Peripheral nerve diffusion tensor imaging is reliable and reproducible

Purpose To determine the reliability and reproducibility of peripheral nerve diffusion tensor imaging (DTI) in healthy subjects. Materials and Methods We assessed the test-retest and interrater reliability studies of peripheral nerve DTI in a cohort of 12 healthy subjects (mean age 44.0 years, range 26-71 years). Magnetic resonance imaging (MRI) studies were performed on a 3T scanner (MR750, GE Healthcare). DTI and T₁-weighted sequences were performed on the tibial and peroneal nerves in the knee. Each subject was scanned on three separate occasions. Image analyses were performed at two anatomic positions: the level of the lateral femoral condyle (position 1), and the superior pole of the patella (position 2). Analysis was performed using three software packages ("raters"): FuncTools, FSL, and Diffusion Toolkit. Metrics obtained included fractional anisotropy (FA), axial diffusivity (AD), and radial diffusivity (RD). Results DTI metrics were not significantly different between each measurement position, except peroneal nerve AD and tibial nerve RD were higher at position 1 (P = 0.02 and 0.04, respectively). There was no significant difference in DTI metrics between male and female subjects (P = 0.10-0.99). The test-retest reliability of DTI metrics was high, maximal for FA (intraclass correlation coefficient [ICC] = 0.96). ICC values for individual DTI metrics were similar between each measurement position, except that tibial nerve AD was significantly higher at position 2 (P = 0.03). Interrater reliability was also high (ICC = 0.95-0.96). Conclusion We found peripheral nerve DTI to be reliable and reproducible, with few effects related to the postprocessing package used.