Comparison of MRI and electrophysiological studies for detecting brainstem lesions in traumatic brain injury.

Abstract

The yield of magnetic resonance imaging (MRI) and electrophysiological studies in detecting brainstem lesions was assessed in 35 patients suffering from traumatic brain injury (Glasgow Coma Scale, 3-10). As an inclusion criterion, all patients had brainstem trauma as revealed by early MRI or electrophysiological studies. Of the 35 cases, 7 (20%) had brainstem lesions detected by MRI only, whereas in 10 patients (29%), electrophysiological examination disclosed impairment of brainstem function with normal MRI. In 18 (51%) subjects, both diagnostic techniques revealed brainstem lesions. The midbrain was the most common location of lesions. Masseter reflex recording had the highest yield (93%) of abnormal findings. No mismatch with respect to site and side of abnormality occurred between MRI and electrophysiological studies. Outcome analysis indicated an unfavorable course for the vast majority (83%) of patients, regardless of the diagnostic means disclosing traumatic brainstem injury. Therefore, both techniques are effective in disclosing traumatic brainstem injury, with diagnostic overlap in about 50% of cases. In contrast to MRI, electrophysiological investigation is easily performed and repeated at low cost in the setting of an intensive care unit, where such patients are typically hospitalized after trauma. In addition to electrophysiological assessment of brainstem function, MRI is recommended in each case having normal electrophysiological findings when brainstem injury is suspected.