Musculoskeletal ultrasonography: Clinical applications

INTRODUCTION — Ultrasonography (US), also referred to as ultrasound imaging or sonography, is an imaging modality that utilizes reflected pulses of high-frequency (ultrasonic) sound waves to assess soft tissues, cartilage, bone surfaces, and fluid-containing structures. Ultrasonographic imaging, once the sole province of radiologists, is nowadays widely available in rheumatology clinics and other ambulatory and emergency settings. Technical aspects of musculoskeletal ultrasonography, the validity and reliability of image acquisition, and interpretation in rheumatic disorders, primarily in patients with rheumatoid arthritis, are addressed in detail separately. (See "Musculoskeletal ultrasonography: Nomenclature, technical considerations, and basic principles of use".)

The use of US to assess patients with rheumatic diseases in the clinic was fostered by the development of compact real-time US systems in the 1980s. Synovitis of the knee was the earliest musculoskeletal disorder assessed ultrasonographically in the clinic [1]. Ultrasonographic assessment of synovitis of the small joints of the hands in patients with rheumatoid arthritis followed by a decade [2], aided by the availability in the 1990s of high-resolution transducers that made detailed assessment of superficial structures feasible.

Selected clinical applications of musculoskeletal US are discussed here. An overview of imaging modalities and guidelines for selecting imaging studies (eg, plain film radiography, computed tomography [CT], magnetic resonance imaging [MRI], and US) for musculoskeletal problems are presented separately. (See "Imaging techniques for evaluation of the painful joint" and "Imaging evaluation of the painful hip in adults" and "Radiologic evaluation of the painful shoulder in adults".)

Diagnostic imaging is only one element in the assessment of patients with musculoskeletal symptoms. The clinical approach to such patients, including, in most instances, suggestions regarding the role of diagnostic imaging in their assessment, is discussed in the topics that deal with the initial evaluation of patients presenting with pain in various parts of the body. Examples include shoulder pain, hip pain, and knee pain. (See "Evaluation of the adult with shoulder complaints" and "Approach to the adult with unspecified hip pain" and "Approach to the adult with unspecified knee pain".)

ACQUIRING IMAGES — For the musculoskeletal ultrasonographer, the selection of anatomic structures to image and the planes in which multiplanar ultrasonography (US) is performed is primarily complaint-driven. Standardized scans and disease-specific protocols for imaging are alternative approaches that are used in certain settings [3-5].

Complaint-driven image selection — Complaint-driven image selection means that decisions about which images to acquire and the order in which images are obtained are determined by the clinician's assessment of the likely causes of symptoms and by the examiner's assessment of the usefulness of particular images in supporting or excluding such causes.

Complaint-driven US scans focus on a specific region of the body (eg, the elbow in lateral epicondylitis or the metacarpophalangeal joints in rheumatoid arthritis). The position of the transducer depends upon the body region, while the view is manipulated by a combination of angulation, pressure, and rotation of the transducer from the same position.

The type of US imaging to employ (eg, B-mode, grayscale, Doppler, color Doppler, or power Doppler US), as well as the size and frequency of US to use, is also dependent upon the size and depth of structures being imaged and upon the type of pathologic changes that are considered to be likely causes of patients' complaints.

Standardized scanning — Standardized scans are time-consuming and are more appropriate for training sessions and for use in clinical trials. Standardized scanning is typically the method of image selection that is utilized by an ultrasonographic technician. Images obtained using standardized views are subsequently interpreted by a radiologist.

Disease-specific scanning — Disease-specific protocols specify particular areas of the body and types of images to obtain based upon an established or suspected diagnosis. An example of a disease-specific protocol is scanning of the hips and shoulders for a patient with suspected polymyalgia rheumatica.

Scanning of the temporal arteries might also be considered part of disease-specific scanning when giant cell arteritis is suspected. (See "Diagnosis of giant cell arteritis", section on 'Ultrasound with Doppler'.)

ULTRASONOGRAPHY FOR SPECIFIC DISORDERS — Use of musculoskeletal ultrasonography (US) by the clinician may be valuable in disorders of the following structures:

●Tendons, ligaments, and muscles

●Entheses

●Bursae

●Peripheral nerves

●Joints and joint recesses

●Contour of bone

●Cartilage

In addition, real-time or preprocedure sonography may aid in needle guidance for aspiration and injection of bursae and joints.

Tendons, ligaments, and muscles — US and MRI are good imaging techniques for tendons, ligaments and muscles. Tendon pathology in inflammatory rheumatic disease is very common. There are two factors that contribute to the frequent involvement in this setting.

●First, tendons may be enveloped by a synovial sheath. If the synovial sheath is involved in the rheumatic inflammatory process, the tendon itself, although poorly vascularized, may also become inflamed. In inflammation, the tendon may undergo neovascularization [6]. In sheathed tendons, the nutrient vessels enter the tendon at specific sites called mesotendons (or vincula in the fingers), whereas in unsheathed tendons, blood vessels enter the tendon at randomly distributed sites. Sheathed tendons have a wide distribution (eg, the long head of the biceps brachii muscle at the shoulder; the extensor tendons and flexor tendons of the fingers at the wrist; the finger flexor tendons, which are sheathed starting at the metacarpophalangeal joints all the way down to their insertion; the separate sheath of the tendon of the flexor pollicis longus muscle; and all the flexor and extensor tendons around the ankle joint).

●Second, tendons, either sheathed or unsheathed, lie adjacent to bursae and joints, structures that are lined with synovial cells. Thus, tendons may become involved in the inflammatory process as innocent bystanders. This mechanism probably plays a predominant role in inflammation of unsheathed tendons, including the rotator cuff tendons, which lie adjacent to the shoulder joint and the subacromial-subdeltoid bursa; another example is inflammation of the Achilles tendon, which is adjacent to the retrocalcaneal bursa. At the wrist, extensor tendons may become attritioned by adjacent erosions of the distal ulna or radius.

Normal tendon and muscle — Sonography usually depicts a normal tendon as hyperechoic and fibrillar (image 1). However, a normal tendon may demonstrate the phenomenon of anisotropy (ie, areas of decreased echogenicity due to oblique beam artifact). Normal muscle is characterized as being predominantly hypoechoic with interspersed hyperechoic connective tissue on sonography and as having an intermediate signal on MRI. The transverse US scan of muscle is featured as a "starry night" (figure 1).

Abnormal tendon and muscle — Abnormal tendon is characterized by hypoechoic areas on sonography and by a fluid signal on MRI. Full-thickness tears are characterized by complete tendon absence or muscle disruption with retraction. The presence of an accompanying hematoma is characterized by anechoic or hypoechoic areas or by heterogeneous abnormal mixed echogenicity with a mass effect on sonography and with a fluid signal on MRI. Abnormal muscle, as in inflammatory myopathy, is sonographically characterized by greater echogenicity and diminished size of the hypoechoic muscle fibers separating the perimysial connective tissue [7].

An essential part of US examination of tendons is dynamically moving the joint or extremity (eg, finger tendons should be examined while bending and stretching of the fingers, and the rotator cuff tendons should be examined while rotating [internally and externally] and lifting [flexing and abducting] the arm). During dynamic examination, tendon movement and anatomical integrity may be assessed. Anatomic integrity is defined as continuity of the parallelly aligned tendon fibers, without interruption, along with a crisp outer tendon border. Examples of pathologic features that may be observed include the snapping phenomenon accompanying trigger finger and the bulging of the supraspinatus tendon on the antero-lateral edge of the acromion as a result of shoulder impingement.

Shoulder tendons — Tendons of the shoulder consist of those of the rotator cuff and the biceps brachii tendon (figure 2). The rotator cuff is best examined with transducer frequencies from 7.5 MHz to 15 MHz, although obese persons may require lower frequencies. In young athletes and older adults presenting with shoulder pain, cuff lesions should always be sought.

The most important and vulnerable tendon of the rotator cuff is the supraspinatus, which originates from the upper dorsal part of the scapula and inserts on the greater tuberosity of the humerus. The majority of cuff lesions occur in the tendon of the supraspinatus muscle, whereas approximately 30 percent are present in the subscapularis tendon [8]. Complete cuff tears are easily visualized, whereas partial tears may sometimes be difficult to diagnose. Sonographically, the rotator cuff tendons are best examined dynamically and in a standardized sequence, starting with the biceps tendon and then moving medially to assess the subscapularis tendon; subsequently, the supraspinatus tendon and the infraspinatus tendon are examined [3,4].

●Supraspinatus tendon – The vulnerable part of the supraspinatus tendon is a region of reduced vascularity, which is near the enthesis of the tendon at the greater tuberosity of the humerus [9]. In rheumatic diseases, inflammation of the shoulder joint or the bursa may result in chronic cuff edema, impingement, and, subsequently, a tendon tear.

The presence of a hypoechoic discontinuity within the contours of the cuff is diagnostic of a tendon tear. A full-thickness tear is present when the discontinuity extends from the joint capsular side to the bursal side (image 13B), and a partial tear is present when it is limited to a part of the tendon thickness (image 2). More subtle signs for rotator cuff tears, including both hyperechoic and hypoechoic areas, have been described [10-12].

Synovitis of the glenohumeral joint is often present, and signs are sonographically detectable fluid in the axillary pouch, in the subscapular-subdeltoid bursa, in the posterior recess, or in the biceps sheath. When fluid is present in the bursa, one has always to look for a rotator cuff tear.

●Biceps brachii tendon – The long biceps tendon is an important stabilizer of the glenohumeral joint. On sonography, a normal biceps is round-elliptical in cross-section and produces multiple echogenic dots (image 3). It may be surrounded by a thin hypoechoic rim representing normal physiologically present fluid. An inflamed biceps tendon is often thickened and hypoechoic, appears less round and more flattened, and may be surrounded by a larger hypoechoic rim (image 4). The synovial sheath may become thickened. Power Doppler ultrasonography (power Doppler US) may demonstrate positive signals in the hypoechoic rim, but the signal should not be mistaken for the normal flow adjacent to the tendon within the synovial sheath consistent with the anterolateral branch of the anterior circumflex humeral artery [13]. Rupture of the long head of the biceps tendon is a frequent finding in longstanding synovitis of the glenohumeral joint, whereas dislocation of the long head is infrequent in inflammatory rheumatic disorders. Both conditions can be diagnosed sonographically.

Finger tendons — Finger flexor tendon involvement is a common feature of both early and late rheumatoid arthritis. Ultrasonography of the finger tendons requires a linear transducer of at least 15 MHz. Grassi et al were the first to describe in detail the sonographic findings of both normal and abnormal finger flexor tendons [14]. The sagittal diameter of the finger flexors at the metacarpophalangeal joint varies from 3.4 to 4.2 mm [3].

High-frequency sonography can allow the superficial and the deep finger flexor tendons to be distinguished at the MCP joint. However, when tenosynovitis is present, the fluid may separate the superficial from the deep flexor tendon, making the distinction more clear [14].

●Inflammatory arthritis – The most common ultrasonographic sign of flexor tendon involvement is widening of the flexor tendon sheath (1.7 mm versus 0.2 mm in controls), which occurs in about 80 percent of patients with rheumatoid arthritis and psoriatic arthritis, followed by altered tendon echogenicity (in 60 percent) and irregularity of the tendon margin (50 percent) [14-16]. In inflammatory arthritis, there are two different patterns of tendon sheath widening, namely a hypoechoic enlargement correlating with effusion and an irregularly echoic widening due to proliferative synovitis (image 5 and image 6) [14].

●Trigger finger – Trigger finger (also called stenosing flexor tenosynovitis) may be assessed by US. Usually, the phenomenon of trigger finger is due to pathological thickening of the A1 pulley, a strong fibrous band, at the volar aspect of the metacarpophalangeal joint. On US, it appears as a hyperechoic band that is prone to anisotropy; dynamic examination may display the trigger phenomenon at this site. Power Doppler US is able to distinguish between hypoechoic fluid around the tendon and inflammatory or infectious tenosynovitis [13]. Blood flow in inflamed synovium is indicated by the colored regions in the peritendinous tissue (image 5). (See "Trigger finger (stenosing flexor tenosynovitis)".)

Wrist tendons — Inflammation of the “wrapped-up” extensor tendons at the wrist is frequently seen in patients with or without inflammatory rheumatic disease. Any of the six extensor compartments may become involved [17]. Inflammation of the first compartment gives rise to de Quervain's syndrome (entrapment tendonitis or tenosynovitis of the abductor pollicis longus and extensor pollicis brevis tendons at the styloid process of the radius), which generally occurs as an overuse syndrome (image 7 and image 8). By contrast, tenosynovitis of the extensor carpi ulnaris is a typical sign at the onset of rheumatoid arthritis and may predict progression of erosive disease [18].

Achilles tendon — The archetypal unsheathed tendon is the Achilles tendon. This tendon is the longest of the human body. It connects the calf muscles (ie, the gastrocnemius and the soleus) with the calcaneal bone. It can best be examined using a probe with a frequency between 7.5 and 13 MHz. Longitudinal sonography permits measuring the length and the sagittal diameter (picture 1 and image 9 and image 10). The mean length in adults is about 15 cm, and the sagittal diameter is about 4.3 mm. A bursa lies deep to the tendon at the enthesis, which is sonographically not visible when it is not inflamed.

Both partial and complete ruptures of the Achilles tendon may be assessed with US. Partial tears are more common than complete tears and are more difficult to diagnose clinically. Sonographic examination of artificial cuts in cadaveric tendons shows 81 percent accuracy between the size of the tear and the sonographic diagnosis [19]. In patients with Achilles tendon injuries, full thickness rupture is suggested by absence of tendon at the site of injury, by proximal tendon migration or retraction, and by posterior acoustic shadowing [20]. A common site of complete tears is about 5 cm proximal to the insertion onto the calcaneus. Furthermore, the Achilles tendon and its enthesis are frequently involved in the spondyloarthropathies, characterized by increased thickness, hypoechogenicity, and power Doppler signals.

Entheses — Tendons, ligaments, and joint capsules adhere to bone at a specific region, called the enthesis. Histologically the enthesis consists of an area of fibrocartilage (close to joints where tendons or ligaments are bent by shear forces) or of fiber-bone interface. To underpin the significance of this anatomical complex, the term enthesis organ complex has been coined to include not only the portion of the attachment between tendon and bone but also adjacent structures including bone, bone marrow, bursae, and adipose tissue [21].

Enthesitis — Peripheral enthesitis is one of the hallmarks of all of the spondyloarthritides (eg, undifferentiated spondyloarthritis, ankylosing spondylitis, psoriatic arthritis, and reactive arthritis). Sonographic findings of enthesitis include thickening or intratendinous focal changes of the tendon insertion, edema of the tendon insertion, calcific deposits at the tendon insertion, spurs, erosions, new bone formation, or periosteal changes (image 11). In addition, adjacent bursitis is considered a sign of enthesitis. A positive power Doppler US signal consistent with increased vascularity may be found in active enthesitis (image 12) [22].

Bursae — Under normal circumstances, bursae consist of two lining layers within only a very thin fluid lining between them; the fluid serves as a lubricant. Sonography usually depicts a bursa as a hypoechoic cleft between two hyperechoic lines. The hypoechoic cleft corresponds to the actual bursal sac, and the two hyperechoic lines correspond to the interfaces of the bursal sac with the surrounding tissues. However, under normal conditions, the subdeltoid bursa, the largest of its kind of human body, is hardly visible. On sonography, a dark semicircular line of 1 mm surrounds the rotator cuff in transverse view.

Bursae may become inflamed in inflammatory and degenerative rheumatic disease. There are a number of bursae that most commonly become inflamed (table 1). Some of these bursae communicate with the joint.

Subdeltoid-subacromial bursa — The subdeltoid-subacromial bursa is often visible in patients with inflammatory rheumatic disease. Sonographic images show a distended bursa, filled with anechoic or hypoechoic fluid (image 13A-B). On a longitudinal view, the distended bursa often takes the form of a droplet hanging over the lateral edge over the shoulder beneath the deltoid muscle.

As mentioned above, when sonographic evidence of subdeltoid-subacromial bursitis is present, one has to assess the presence of a rotator cuff tear, visible as a communicating cleft between the glenohumeral joint and the bursa (image 13B). Such evidence should be sought if not already noted when this area is examined.

In patients with rheumatoid arthritis or polymyalgia rheumatica [23,24], subacromial-subdeltoid bursal fluid may be present without an underlying tear in the rotator cuff. A distended bursa may also be noted in association with advanced age, longstanding osteoarthritis of the glenohumeral joint, or osteoarthritis of the acromioclavicular joint. In the case of acromioclavicular osteoarthritis, inferiorly directed osteophytes arising from the proximal acromion and/or the distal clavicle may result in increased friction and fraying of the rotator cuff, resulting in edema, impingement of the tendon, and, in some patients, subsequent cuff tear.

Olecranon bursa — Effusions within the superficially located olecranon bursa may be due to trauma, infection, or rheumatic diseases, particularly rheumatoid arthritis and gout. Sonography of septic bursitis or chronic bursitis due to rheumatic disease often shows a fluid collection surrounded by a thickened bursal wall. Power Doppler is usually highly positive. When the bursa is filled with fibrin-rich material or when the bursa is septated by fibrinous bands into several compartments, needle aspiration may result in a dry tap. Under these circumstances, ultrasound can be useful in steering the needle to an exact location or cavity filled with fluid.

Baker's cyst — The semimembranosus-gastrocnemius bursa, also referred to as Baker's cyst, was first sonographically demonstrated in 1972 [25] (see "Popliteal (Baker's) cyst"). The bursa deep to the tendon of the semimembranosus muscle anteriorly and the bursa below the medial head of the gastrocnemius muscle more cranially posteriorly may individually or jointly contribute to the formation of a Baker's cyst. Thus, on transverse sonograms, the Baker's cyst may take various forms, which include the following:

●If there is no connection between the medial gastrocnemius and semimembranosus bursae, only the medial gastrocnemius bursa is filled with fluid and then takes the form of a half-moon (image 14).

●If the two bursae communicate, the sonographic appearance of the cyst is similar to a so-called cross of St Andrew.

●In patients with rheumatoid arthritis, Baker's cysts can become septated and can have the US appearance of a bunch of grapes.

●If fluid accumulation is minimal, a fourth type of Baker's cyst (ie, the slit or crescent shape) is recognized [26,27].

A slit-like connection with the posterior joint space may be present in 30 to 50 percent of cases but is often difficult to detect sonographically. The communication pathway is through a 1.5 to 2 cm long tunnel in the medial part of the dorsal joint capsule, near the medial femur condyle, where the joint capsule is thinnest. The connection with the joint space accounts for the pumping of fluid into the bursa while bending the knee. When there is also a valve mechanism, the Baker's cyst will extend superficially and laterally into the loose connective tissue surrounding vessels and nerves.

The Baker's cyst position in a transverse view is between the tendons of the medial gastrocnemius anteriorly and the semimembranosus tendon posteriorly, adjacent to the femoral bone. Rupture of a Baker's cyst allows fluid to spread out downwards into the calf, usually following the fascial planes within or between the calf muscles. Sonography is able to detect the extended fluid collection within the calf area.

Peripheral nerves — Peripheral nerves may be imaged using a linear probe with frequencies between 10 and 16 MHz [28]. The longitudinal sonographic appearance of a normal peripheral nerve consists of parallel linear hypoechoic bands in a hyperechoic background. On transverse section, the image consists of hypoechoic rounded areas in a hyperechoic background. The hypoechoic structures correspond histologically to the neuronal fascicles. Typical applications in clinical practice are the entrapment neuropathies, including carpal tunnel syndrome, ulnar neuropathy due to entrapment at the cubital tunnel, and tarsal tunnel syndrome. Moreover, perineural fibromas such as Morton's neuroma can easily be visualized [29].

Joints — Sonography provides images of various joint components including the joint capsule, the joint recesses, the joint cartilage, and the bone surface. As an example, all recesses of the ankle joint including the anterior and posterior tibiotalar recesses, as well as the subtalar recess, are well-visualized by US if they are filled with fluid.

Bone — The contour of bone may be imaged by ultrasound and assessed for interruption of irregularities. For instance, stress fractures of ankle bones or metatarsal bones can be easily detected.

Bone erosions — In rheumatoid arthritis, pathologic changes may affect soft tissues, cartilage, and bone, but the hallmark of rheumatoid disease is bone erosion. Although plain film radiography has long been the standard imaging modality for detecting erosive disease, US and MRI are each able to demonstrate bone erosions in an early phase of rheumatoid arthritis with sensitivities that are significantly greater than that of conventional radiography. The following studies are illustrative:

●A prospective study compared conventional radiography, bone scintigraphy, US, and MRI in 60 patients with various forms of arthritis [30]. Both US and MRI detected significantly more erosions and synovitis in finger joints than did conventional radiography.

●Greater sensitivity of US over conventional radiography in the detection of bone erosions in metacarpophalangeal joints of patients with rheumatoid arthritis was corroborated in another study in which sonography detected 6.5-fold more erosions than radiography in early rheumatoid arthritis and 3.4-fold more erosions in late disease [31]. Moreover, the sonographic erosions corresponded to MRI bone abnormalities.

●The sensitivity to change for bone erosions was also assessed in a follow-up study, showing that US was superior to conventional radiography [32]. At sites at which US has limited access to cover the entire area of cartilage (eg, the shoulder), US is significantly inferior to MRI and is only slightly superior to conventional radiography in detecting bone erosions [33].

Synovitis — Clinically differentiating between a thickened synovial membrane and joint effusion is, at times, problematic. The echogenicity of grayscale (B-mode) images allows a preliminary assessment of a swollen joint. Synovitis is depicted as a hypoechoic area in relation to the isoechogenicity of connective tissue. Further information obtained by the use of color and power Doppler images facilitates differentiation between active and inactive synovitis. The ability of power Doppler US to demonstrate active inflammation suggests that this imaging modality may become an extremely important tool in assessing the amount of inflammation and the response to therapy in patients with inflammatory joint diseases (image 15). Assessing the activity of synovial inflammation by power Doppler may be graded semiquantitatively on a scale from zero to three (image 16) [34,35]. Furthermore, power Doppler US may be used for monitoring the clinical response to anti-tumor necrosis factor (TNF) therapy in patients with rheumatoid arthritis (image 17) [36]. Several studies suggest that US is superior to clinical examination in detecting synovitis [34,37-39]:

●When knee arthroscopy was used to determine the presence or absence of synovitis in patients with a variety of different arthritides, grayscale US was significantly better than clinical examination in detecting synovitis in the knee [37].

●In a study that utilized both gray scale and power Doppler US but that did not utilize MRI or arthroscopy as gold standards, US detected more synovitis in finger and toe joints in patients with rheumatoid arthritis than did clinical examination [38].

●Thickening of the synovial membrane, whether detected by clinical examination or by US, does not necessarily equal inflammation. Power Doppler US proved to be a reliable modality for assessing active metacarpophalangeal joint synovitis in patients with rheumatoid arthritis when compared with dynamic MRI as the gold standard [34].

●Evidence of a correlation between US detection of synovitis and actual pathology was provided by a study comparing the results of power Doppler US and histopathologic examination of synovial biopsy samples [39].

●Ultrasound assessment can also improve the accuracy of the 2010 American College of Rheumatology (ACR) and European Alliance of Associations for Rheumatology (EULAR; formerly known as European League Against Rheumatism) criteria for classifying patients as having rheumatoid arthritis, and hence, to start methotrexate treatment [40].

Cartilage — US is able to visualize normal and pathologic joint cartilage. Normal hyaline cartilage appears as a homogeneously anechoic band delineated by hyperechoic margins. Crystal arthropathies, such as chondrocalcinosis and gout, cause characteristic changes in cartilage that are visible with US (image 18) [41]. These disorders may be well visualized at the condyles of the knee or the first metatarsophalangeal joint. In addition, US is a valuable technique for evaluation of the various abnormalities in patients with osteoarthritis [42].

Additional ultrasound resources — Instructional videos demonstrating proper performance of the US examination and related pathology can be found at the website of the American Medical Society for Sports Medicine: principles of sports US and introduction to scanning techniques. Registration must be completed to access these videos, but no fee is required.

VALIDATION, RELIABILITY, AND STANDARDIZATION — A growing number of studies on validity and reliability of ultrasonography (US) in patients with rheumatic disease have been published. As a first step towards standardization, an international group of experts in the musculoskeletal US field proposed consensus-based definitions of rheumatoid arthritis erosions, synovitis, tenosynovitis, tendon damage, and enthesitis [35,43]. Studies performed thus far suggest moderate to good intraobserver reproducibility and variable agreement between observers, which depends, in part, upon the joint examined and upon the feature of disease that is being assessed. In addition, studies have shown a good US sensitiveness to change when treating synovitis in patients with rheumatoid arthritis [44,45]. Additional data are needed on these test characteristics in patients with various musculoskeletal disorders [46]. The following studies are illustrative:

●In one study, the second metacarpophalangeal joints of 55 patients with rheumatoid arthritis were assessed ultrasonographically by two examiners [31]. The intraobserver kappa statistic for the detection of cortical bone erosions on the second metacarpophalangeal joints of 55 rheumatoid arthritis patients was 0.75, and the interobserver kappa for agreement between two observers was 0.76, implying substantial agreement among the observers. These kappas are comparable with those of radiologists scoring lesions on mammograms [47]. All bone erosions that were detected sonographically but that were not visible on radiography corresponded to MRI abnormalities (ie, were true positive findings).

●In another study, 14 experts examined four regions (shoulder, hand, knee, and foot) in four patients and compared the acquired images with previously performed MRI for synovitis, erosions, bursitis, and tendonitis [48]. The overall kappa was 0.76. However, there was considerable difference between joints, with the kappa statistic ranging from a high of 0.76 for the shoulder to a low of 0.28 for the ankle/toes. Kappa values for erosions, bursitis, and tendonitis were high, but agreement for synovitis was low. Sensitivity and specificity of the US images compared with MRI was moderate to good.

●In another study, 23 experts acquired and interpreted images of four joints (shoulder, hand/wrist, knee, and ankle/foot), without a set of predefined definitions or scanning method [49]. Overall agreements were 91 percent for joint effusion/synovitis and tendon lesions, 87 percent for cortical abnormalities, 84 percent for tenosynovitis, 83.5 percent for bursitis, and 83 percent for power Doppler signal; kappa values were good for the wrist/hand and knee (0.61 and 0.60) and were fair for the shoulder and ankle/foot (0.50 and 0.54).

●Another experiment assessed the intra- and interobserver reliability regarding synovitis of small hand joints in patients with rheumatoid arthritis. Seventeen rheumatologist experts in US scored a sequence of 86 images of metacarpophalangeal, proximal interphalangeal, metatarsophalangeal, and wrist joints for the presence or absence of gray scale and power Doppler synovitis according to a predefined (OMERACT) definition. In the follow-up experiment, the experts acquired images of 32 metacarpophalangeal and 32 proximal interphalangeal joints on eight patients with rheumatoid arthritis. Results showed good agreement both between the interobserver and intraobserver interpretation of static images, but the kappa values for the acquisition of images were low [50].

●US has been shown to be a valid and reproducible technique for detecting synovitis of the knee, and is more sensitive than clinical examination [37]. As an example, a small observational study of US assessment of patients with knee osteoarthritis found that US was reliable for detecting osteophytes and protrusion of the medial meniscus, as well as for detecting synovitis [42].

More data are needed regarding the responsiveness to change for US synovitis and tenosynovitis in rheumatoid arthritis, and for validation of these observations, before we can draw a more definite conclusion about the position of US in the clinical trial management of patients with rheumatoid arthritis. (See "Clinical manifestations of rheumatoid arthritis", section on 'Ultrasonography' and "General principles and overview of management of rheumatoid arthritis in adults", section on 'Clinical assessment of disease and related testing'.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Musculoskeletal ultrasound".)

SUMMARY

●The selection of anatomic structures and the planes in which to image them in clinical practice is driven by the patient's complaints. (See 'Complaint-driven image selection' above.)

●The type of ultrasound (US) equipment (eg, transducer frequency and shape) and the mode of operation (eg, brightness or B-mode, grayscale, color Doppler, or power Doppler) are also dependent upon the location, size, depth, and type of pathologic changes that are being assessed in the structures of interest. (See "Musculoskeletal ultrasonography: Nomenclature, technical considerations, and basic principles of use".)

●Musculoskeletal US can provide images of tendons, ligaments, muscles, entheses, bursae, joints, cartilage, and peripheral nerves and can detect bone erosions and synovitis. (See 'Ultrasonography for specific disorders' above.)

●Power Doppler signal from inflamed synovium may correlate with the severity of synovitis as determined by biopsy. If confirmed, this suggests that US could play a role in diagnosing and monitoring goal-directed therapy for chronic forms of inflammatory joint disease. (See 'Synovitis' above.)