Facing the Challenge of Estimating Human Brain White Matter Pathways

Diffusion anisotropy has been used to characterize white matter neuronal pathways in the human brain, andinfer global connectivity in the central nervous system. However, mapping complex fiber configurations invivo remains a challenging task. We present a new methodology to reduce uncertainty in estimating theorientation of neuronal pathways in high angular resolution diffusion imaging (HARDI) reconstructions. Themethodology relies on three main features. First, an optimized HARDI reconstruction technique based onthe generalized q-sampling imaging approach is adopted. Second, directional statistics are used to estimateorientation distribution function (ODF) profile directions from data distributed on the unit sphere. Third, amodified streamline algorithm able to accommodate multiple fiber tracts and multiple orientations per voxel isused, to exploit the directional information gathered from estimated ODF profiles. The methodology has beentested on synthetic data simulations of crossing fibers and on a real data set.