Median nerve compression can be detected by magnetic resonance imaging of the carpal tunnel

Clinically symptomatic carpal tunnel syndrome is not necessarily accompanied by impaired nerve conduction values. Surgical decompression, however, may immediately lead to complete and lasting relief of symptoms in these patients. Because minimally invasive techniques have reduced perioperative morbidity and vocational impairment related to operative decompression, the decision to decompress symptomatic patients (despite still unimpaired nerve conduction values) might be subject to discussion in the future. New diagnostic tools may be helpful in deciding which therapeutical options to choose. When the wrist is held either in flexion or in extension, the carpal tunnel pressure increases. To investigate the dynamic changes of the carpal tunnel shape during wrist motion, as well as the variations of space for the median nerve and its signal intensity in T2-weighting, magnetic resonance imaging (MRI) was performed on patients and healthy volunteers alike. Restitution and the persistence of pathological findings were assessed pre- and postoperatively. MRI (1.0 T) was performed on 20 wrists of patients with clinical symptoms of carpal tunnel syndrome (CTS) and pathological nerve conduction values. Healthy volunteers (20 wrists) were matched according to sex and age. MRI was performed in neutral, 45-degree extension, and 45-degree wrist flexion positions. T2-weighted signal intensity of the median nerve was measured in 18 patients pre- and postoperatively. The cross-sectional area of the carpal tunnel in patients with CTS tends to be smaller than that found in nonsymptomatic volunteers. The cross-sectional area of the carpal tunnel decreases during wrist flexion at the pisiform and hamate level. During wrist extension, the cross-sectional area of the carpal tunnel decreases at the level of the pisiform. During extension, it increases at the level of the hamate. The cross-sectional area of the median nerve showed an increase at the pisiform level (P < 0.05), a flattening of the median nerve at the hamate hook level (P < 0.05), and palmar deviation of the flexor retinaculum at the pisiform and hamate hook level (P < 0.001). This was significantly greater in CTS patients than in individuals with normal wrists. Postoperatively, the distal flattening of the median nerve recovered in 94% of the cases reviewed. Although the signal intensity of the median nerve on T2-weighted images decreased by 67%, the motor latency recovered in only 39% of the cases. The carpal tunnel