Dean Kazuo Shibata MD, Presenter: Consultant, HealthHelp

Todd L. Richards PhD, Abstract Co-Author: Nothing to Disclose

Yoshimi Anzai MD, Abstract Co-Author: Nothing to Disclose

Although there has been much interest in applying diffusion tensor imaging (DTI) to multiple sclerosis, most studies have focused on fractional anisotropy and mean diffusivity. The purpose of this study was to perform a histogram-based analysis of axial and radial diffusivity and correlate with disease severity.

Twenty-six consecutive patients with clinical diagnosis of multiple sclerosis were studied. All scans were performed at 3T using a single shot echo planar 15-direction DTI 2mm 256x256 axial pulse sequence. DTI data was processed using FSL (Oxford, UK) to generate DTI tensors, fractional anisotropy values, eigenvalues (L1, L2, L3), and mean diffusivity (MD) for every voxel within the brain. Tract-based spatial statistics (TBSS software as part of FSL) was performed on the DTI data in order to coregister white matter tracts from all subjects and to generate skeletonized white matter maps (all normalized to the same skeleton area) which provides a way to coregister white matter tracts to a common space. Histogram analysis was then performed on the individual skeletonized FA, L1, L2, L3, and MD maps to derive histogram peak intensity, histogram peak position, and histogram full-width at half maximum. Image-based disease severity was scored on a seven-point scale based on the axial FLAIR images.

When compared with disease severity, the best correlation was with L2 histogram peak amplitude, a measure of radial diffusivity diversity, with P<0.04. In particular, there was a negative correlation with more severe patients having lower L2 peak amplitudes. There was a converse but slightly less significant correlation with the full-width at half maximum. There was similar but reduced correlation with L3, the other measure of radial diffusivity. Correlations with FA, MD, and radial diffusivity parameters were not statistically significant.

Although this is a preliminary study, it demonstrates the value of a histogram-based analysis of the DTI eigenvalues, particularly the increased diversity of radial diffusivity, as parameters which may be more sensitive to multiple sclerosis disease severity than FA or MD.

For brain diffusion tensor imaging in multiple sclerosis patients, a histogram analysis of the radial diffusivity may correlate best with disease severity.

Shibata, D, Richards, T, Anzai, Y, Diffusion Tensor Radial Diffusivity Histogram Parameters Correlate with Severity in Multiple Sclerosis.  Radiological Society of North America 2010 Scientific Assembly and Annual Meeting, November 28 - December 3, 2010 ,Chicago IL. http://archive.rsna.org/2010/9013347.html