Abstract. The reconstruction of the corticospinal tract in the human brain is a clinically important task for both surgical planning and popula-tion studies. Diffusion MRI tractography provides an in-vivo and patient-specific technique for mapping the tract’s geometry; however, its rela-tionship to other bundles, such as the superior longitudinal fasciculus, presents issues for the standard tensor model, as it cannot represent their crossing fibers. We explore multi-fiber models that have been shown to overcome some of these issues, and evaluate methods for improving on previous work with model-based filtering of orientations. We conduct ex-periments with real clinical data including normal and tumor-infiltrated corticospinal tracts and compare the single tensor, multi-tensor, and fil-tered multi-tensor approaches. We found the multi-fiber approach to al-low for lateral projections of the tract to be reconstructed and found the addition of orientation filtering to reduce outlier fibers and increase the number of lateral projections. Our results suggest this approach could be considered for clinical applications of corticospinal tract modeling.

Abstract We present and evaluate a bilateral filter for smoothing diffusion MRI fiber orientations with preservation of anatomical boundaries and support for mul-tiple fibers per voxel. Two challenges in the process are the geometric structure of fiber orientations and the combinatorial problem of matching multiple fibers across voxels. To address these issues, we define distances and local estimators of weighted collections of multi-fiber models and show that these provide a basis for an efficient bilateral filtering algorithm for orientation data. We evaluate our approach with ex-periments testing the effect on tractography-based reconstruction of fiber bundles and response to synthetic noise in computational phantoms and clinical human brain data. We found this to significantly reduce the effects of noise and to avoid artifacts introduced by linear filtering. This approach has potential applications to diffusion MR tractography, brain connectivity mapping, and cardiac modeling. 1