FIBER TRACTOGROPHY AND TRACT SEGMENTATION IN MULTIPLE SCLEROSIS LESIONS

Diffusion tensor imaging provides rich information about human brain connectivity in vivo, yet most current methods for fiber tractography or tract segmentation do not address white matter pathologies such as multiple sclerosis lesions, which can alter the diffusion tensor characteristics. We study here the effects of MS lesions on estimated diffusion tensors and how they affect the processing of fibers and tracts. An efficient correction algorithm is proposed to compensate for lesion areas in two different approaches to fiber tracking and tract segmentation. Application of the algorithm to real data acquired from MS patients demonstrates improved fiber tracking through lesion regions.