Electroneuromyography has long been considered the “gold standard” test to determine nerve injury and function. In recent years, it has become more and more popular to attempt to predict nerve injury by the use of MRI. This paper reviews the published literature to compare the abilities of MRI and ENMG in predicting several common nerve injuries.
In 1993, Buirski et al. (4) used magnetic resonance imaging to evaluate abnormal lumbar disc’s in 115 symptomatic patients and 63 symptom-free control patients. This study tried to determine which abnormal disc’s would be painful using MRI. Results showed no significant difference between the two groups. In 1998, Saifuddin et al. (5) tried to determine the sensitivity of MRI to painful annular tears. They found that MRI had a specificity of 95.2%, however, showed only 26.7% sensitivity to painful annular tears. Toyokura et al. (6) found EMG/NCS to have 70% sensitivity to radiculopathies, but no specificity was mentioned. Additional studies, with Ito et al. (7) and Horton et al. (8) further stated that MRI is a poor predictor of cause for patient’s symptoms.
Jarvik et al. (9) showed that MRI had a sensitivity of 96% for determining carpal tunnel syndrome, however demonstrated a specificity of only 33 to 38%. Deryani et al. (10) found a 33% sensitivity of MRI in predicting carpal tunnel syndrome. No specificity was mentioned. This group determined that MRI was most helpful in determining cases of suspicious clinical and electrophysiological diagnoses. Stevens (11) published a paper in 1998 in which he determined conventional nerve studies had approximately a 90% sensitivity and a 90% specificity for predicting carpal tunnel syndrome. In 2003, Chang et al. (12) evaluated the sensitivities of the most commonly used nerve conduction tests. They determined the sensitivity for median to ulnar motor comparison to be 70%. Additionally, they found sensitivities of 68.8% for median motor distal latency, 73.8% for median sensory distal latency, and 77.5% for second lumbrical to second dorsal interosseous comparison. Furthermore, they determined that a combination of these tests increased the sensitivity for predicting carpal tunnel syndrome to 91.9%.
Grant et al. (13) performed a small pilot study in 2002 evaluating 21 ulnar nerves. They found that MRI had a 100% sensitivity in determining cubital tunnel however did not determine specificity. They also determined that standard nerve conduction studies provided a sensitivity to cubital tunnel syndrome of 71% preoperatively and 90% sensitivity intraoperatively. Many other studies have reported electrophysiological testing sensitivities of 20% to100% for determining ulnar neuropathy at the elbow (13).
Bendszus et al. (14) determined that MRI can detect changes in denervated muscle within 24 hours post injury. Aagaard et al. (15) demonstrated the potential of MRI to noninvasively differentiate between axonotmetic and neurotmetic nerve injuries.
While MRI has made significant inroads with its ability to identify peripheral nerve injury, electrophysiological testing continues to be the “gold standard.” MRI is a relatively new area of medicine and has been used clinically in medicine for only two and a half decades. As further research is conducted and improved techniques for evaluating anatomical structures are developed, MRI will become more commonly used for evaluating nerve injuries because of its noninvasive nature. Perhaps the defining limitation for MRI in evaluating peripheral nerve injury is its inability to provide more than a visual image of anatomical structures and cannot truly evaluate nerve function.
We are an EMG and NCS testing company. Electromyography, or EMG, is paired with nerve conduction studies, NCS, to evaluate the integrity of the peripheral nervous system and how it interacts with the muscles. We provide in-house services on a referral basis.