Evaluation of Regional White Matter Volume Reduction after Diffuse Axonal Injury using Voxel-based Morphometry.

PURPOSE: We developed a new and convenient method that employs voxel-based morphometry (VBM) to evaluate regional reduction in the volume of white matter after diffuse axonal injury (DAI).

METHODS: We studied 29 patients with moderate cognitive disability after DAI. Each subject underwent 3-dimensional volumetric magnetic resonance (MR) imaging. Images were preprocessed automatically using stand-alone software running on a Windows PC for VBM of volumetric MR imaging utilizing a statistical parametric mapping (SPM) version 8 software engine and an algorithm for diffeomorphic anatomic registration through exponentiated lie algebra (DARTEL). We then computed a Z-score for all coordinates on the white matter, which represented the relative reduction in white matter volume. Finally, we used voxel-based stereotactic extraction estimation (vbSEE) to compute the extent of regional reduction in the volume of white matter (rWMVR) for each region of interest (ROI), defined as the rate of coordinates with Z-scores exceeding 2.0 in the ROI. For each ROI, we used Pearson's correlation analysis to examine the correlation between the extent of regional volume reduction and patient scores on the Wechsler Adult Intelligence Scale III (WAIS-III).

RESULTS: We detected marked rWMVR in several ROIs, including the corpus callosum, and rWMVR correlated significantly with performance IQ and processing speed index in the splenium of the corpus callosum.

CONCLUSIONS: The results indicate the utility of our applications for the daily clinical evaluation of DAI. That they can be used on a PC and allow acquisition of volumetric data from standard MR images are their advantages.