A Monte-Carlo approach for estimating white matter density in HARDI diffusion data

Diffusion weighted MRI (DWI) has been widely used to study human brain structure and disease processes. Due to its simplicity and the presence of semi-quantitative measures (such as fractional anisotropy (FA)), the diffusion tensor (DT) model has become the de facto standard for analysis of DWI data. However the DT model is limited by its inability to resolve crossing fiber populations and thus cannot be trusted in the third of white matter voxels with more than a single fiber population [1]. More recently, the use of higher order models such as the constrained spherical deconvolution (CSD) [2] using high angular resolution diffusion imaging (HARDI) have overcome some of these limitations. In this abstract we use test the influence of the number of averages and cutoff values on Whole-brain Track-Density (visitation) maps [3], generated by Monte Carlo methods using repeated probabilistic tracking. Furthermore, we test the applicability of VBM analysis on visitation maps.