Use of diffusion tensor imaging to evaluate weakness

Abstract

Object Recently, diffusion tensor (DT) imaging was introduced to demonstrate white matter tracts. However, research interest has focused on the anatomical rather than the functional aspects of this imaging modality. The authors undertook a functional analysis of DT imaging to determine the relationship between weakness and changes on DT images. Methods Diffusion tensor images were obtained in 23 patients with lesions located adjacent to the pyramidal tract. Patients were classified according to their motor deficit. Axial magnetic resonance image sections through the maximum tumor diameters were selected and the mean apparent diffusion coefficients (ADCs) and mean fractional anisotropies (FAs) were measured. One ovoid region of interest (ovROI) was placed in the center of the pyramidal tract and another was designed to include the whole pyramidal tract at the same axial level (wROI). To determine intraobserver variability, a single neurosurgeon measured mean ADCs and FAs four times by using these two different ROI types without knowledge of any clinical information. To determine interobserver variability, a second neurosurgeon who was also unaware of any clinical information measured the mean ADCs and FAs by using the wROI method. The five measurements produced the same results. The mean FA at the lesion side of the pyramidal tract was significantly lower in patients with weakness (p < 0.01). Little intraobserver measurement variability occurred using the ovROI method, and no interobserver variability occurred using the wROI method. Conclusions Motor weakness was significantly related to a low mean FA in the pyramidal tract on the lesion side. Designing an ROI that includes the whole pyramidal tract is an easier and more reproducible method than using an ovROI method.