Evaluation of Diffusion Tensor Imaging and Fiber Tractography of the Median Nerve: Preliminary Results on Intrasubject Variability and Precision of Measurements

Abstract

The purposes of this study were to determine the intrasubject side-to-side variability of quantitative and qualitative measures of diffusion tensor imaging (DTI) and fiber tractography of the median nerves and to determine the precision of quantitative measurements and fiber tractography. Fifteen healthy volunteers (seven men, eight women; mean age, 31.2 years) underwent DTI of both wrists with a single-shot spin-echo-based echo-planar imaging sequence (TR/TE, 7,000/103; b value 1,025 s/mm2). Postprocessing included fiber tractography and quantitative analysis of fiber length, fiber density index, fractional anisotropy, apparent diffusion coefficient, and signal-to-noise ratio. Two readers in consensus graded the quality of fiber tract images of the two wrists as equal, slightly different, or very different. Fiber tractography and all analyses were repeated after 3 weeks, and the images from the two sessions were compared. No statistically significant side-to-side differences in quantitative data were found (p=0.054-0.999). In all subjects, the quality of fiber tract images of the right and left median nerves was either slightly or very different. Between the initial and the second quantitative analyses, no statistically significant differences (p=0.086-0.898) were found, and the quality of fiber tract images was rated equal for nine of 15 subjects (60%) and slightly different for six of 15 subjects (40%). Preliminary results indicate that quantitative evaluation of DTI of the median nerve is precise. The absence of statistically significant intrasubject side-to-side variability in quantitative data suggests that the healthy contralateral nerve can be used as an internal control. Observed side-to-side variability in the quality of fiber tract images, however, rules out side-to-side comparisons in fiber tractography.