Comparison of Magnetic Resonance Imaging and Ultrasonography in Diagnosing and Grading Carpal Tunnel Syndrome: A Prospective Study.

OBJECTIVE

To use anatomic measurements on magnetic resonance imaging (MRI) and ultrasonography (USG) in diagnosing and grading carpal tunnel syndrome (CTS) using nerve conduction studies (NCS) as the gold standard.

MATERIAL AND

METHODS

In this prospective study, 26 patients with CTS (45 wrists; 22 female and 4 male patients; mean ± SD age of 49.42 + 14.47 years) and 19 age and sex matched healthy volunteers (32 wrists; 15 female and 4 male volunteers, mean ± SD age of 42.52 + 10.85 years) underwent MRI and USG. Cross-sectional area (CA) of median nerve was measured using free hand ROI at four levels: hamate hook (H0), pisiform bone (PI0), 1 cm proximal (PI1) and 2 cm proximal to PI0 (PI2). Relative median nerve signal intensity (MNSI) was calculated as ratio of median nerve signal intensity with hypothenar muscle signal intensity. Flexor retinacular bowing was calculated at hamate hook level. Echogenicity and Power Doppler vascularity of median nerve were assessed on USG. Independent t-test, chi square test and receiver operating characteristic curve analysis were used as appropriate.

RESULTS

On USG, CA measured at PI0 (95% confidence interval of 0.872-0.987) and retinacular bowing (0.816-0.912), while, on MRI, CA at PI1 (0.874-0.997) were most useful in diagnosing CTS based on the ROC and Zombie plot analysis. Area under curves for CA measurements on USG and MRI were not significantly different. CA at PI1 on MRI (0.752-0.965) was significantly different between minimal to moderate CTS and severe to extreme CTS groups (on NCS).

CONCLUSION

CA of median nerve is the most useful parameter to diagnose and grade CTS and USG and MRI are comparable for measurements. Increased retinacular bowing on USG and hypoechogenicity of median nerve increase the diagnostic confidence while MRI helps in picking up important associated conditions.