Clinical aspects

Description

Imaging methods. Advances in imaging techniques to visualize peripheral nerves has enhanced our ability to study pathological, structural, and functional changes. Evaluation of peripheral nerves can be performed using high resolution MRI as well as neuromuscular ultrasound. These techniques complement the clinical and electrodiagnostic assessment of peripheral nerve abnormalities. MR neurography is not universally available, and it requires more time for image acquisition. Ultrasound is a low-cost technique that is more widely available and offers dynamic capabilities, but it is largely operator dependent.

MR neurography. MR neurography utilizes axial T1-weighted and fluid-sensitive, fat-suppressed T2-weighted images to evaluate peripheral nerve characteristics, including signal intensity, anatomic course, nerve diameter, fascicular pattern, and nerve size (09). MR neurography can visualize nerve sheath tumors and perineural fibrosis (10). On T1-weighted images, normal nerve signal is isointense to muscle. On T2-weighted images, nerve may be isointense to slightly hyperintense compared to muscle, depending on the degree of background fat suppression (23).

In neuropathy, the T2 MRI signal intensity increases abnormally, almost to the degree of adjacent vessels (37). Longitudinal imaging visualizes the nerve caliber and contour throughout the nerve course. Nerve injury can also be graded and classified based on degree of severity using the Seddon and Sunderland grading system (45). This grading system separates nerve injuries into neuropraxia, axonotmetsis, and neurotmesis, based on severity of nerve injury.

In the mildest form of nerve injury, neuropraxia, pathologic changes are limited to the myelin sheath around the axon, typically resulting in transient loss of function. Examples of injury include stretch or mild focal compression. Chhabra and colleagues demonstrated that MR neurography in neuropraxic injury shows nerve signal hyperintensity on T2-weighted imaging, with only mild peripheral nerve enlargement (09).

Axonotmesis occurs with a greater degree of injury and is characterized by axonal damage, with sparing of the supporting nerve structures, the perineurium and epineurium (45). The axon may regenerate at a speed of approximately 1 mm per day from the point of injury to its target tissue. Chhabra and colleagues describe MR neurography findings of fascicular effacement, enlargement, or overall fascicular disruption in addition to the nerve signal hyperintensity and focal enlargement that is seen in neuropraxic injury (09). Complete nerve transection, known as neurotmesis, is the most severe form of nerve injury. In this case, there is complete loss of nerve continuity. If imaging is performed acutely, MR neurography may confirm nerve discontinuity with evidence of fluid and granulation tissue adjacent to the disrupted nerve endings (09). Several weeks to months following nerve transection, MR neurography can depict strands of hypointense soft tissue within the nerve cavity on T2-weighted imaging.

The table below summarizes the MR characteristics of normal and abnormal nerves (Table 1).

Table 1. Characteristics of Normal and Abnormal Nerves on MR neurography

Neuromuscular ultrasound. Fornage first described ultrasonographic peripheral nerve imaging in 1988 (16). Technological advancements have improved visualization of peripheral nerve anatomy. With the emergence of high frequency transducers (greater than 15MHz), even the smallest of cutaneous nerve branches can be imaged. Additionally, improved sensitivity of Doppler studies has allowed for evaluation of vascular changes within nerve (26).

Ultrasound of the peripheral nerves should be done using high-resolution (15 to 18 MHz) transducers. Imaging should be performed along the nerve short axis and be compared to the contralateral extremity. Real time visualization of blood flow can be achieved using Doppler ultrasonography. Specifically, quantification of the Doppler shift, or the change in frequency of the reflected ultrasound from the incident ultrasound, allows for the calculation of blood flow velocity (32). Color Doppler converts the blood flow measurements into a color map, correlating with the speed and direction of vessel blood flow. Power Doppler, more sensitive than color Doppler, allows for the detection of the direction of blood flow through small blood vessels. The direction of flow is extracted from the sequence of returning echoes and then converted by the machine into a color. Conventionally, shades of red represent flow towards the transducer and blue away from the transducer. Power Doppler can be used to differentiate small peripheral nerves from surrounding vasculature as well as to evaluate focal areas of nerve hyperemia (25). Instead of visually displaying blood flow as color, spectral Doppler graphically displays blood flow velocities recorded over time (40). Though there is limited literature on the utility of spectral Doppler in peripheral neuropathies, Hough, and colleagues demonstrated the reliability of spectral Doppler in quantifying median nerve motion at the elbow during wrist extension in normal subjects and compared this to in-vivo microelectrode recordings of median nerve motion during upper limb movements from a study by McLellan and colleagues in 1976 (33; 22).

Normal peripheral nerves have a tubular form, with alternating zones of hypo- and hyperechogenicity, which correspond to nerve fascicles and perineurium and are often described as having a “honeycomb” appearance in transverse views.

Nerve tissue can be differentiated from surrounding tissue using the principle of anisotropy. In ultrasound, anisotropy is the quality of varying echogenicity depending on the axis of insonation. Identification of nerve and differentiation from nearby tendons can be performed by changing the angle of insonation. Tendons have a more fibrillary pattern of echogenicity as compared to nerve. Nerves have less variability in echotexture and echogenicity with variation of the angle of insonation than tendons or blood vessels because the probe is tilted off the perpendicular short axis (52). Additionally, when nerve pathology is evident, the surrounding musculature may appear hyperechoic, suggesting possible denervation atrophy.

The basic components of the ultrasonographic examination of peripheral nerves include:

Indications

The various indications for peripheral nerve imaging include common entrapment neuropathies, brachial plexopathies, peripheral nerve sheath tumors, diabetic polyneuropathy, and acquired immune-mediated demyelinating polyneuropathies.

Contraindications

Contrast MR neurography is contraindicated in pregnancy.

Ultrasonography is a safe and noninvasive imaging technique; however, it is not suitable in areas in which recent invasive procedures have been performed, which are specifically limited by loss of skin integrity following the procedure, limiting transducer application.

Results

Entrapment neuropathies. Conceptually, nerve entrapment can be envisioned as initially inducing nerve “pinching” or “flattening,” but in reality the most common ultrasonographic finding in focal nerve entrapment is focal nerve enlargement, usually just proximal to the site of nerve entrapment (07). Nerve enlargement is usually fusiform with cross-sectional area enlargement defined by an area more than two standard deviations above the reference mean or more than 1.5 times greater than the area of the unaffected portion of the nerve. Walker and Cartwright provide reference values for the cross-sectional areas of the median, ulnar, radial, musculocutaneous, vagus, sciatic, peroneal, tibial, and sural nerves as well as for the brachial plexus at various sites (52).

Nerve enlargement occurs due to focal inflammation and vascular changes from chronic compression (08). Common sites of entrapment that are amenable to ultrasound evaluation include: median neuropathy at the wrist, ulnar neuropathy at the elbow, ulnar neuropathy at the wrist, and peroneal neuropathy at the knee.

Table 2, adapted from Cartwright and Walker in 2013, reveals the various grades of median nerve movement at the wrist. Cartwright and Walker evaluated three ultrasonographic parameters following steroid injection for carpal tunnel syndrome, including nerve cross-sectional area, mobility, and vascularity, with demonstration of improvement of all three parameters following steroid injection (07). They also noted that at 6 months following steroid injection, nerve mobility and vascularity values stopped improving and began to return to baseline, suggesting that these parameters may be sensitive in the assessment of median neuropathy at the wrist.

Table 2. Grading Median Nerve Movement at the Wrist

Interestingly, Chompoopong and Preston demonstrated the presence of structural abnormalities or anatomical variations in 51 of 114 patients with an electrophysiologic diagnosis of median neuropathy across the wrist involving the nondominant hand whose management was altered based on ultrasonographic data alone (11).

The utility of MR neurography in the detection of entrapment neuropathies is less well-described. An illustrative case is a report of a 9-year-old with an ulnohumeral dislocation and medial epicondyle fracture, who developed a proximal median neuropathy following cast immobilization. MR neurography demonstrated median nerve compression at the interosseous site (between the medial epicondyle and apophyseal fracture site), with evidence of diffuse T2 signal abnormality (35). In this clinical scenario, early detection of a proximal median neuropathy allowed for immediate surgical exploration and relief of focal nerve compression. Additionally, electrodiagnostic studies in a 9-year-old child may be difficult to perform due to patient tolerance of the procedure, and imaging was preferable as an initial diagnostic test as it is noninvasive.

Brachial plexopathies. Idiopathic brachial plexopathy (also called Parsonage-Turner syndrome, neuralgic amyotrophy, or brachial plexus neuritis) involves varying portions of the plexus and has a wide phenotypic presentation, often with severe pain and varying degrees of muscle weakness (44). The most commonly affected nerve distributions are the predominantly motor nerves, such as the suprascapular, axillary, and long thoracic nerves, but upper trunk and other locations are often affected. The disease is most commonly unilateral. The brachial plexus may also be affected by: compression or traction injury, direct trauma, radiation, inflammatory demyelinating neuropathy (ie, Guillain Barre syndrome, chronic inflammatory demyelinating polyneuropathy, chemotherapy hereditary, infectious, or other causes) (12). Visualization of the brachial plexus can be achieved using both MR neurography and neuromuscular ultrasound.

MR imaging of brachial plexus pathology is best performed using a multichannel, phase-array surface coil, with both T1-weighted and short-time inversion recovery sequences (12). Direct signs of plexus injury can include visualization of enlargement of large segments of one or multiple portions of the plexus, with increased T2 signal intensity, along with abnormalities or discontinuity of individual fascicles (03). Additionally, contrast enhancement (eg, gadolinium) can also be seen, though absence of enhancement does not exclude pathology. Crim and Ingalls reported a case series of 43 patients who underwent both electrodiagnostic studies and MR neurography for a suspected diagnosis of brachial plexopathy, based on their clinical presentations. Electrodiagnostic studies were negative in 12 patients, and another 13 patients had electrodiagnostic abnormalities other than brachial plexopathy; one patient had equivocal data with no definite diagnosis, and the remaining 17 patients had both clinical and electrodiagnostic diagnoses of brachial plexopathy. Two radiologists evaluated the MR imaging, comparing their impression with the electrodiagnostic data as the gold standard. The sensitivity of MR neurography ranged from 41.2% to 70.6%, with a significantly higher specificity at 97.7% to 100% (12). Thus, it is cautioned that a normal appearance of the plexus on MR neurography does not refute a clinical diagnosis.

Though MRI is the most commonly used imaging method for the evaluation of brachial plexus pathology, neuromuscular ultrasound imaging is steadily growing in popularity. MR neurography is expensive, provides a static image, and cannot be used in patients with non-MR compatible devices or metal. Ting and colleagues in 1989 first demonstrated the utility of ultrasound in the guidance for the approach in performing local axillary brachial plexus blocks (47). Ultrasound imaging of the brachial plexus is restricted largely to the supraclavicular portion of the brachial plexus. Gruber and colleagues demonstrated the utility of ultrasound in the visualization of the supraclavicular portion of the brachial plexus in 221 patients using a high-frequency (12 to 17 MHz) probe (20). Twelve patients with clinical history and neurologic examination findings consistent with posttraumatic brachial plexus involvement had ultrasonographic evidence of nerve root and trunk enlargement, hypoechogenicity, and fascicular distortion. These findings were confirmed with surgical exploration. The mean time elapsed between the patient’s trauma and the high-resolution ultrasound evaluation was 5.5 months (range 0 to 27 months), with a mean of 3.3 months between HRUS evaluation and surgery (range 0 to 17 months). The authors concluded that ultrasound should be used as first-line technique for brachial plexus visualization following trauma in an effort to expedite surgical exploration.

Brachial plexus ultrasound has been used in suspected inflammatory or immune-mediated brachial plexus neuritis. Aranyi and colleagues studied 14 patients with clinical presentation concerning for neuralgic amyotrophy (04). Clinical nerve involvement included radial (10 patients), musculocutaneous (two patients), suprascapular (two patients), long thoracic (one patient), and axillary (one patient) nerves. Electrodiagnostic studies were also performed, showing severe axonal involvement in nearly all patients. High-resolution ultrasonography revealed:

Pan and colleagues performed surgical exploration on five patients with clinical and electrodiagnostic findings consistent with brachial neuritis involving at least two nerves and reported presence of hourglass-like constrictions in individual peripheral nerves. Conservative treatment over 2 to 11 months (vitamin B12 and physical therapy) was not successful. On surgical exploration, they noted well-described hourglass-like constrictions in the involved nerve, with areas of nerve edema and thickening for 3 to 5 cm beyond the area of constriction, along with presence of adhesions (36). Neurolysis was performed, with modest clinical improvement in strength and sensation noted after 24 to 84 months follow-up postoperatively.

High resolution ultrasound has been used to visualize the anterior interosseous nerve, as a presentation of distal neuralgic amyotrophy. Noda and colleagues studied the anterior interosseous nerve with high resolution ultrasound in five patients with clinical anterior interosseous nerve palsy (34). There was a spectrum of sonographic findings ranging from neural swelling to incomplete hourglass-like constriction of the nerve fascicle, to complete hourglass-like constriction of the nerve fascicle (compression), and ultimately to twisting of the nerve fascicle. Koneru and colleagues explored the MR imaging and ultrasound correlation of specific brachial plexus nerve injuries and demonstrated similar focal nerve constriction in a patient with suprascapular neuropathy at the spinoglenoid notch, with corresponding MRI findings of diffuse thickening and hyperintensity of all the cords and trunks, along with denervation changes and atrophy in the affected muscles (27). This finding suggests a broader involvement of the brachial plexus, such as in Parsonage-Turner syndrome, though localization with ultrasound would have been seen as a more focal entrapment neuropathy.

Over time, improvements in ultrasound technology will continue to improve visualization of pathology. Specific training for nonradiologists performing these studies is required, and there is a credentialing process for certification in neuromuscular ultrasound by the American Association of Neuromuscular and Electrodiagnostic Medicine (AANEM).

Peripheral nerve sheath tumors. Imaging with either MR neurography or neuromuscular ultrasound can show peripheral nerve enlargement related to nerve sheath tumors. There is caution that a definitive diagnosis may not be reached based on imaging alone, and that biopsy may be necessary.

MR neurography can assist in the evaluation of peripheral nerve enlargement. Specifically, it may assist with differentiation between benign peripheral nerve sheath tumors, malignant peripheral nerve sheath tumors, hereditary or inflammatory neuropathy, post-traumatic neuroma, intraneural ganglion, or other secondary non-neurogenic processes such as neurolymphoma, or epithelial intraneural tumors (02). Specifically, whole body MRI is the reference standard to identify nerve sheath tumors in neurofibromatosis type 1 (NF1), allowing for the surveillance of plexiform neurofibromas (42). Plexiform neurofibromas are typically benign, though complex tumors, and can cause enlargement of deep nerves and plexus. They are classically seen in individuals with NF1. They typically appear as isointense to muscle on T1-weighted imaging, and hyperintense on T2-weighted imaging. They have variable enhancement with gadolinium.

Fluoro-2-deoxy-D-glucose positron emission tomography/computed tomography (FDG-PET/CT) is an imaging modality that noninvasively assesses in vivo glucose metabolism. It is commonly used to stage and monitor treatment response and investigate for recurrence in solid-tumor malignancies. This imaging modality may be helpful in distinguishing between malignant and benign plexiform neurofibromas (49). Specifically, the uptake of FDG is measured using a unitless maximal standardized uptake value (SUVmax). Ferner and colleagues demonstrated that malignant lesions have a statistically higher maximal standardized uptake value as compared to benign lesions, with mean value 5.4, as reported in seven malignant peripheral nerve sheath tumors identified out of 23 total plexiform neurofibromas from 18 subjects (14). These findings were histologically confirmed.

Repeated whole-body MRI is both expensive and time-consuming to perform. As such, neuromuscular ultrasound can be a safe, cost-effective and far less time-consuming modality to evaluate peripheral nerve sheath tumors and monitor them for interval change. Some examples include posttraumatic neuromas, solitary neurofibromas, and peripheral schwannomas. Posttraumatic neuromas can disrupt the peripheral nerve either partially or completely. On ultrasound, a hypoechoic neuroma can be seen at the nerve terminus and is usually bulbous appearing (52).

Solitary neurofibromas, or peripheral schwannomas, such as in NF1 and NF2, can be imaged with neuromuscular ultrasound. These lesions are also hypoechoic and focal, with normal-appearing nerve adjacent to the area of focal abnormality. Winter and colleagues studied 37 patients with neurofibromatosis (27 with NF1 and 10 with NF2) using ultrasound and found evidence of significantly increased cross-sectional area of most nerves in NF1 as compared to NF2. Patients with NF1 had predominantly generalized plexiform tumors, whereas patients with NF2 often had localized enlargements (schwannomas) between normal nerve segments (53). Additionally, power Doppler findings of hypervascularity would suggest that the lesion is more likely to be a schwannoma rather than neurofibroma as schwannomas are more vascular (41). Interestingly, Dr. Podnar published a retrospective analysis of 15 patients who were found to have peripheral nerve sheath tumors (either neurofibromas or schwannomas), of whom seven patients had only clinical findings of pain and palpable mass, with largely unrevealing electrodiagnostic testing (38). Of those seven patients with normal electrodiagnostic studies, four had definite or probable schwannoma, and three had probable neurofibroma. Had ultrasonography not been performed in those individuals with unrevealing electrodiagnostic testing, the finding may have been missed.

Diabetic polyneuropathy. The most common form of diabetic neuropathy is a distal symmetric dying-back phenotype, progressing in a length-dependent pattern and involving more sensory than motor nerves. A distal axonopathy may be confirmed with electrodiagnostic studies. There are, however, many other variants of neuropathy in diabetics. MR neurography can demonstrate a large spectrum of abnormalities in patients with diabetes. Patients with diabetic peripheral neuropathy can have abnormal T2 signal hyperintensity with fascicular enlargement in acute to subacute stages of neuropathy. As the disease progresses, the fascicles appear more atrophic. In chronic diabetic neuropathy, imaging may suggest intra-epineurial fat deposition (48).

Ultrasound changes of nerve with progression of diabetic sensorimotor polyneuropathy have also been studied. Arumugam and colleagues performed electrodiagnostic studies on 100 patients with diabetic sensorimotor polyneuropathy and compared them to 40 age-matched healthy controls. Nerve ultrasound evaluated the cross-sectional area diameters of the median, ulnar, peroneal, tibial, and sural nerves (05). Polyneuropathy severity was rated using the Toronto Clinical Scoring System. On nerve conduction studies, absent sural nerve responses were noted in 69% of patients with severe diabetic sensorimotor polyneuropathy. Ultrasound revealed enlargement of the cross-sectional area of all nerves in diabetic polyneuropathy, with the greatest correlation between electrodiagnostic parameters and cross-sectional area of the ulnar, peroneal, tibial, and sural nerves. Enlargement of nerve cross-sectional area correlated with increasing disease severity and was especially relevant in those individuals who had absent nerve responses on electrodiagnostic testing. Additionally, entrapment neuropathies are frequently seen in any stage of diabetic neuropathy and may at times be the earliest neurophysiologic abnormalities noted in diabetic patients, even in the absence of a generalized polyneuropathy (39).

Acquired immune-mediated demyelinating polyneuropathies. The diagnosis of acquired immune-mediated demyelinating polyneuropathies is based primarily on the clinical features, historical course, and electrodiagnostic evidence of demyelination (50). MR neurography of proximal nerve segments (ie, brachial and lumbosacral plexus) has revealed root enlargement or gadolinium enhancement, which can further support the diagnosis. Various studies reported that 52% to 86% of patients with chronic inflammatory demyelinating polyneuropathy will have T2 signal hyperintensity with plexus enlargement on MRI (01; 29). One study compared peripheral nerve MR neurography parameters in 36 subjects, 18 with chronic inflammatory demyelinating polyneuropathy and 18 age-matched controls (28). The patients with chronic inflammatory demyelinating polyneuropathy had increased nerve cross-sectional area with increased T2 signal hyperintensity, predominantly in proximal nerves, which was most notably observed at the lumbosacral plexus and sciatic nerve. The imaging findings correlated with abnormalities in the electrophysiologic data.

Another study visualized peripheral nerves with MR neurography and quantified the volumes of the brachial and lumbar plexus as well as their contributing nerve roots in 13 individuals with chronic inflammatory demyelinating polyneuropathy compared to 12 healthy controls (24). The findings corroborate other reports that MR neurography reveals enlargement of both the brachial and lumbar plexus and nerve roots and that the imaging findings correlate with disease duration. A retrospective study evaluated 48 patients with clinical findings suggestive of chronic inflammatory demyelinating polyneuropathy. Thirty-eight patients did not meet the definite electrodiagnostic criteria for chronic inflammatory demyelinating polyneuropathy. The patients underwent lumbosacral or brachial plexus MRI (13). Plexus MRI showed abnormalities in 58% (22 of 38) of patients who did not meet the electrodiagnostic criteria for chronic inflammatory demyelinating polyneuropathy. Of the 22 patients with abnormal MRI findings, all had increased nerve T2 signal hyperintensity, 91% had nerve enlargement, and 36% had contrast enhancement. The patients who did not meet electrodiagnostic criteria for chronic inflammatory demyelinating polyneuropathy were found to have more asymmetric and less diffuse MRI abnormalities as compared to those individuals with clinical and electrodiagnostic chronic inflammatory demyelinating polyneuropathy.

Nerve ultrasonography in chronic inflammatory demyelinating polyneuropathy was first performed in 2000 in a patient who was followed longitudinally. Findings of focal enlargement of the brachial plexus were noted (46). In 2004, Matsuoka and colleagues published ultrasound findings of cervical nerve root hypertrophy in 9 of 13 patients with chronic inflammatory demyelinating polyneuropathy (31). In 2009, Zaidman and colleagues performed ultrasonography of the median and ulnar nerves in 190 individuals, 100 with neuropathies (11 with Charcot-Marie Tooth, 36 with chronic inflammatory demyelinating polyneuropathy, 17 with acute inflammatory demyelinating polyneuropathy, and 36 with axonal neuropathy) and 90 healthy controls. Enlarged nerve cross-sectional areas were seen in most patients with chronic inflammatory demyelinating polyneuropathy (86%), about half of patients with acute inflammatory demyelinating polyneuropathy (47%), and in 19% of patients with axonal neuropathy (54). A study evaluated 20 adult patients who met diagnostic criteria for chronic inflammatory demyelinating polyneuropathy and divided patients into a group with stable or remitting disease and a group with progressive disease based on overall disability sum score (15). Patients underwent electrodiagnostic testing every 6 months with concomitant high-resolution ultrasonography, evaluating cross-sectional area of bilateral tibial (ankle, popliteal fossa), peroneal (fibular head, popliteal fossa), sural (lower leg), median (wrist, forearm, upper arm), ulnar (wrist, forearm, elbow, upper arm), and radial nerves (spiral groove). Patients were followed for a median 34 months (though only 11 of 20 patients). The intranerve cross-sectional area variability in the legs increased over time in the group with disease progression and was stable over time in the group with stable or remitting disease. The study concluded that neuromuscular ultrasound can assist with objective monitoring of the clinical course of patients with chronic inflammatory demyelinating polyneuropathy using intra-nerve cross-sectional area variability.

Additionally, several case reports have demonstrated focal nerve enlargement at sites of electrodiagnostic conduction block in chronic inflammatory demyelinating polyneuropathy. Specifically, Granata and colleagues demonstrated ultrasonographic median and ulnar nerve enlargement with hypoechogenicity in a 19-year-old man with left hand weakness, with correlation of findings at areas of electrophysiologic conduction block (18). Scheidl and colleagues also presented two cases of multifocal acquired demyelinating sensory and motor neuropathy (variant of chronic inflammatory demyelinating polyneuropathy). The first was a case of predominantly upper limb involvement, with electrophysiologic conduction block of the radial nerve at the spiral groove, the median nerve in the upper arm, and the ulnar nerve at the distal forearm (43). Ultrasonography revealed nerve cross-sectional area enlargement with hypoechogenicity at the site of conduction block. In the second case of a 51-year-old woman with partial right median, left ulnar, and peroneal nerve involvement, demonstration of ultrasonographic enlargement of nerve cross-sectional area was proximal to the site of conduction block. These cases demonstrate the role of ultrasonography in localizing nerve pathology at sites otherwise not accessible to electrodiagnostic study.

Assessing respiratory function in neuromuscular disorders. Respiratory function is often affected in neuromuscular disorders, and respiratory muscle weakness can become prominent in disease progression. Dyspnea, lung infections, and sleep disturbances are common symptoms experienced by patients. Typical methods of evaluating respiratory muscle function, such as pulmonary function testing, rely on patient effort, which can be challenging with children, and can only assess global respiratory function. Pulmonary function testing is not an accurate measure of respiratory muscle changes as muscle atrophy can present well before functional testing is abnormal (21). Ultrasound and MRI have been shown to successfully evaluate both the function and structure of individual respiratory muscles (51).

There are two common approaches to using ultrasound in imaging the diaphragm: an intercostal approach measuring thickness of the diaphragm and a subcostal approach measuring diaphragm movement during breathing. In measuring diaphragm thickness, the thickening fraction is often used a marker to detect diaphragm atrophy or dysfunction. Some studies measure the thickening fraction as the difference between end-expiratory and end-inspiratory thickness divided by end-expiratory thickness, whereas other studies measure diaphragm thickness as the end-inspiratory thickness divided by the end-expiratory thickness. Patient sex and position should be noted during evaluations, as studies have shown differences in normative values with differences in patient sex and anatomical positions (06). Thickness of the parasternal intercostal and abdominal wall muscles can be measured, and the thickness fraction can be calculated as well. Diaphragm excursion can be measured using a subcostal approach in M-mode during tidal breathing, deep inhalation, or sniff maneuvers.

MRI can also be used in assessing respiratory muscle structure and function, specifically, to image both muscle structure and function. T1-weighted images can show intramuscular fat, which appears hyperintense as compared to muscle tissue and water. T2-weighted images can visualize both fat and water within muscles. Using different MRI modalities and protocols, respiratory muscle degeneration can be scored according to established rating systems (17). Additionally, Dixon MRI techniques, used to calculate the muscle’s fat fraction whereby the fat and water contribution to each voxel of tissue can be quantified and used to calculate the muscle’s fat fraction, can be used to quantify the degree of fatty infiltration in atrophied muscles. This technique utilizes the fat and water contribution to each voxel of tissue, which is then quantified and used to calculate the muscle’s fat fraction. Functional testing can be done by quantifying total lung area using 2D images, or total lung volume using 3D images, at maximal inspiration and expiration.

Special considerations

MR neurography should not be performed in individuals with non-MRI compatible devices such as pacemakers or with metal fragments within the body. There are no special considerations for the performance of neuromuscular ultrasound, though skin integrity should be intact prior to application of transducing gel and the ultrasound probe. To date, there is no literature demonstrating safety of use of high frequency probes for ultrasonography in pregnancy.

Clinical vignette

The cases below illustrate the utility of imaging in various peripheral neuropathies.

Case 1. Visualizing brachial plexus injury using MR neurography. A 38-year-old woman presented 1 week after a fall during which she hit her right anterior shoulder and neck against a chair. She acutely developed proximal right arm pain and noted weakness of all shoulder movements. Over the following 2 weeks, she developed elbow flexion weakness and forearm sensory changes. The strength examination was limited by pain. There was reduced sensation to cold temperature and pinprick over the right anterior shoulder and in the anterior forearm extending to the dorsal thumb.

Electrodiagnostic studies revealed reduced right lateral and medial antebrachial sensory nerve amplitudes and needle electromyography (EMG) findings consistent with severe, active denervation in the right C5/6 myotomes, with cervical paraspinal muscle sparing. The study best localized pathology to the upper trunk of the brachial plexus, with most prominent involvement of the axillary, musculocutaneous, and suprascapular nerves. MR neurography was performed and revealed T2 signal hyperintensity that was diffuse along the right C5 and C6 nerve roots, with extension of uniform T2 signal hyperintensity to the upper trunk of the brachial plexus. She underwent aggressive physical therapy but ultimately benefitted from surgical neurolysis and reconstructive surgery.

Case 2. Intraneural sclerosing perineuroma of the posterior interosseous nerve imaged with neuromuscular ultrasound. A 35-year-old woman presented 1 year after left proximal forearm radial nerve transposition surgery for treatment of dorsal forearm and hand paresthesias with presumed radial nerve involvement. In the months following surgery she noted progressive left-hand numbness and weakness. On examination she had finger extension weakness, which was most prominent at the index finger. There was intact lumbrical function and normal finger flexion and thumb abduction strength. There was sensory loss to all modalities over the dorsum of the left forearm, hand, and fingers.

Electrodiagnostic studies revealed marked reduction of the left radial motor nerve compound muscle action potential (CMAP) amplitudes with recording over the extensor indicis proprius muscle. Side-side radial sensory responses were symmetric. Needle EMG revealed complex repetitive discharges with a fast-firing single motor unit recruitment of the left extensor indicis proprius with evidence of chronic denervation or reinnervation of the extensor digitorum communis. EMG of the brachioradialis was normal. The electrodiagnostic studies localized pathology to the posterior interosseous branch of the left radial nerve.

Ultrasonography revealed a focal, well-circumscribed, avascular enlargement of the posterior interosseous nerve arising at the elbow. The abnormal, focal lesion measured 1.35 cm in length and 0.56 cm in diameter.

Case 3. Neuromuscular ultrasound detection of neurofibromas in a patient without visible or palpable cutaneous manifestations. A 30-year-old woman with no past medical history presented with complaints of arm paresthesias. Examination revealed sensory changes in the bilateral medial forearms, extending to digits 4 and 5 with normal strength and symmetric reflexes. Electrodiagnostic studies revealed conduction blocks in several locations, including the bilateral ulnar wrist to elbow segments.

Neuromuscular ultrasound was performed and revealed findings consistent with diffuse plexiform neurofibromas in nearly every peripheral nerve imaged. The lesions appeared to arise from multiple fascicles, and there was significant distortion of normal fascicular architecture. The cross-sectional area of all nerves assessed was significantly increased.

In the context of the ultrasound results, the electrodiagnostic findings of conduction block were reinterpreted as being likely being related to technical issues with submaximal stimulation at locations of the plexiform neurofibromas.