone.

#### *Acoustic Monitoring of Joint Health DOI: http://dx.doi.org/10.5772/intechopen.92868*

ultrasound-induced forces, allowing TU to be used as a surgical tool [53]. Highintensity TU can also result in the movement of particles and fluid within the tissue. This phenomenon has been used to drive pharmaceuticals, such as non-steroidal anti-inflammatory drugs (NSAIDs) and corticosteroids, into the tissue [58, 59], facilitating local delivery.

#### **3.1 Acoustic detection**

*Data Acquisition - Recent Advances and Applications in Biomedical Engineering*

**3. Acoustic medical technologies for joint health**

medical applications associated with the musculoskeletal system.

tool for predicting disease progression [43]. Furthering this approach using other potential biomarkers for osteoarthritis, including imaging and biochemical markers of

With this in mind, research has progressed in innovative approaches to develop biosensors that address aspects of osteoarthritis that are currently unmeasured. To date, all biomarkers for the disease consider circulating biochemicals or images of the knee in a static state. As the symptoms of osteoarthritis relate directly to movements of the joint, a novel approach to assessing changes in interactions between tissues during joint movement is being investigated using acoustics within the joint.

Due to its non-invasive nature, the use of sound or vibration has found many

For instance, as discussed above, ultrasound imaging, or ultrasonography (US), can be a useful tool in rheumatology. It is increasingly used to image and evaluate the inflammatory aspects of rheumatic diseases as an assessment tool for tendons and soft tissue [44, 45]. It has been applied to osteoarthritis specifically, having been shown to be a sensitive tool for the evaluation of synovitis (joint inflammation) and joint effusion (the flow of blood and other fluids in joints), through direct imaging and the use of Doppler signal analysis, a form of flow velocimetry [44–48]. Whilst US can be used for imaging musculoskeletal changes in osteoarthritis, such as changes in cartilage thickness, it is limited. It has been noted that US may be limited in assessing cartilage in larger weight-bearing joints [49] because of the inherent inability of ultrasound to pass through denser bony structures and therefore penetrate to the deeper portions of the joint [50]. The central portion of thick joints cannot be visualised with US [51], but US can detect osteophytosis (bone spurs forming around joints) at greater rates than conventional radiography. Being non-ionising and able to image soft tissues, US is a good alternative to radiographic imaging. Magnetic resonance imaging (MRI) offers excellent tissue contrast and anatomical resolution compared to US [49]. MRI can detect changes in the volumes of cartilage, whereas US is only capable of quantifying changes in thicknesses. Therefore, whilst MRI is more expensive, US is primarily only used as an alternative for anatomical imaging when there is hardware present within the patient, i.e. implants and some older cardiac defibrillators and pacemakers, which precludes the use of MRI [52]. As well as for imaging, ultrasound can be utilised directly as a treatment for OA [53, 54]. The management of OA involves the relief of pain and the maintenance or improvement of joint function. The American College of Rheumatology (ACR) and the European League Against Rheumatism (EULAR) recommend a combination of pharmacological and nonpharmacological treatments [55]. Various nonpharmacological treatments, including exercise, physical therapy, hot packs and therapeutic ultrasound (TU) etc., exist with varying evidence of efficacy. In TU, mechanical energy in the form of pulsed or continuous high-frequency vibrations is applied directly to the joint [56]. This is reputed to reduce oedema or cysts [57], as well as reduce inflammation, relieve pain and accelerate tissue repair; however, results of clinical studies are conflicting [55, 56]. The applied ultrasonic vibrations cause atomic oscillations in the tissue; the amplitude of which depends on the intensity or power of the applied beam. When applied continuously, this can result in thermal effects in the tissue, which are reduced when the beam is pulsed [56]. When the ultrasonic beam has high intensity, the atoms in the attenuating medium no longer oscillate around their equilibrium position but have a net motion along the axis of the beam [53]. This can result in damage or micro-machining due to the

cartilage and bone change, may allow even greater sensitivity and specificity.

**36**

Spontaneous emission of acoustic waves and other vibrations has been recorded during the flexion and extension of joints, as well as the fracture and wear of bones and implants [60, 61]. Studies have shown that these vibrations are affected by musculoskeletal disorders in joints, making vibration monitoring a useful diagnostic tool [62]. However, joints are highly complex heterogeneous structures over a wide range of length scales. Parameters like wave velocity, dispersion and attenuation all affect how waves travel through tissues, making interpretation of the waveform complicated. The following techniques have been developed to resolve this issue:

#### *3.1.1 Phonoarthrography*

The earliest studies on the monitoring of the spontaneous emission of acoustic waves were based on the use of stethoscopes to amplify audible sounds generated within joints [63, 64]. Early joint auscultation in this manner was initially a manual process and was inherently subjective. Still, these studies showed that whilst there are 'normal joint sounds', the sound produced is affected by different kinds of injury and arthritis [65]. That said, this method is not yet used in primary care and has only received modest attention in the literature since its first appearance in 1902 [63, 66].

Later studies attempted to reduce the subjectivity of this method by recording the sounds using microphones in conjunction with joint measurement technologies such as goniometers and video tracking [67]. Several of these studies note that pathological signals have major frequency components at low frequencies, that is, below 1000 Hz [64, 68]. The sensitivity range of the microphones used is usually in the range 50 Hz to 15 kHz; however, it has been suggested that standard acoustic recording microphones are not appropriate for the monitoring of joint signals, being too sensitive to background noise, with vibration transducers, or contact sensors, and accelerometers being preferred [61, 69]. Studies such as that by Chu et al. employed a differential microphone pair for noise cancellation and bandpass filters to minimise low-frequency movement artefacts and high-frequency transducer noise to mitigate this issue [61]. Conversely, other studies [70] suggest that as microphones are able to detect higher frequencies and no direct contact with the body is required, the combination of signals from both microphones and accelerometers might perform better than anyone signal alone.

Data analysis in early studies generally only used traditional stationary spectrum estimation methods using oscilloscopes or narrow-band spectrum analysers, with key measures being the frequency, wavelength, wave number and amplitude [64]. However, it is clear that the signals are nonstationary in nature, especially as different signals are generated at different joint positions [69]. As a result of this observation, more sophisticated spectral analysis methods were developed. One method is short-time Fourier analysis on segmented data where it is assumed that the data is stationary within each segment. This allows trends in the frequency component of the signal to be correlated with joint angle. The determination of the segments introduces subjectivity into the analysis. Therefore, techniques to track the nonstationarities in the signal, such as adaptive segmentation, linear prediction and autoregressive moving averages (ARMA), have been incorporated into the analysis [69].

#### *3.1.2 Vibroarthrography (VAG)*

Whilst phonoarthrography is based on the sound produced during the flexion or extension of joints, in VAG all vibrations produced during movement are considered [62]. Consequently, it is more common for a single accelerometer to be used as the sensor rather than a microphone [71]. It is also very common for signals in a frequency range below 1000 Hz to be of primary focus [72], with sampling rates of the order 1–4 kHz. A key advantage of the low sampling rate is that it allows for wireless data acquisition and processing using simple microcontrollers or singleboard computers [73, 74]. That said, it has been suggested [71, 75] that single signal processing may be limited and multi-channel recordings may lead to better discrimination of the severity and location of joint injury or disorder. In many cases noise mitigation is achieved through prefiltering (commonly using a bandpass filter from 10 Hz to 1 kHz) and amplification prior to digitization at a specified sampling rate [76, 77]. The digital signal may go through additional filtering, such as that conducted by Andersen et al. [78] who used a Kaiser-windowed finite impulse response (FIR) bandpass filter.

There are other rationales for using multiple sensors during VAG as it has been observed that VAG may pick up vibrations not necessarily just due to the joint directly or to external interference [79]. For instance, the 10 Hz signal generated by the rectus femoris muscle which activates during the extension of the leg could interfere with the VAG signal recorded from the skin surface over the patella [80]. As this signal may vary in a similar fashion to the VAG signal, simple bandpass filtering may not be sufficient. It may be necessary to record the vibromyogram at the rectus femoris at the same time as the VAG signal and use adaptive filtering and noise cancellation techniques to isolate the VAG signal [79].

Therefore, the VAG signal is inherently nonstationary and potentially multicomponent in nature. The nature of the VAG signal means that it is not easily analysed using common signal processing techniques. This coupled with the difficulty in ascertaining the biological origin of the source of the signal is the main barrier to its use as a common diagnostic tool. As a result, much of the recent research activity has been focussed on feature extraction and statistical pattern classification [60]. Adaptive segmentation using least-square, linear prediction and autoregression algorithms is common [81, 82]. A host of statistical measures has been considered to characterise the VAG signal, including the form factors, skewness, kurtosis and entropy [71, 76]. It has also been shown that time-frequency distribution (TFD) [81, 83] and wavelet decomposition [84] are potentially powerful techniques for analysis and may negate the need for segmentation [83] but may be susceptible to noise [85]. These advancements have mostly been driven by developments in digital signal processing technologies that sped up analysis time as well as nonstationary signal analysis techniques developed for other biological signals like EEGs [84].

Using these techniques, spectral features such as frequency, energy and their respective spreads can be classified and linked to joint position, loading and pathology. The commonly used classifiers are neural network-based classifiers and support vector machines (SVM), as well as logistic regression and rule-based techniques [62, 71]. These neural networks and SVMs are supervised learning algorithms which search for a number of independent training data patterns taken from signals measured from participants with known pathologies to characterise new signals. These classification algorithms are increasingly dependable and can perform well with a limited amount of data. A number of different variants of these algorithms and classifiers have been investigated [60, 62]. Wu et al. [73] used an SVM based on the entropy and envelope amplitude features and achieved an overall accuracy of 83.56%. Nalband et al. [86] utilised an a priori algorithm with

**39**

*Acoustic Monitoring of Joint Health*

*3.1.3 Acoustic emission (AE)*

involves the use of ultrasonic sensors.

*DOI: http://dx.doi.org/10.5772/intechopen.92868*

a least-square SVM classifier and claim accuracy of 94.31% with a false discovery rate of 0.0892. Kręcisz [87] achieved accuracies of >90% using a logistic regressionbased method. In each of these cases, the VAG signals were collected during knee flexion/extension motion using an accelerometer secured to the participants patella.

AE for biomedical applications is derived from non-destructive techniques developed for detecting damage in engineering materials, such as metals and composites [88]. AE occurs when materials locally under stress emit energy in the form of transient elastic waves. This allows for the monitoring of microcrack initiation and propagation in the bones and joints [89]—essential parts of bone remodelling [90], and wear [91, 92]. Other characteristic sounds in joints, such as the bursting of gas bubbles in synovial joints during movement, can also be detected using AE [93]. AE frequencies are usually in the ultrasonic range and so detection often

A number of researchers have proposed AE sensor-based joint monitoring

Toreyin et al. [94, 95] used an off-the-shelf low-noise MEMS microphone in conjunction with gyroscope and accelerometer pairs in order to monitor sounds generated during various complex motions. The microphone used had a sensitivity range of 100 Hz to 10 kHz, and the researchers suggested that the MEMS-based microphone had a similar performance to an electret microphone [94]. The acoustic data were sampled at 100 kHz, and the inertial data (monitoring joint angle and limb movement) at 1 kHz, with the data being collected by a field programmable gate array (FPGA)-based real-time processor. It was noted that air microphones do not exhibit signal losses due to motion artefacts, but they are sensitive to ambient noise. Teague et al. [96] compared a piezoelectric film-based contact microphone to two air microphones: one electret and one MEMS-based. The air microphones were used with a 15 Hz high-pass filter and a second-order low-pass filter with a cut-off frequency of 21 kHz and sampled at 44.1 kHz using an acoustic recorder. The piezoelectric microphone was used with a 100 Hz high-pass filter followed by a fourth order low-pass filter with a 10 kHz cut-off frequency. It was sampled at 50 kHz using custom circuits. The 100 Hz high-pass filter was chosen to attenuate the motion artefact noise. It was noted that the electret and MEMS microphones performed similarly in detecting joint sounds, although the electret sensor was significantly more expensive. They were both sensitive to ambient and interface noise, including rubbing of the tape securing the sensors. It was noted that the air microphones did not need to be in contact with the skin. Experiments with sensors positioned 5 cm off the skin captured similar acoustic signals, albeit with lower amplitude. The piezoelectric sensor was more sensitive to interface noise but less sensitive to background noise. Importantly, the contact microphone did not pick up higher frequency vibrations as distinctly as the air microphones which provided

systems using piezoelectric films, electret or MEMS-based microphones.

higher quality recordings as indicated by higher SNIRs.

Jeong et al. [97] used a low-noise electret microphone with a frequency range of 50 Hz to 20 kHz recorded by an audio recorder at a rate of 44.1 kHz. Signals were digitally filtered using a finite impulse response bandpass filter with a bandwidth from 1 to 15 kHz to prioritise short duration joint sounds whilst supressing interface noise. Feng and Chen [98] developed a piezoelectric sensor comprised of a lead zirconium titanate (PZT) film deposited on titanium cantilever arrays as an acoustic sensing layer. This sensor uses a 1-mm-tall SU8 cylindrical probe on each cantilever to be in direct contact with the skin of the participant and transmit vibrations to the sensor. A thermoresponsive poly(N-isopropylacrylamide) (PNIPA) film was

a least-square SVM classifier and claim accuracy of 94.31% with a false discovery rate of 0.0892. Kręcisz [87] achieved accuracies of >90% using a logistic regressionbased method. In each of these cases, the VAG signals were collected during knee flexion/extension motion using an accelerometer secured to the participants patella.
