Fuzzy anatomical connectedness of the brain using single and multiple fibre orientations estimated from diffusion MRI.

A new fuzzy algorithm for assessing white matter connectivity in the brain using diffusion-weighted magnetic resonance images is presented. The proposed method considers anatomical paths as chains of linked neighbouring voxels. Links between neighbours are assigned weights using the respective fibre orientation estimates. By checking all possible paths between any two voxels, a connectedness value is assigned, representative of the weakest link of the strongest path connecting the voxel pair. Multiple orientations within a voxel can be incorporated, thus allowing the utilization of fibre crossing information, while fibre branching is inherently considered. Under the assumption that paths connected strongly to a seed will exhibit adequate orientational coherence, fuzzy connectedness values offer a relative measure of path feasibility. The algorithm is validated using simulations and results are shown on diffusion tensor and Q-ball images.