Fully automated probabilistic white-matter tractography with anatomical priors: Application to Huntington's disease

Introduction White-matter tractography has become an increasingly popular application of diffusion MRI. However, conventional tractography methods are often cumbersome to apply to large populations because of the need for manual intervention, e.g., to place start, end, or waypoints, to define inclusion or exclusion masks, or to tune parameters such as constraints on the tract bending angle, particularly for weaker connections that are more difficult to trace. In this work we combine a Bayesian framework for diffusion tractography with prior information on the tracts of interest from a set of training subjects to perform tractography in a robust and automated manner. We apply this method to trace the corticospinal tract (CST) and the three branches of the superior longitudinal fasciculus (SLF1, SLF2, SLF3) on 33 Huntington’s disease patients and 22 healthy controls and investigate population differences in fractional anisotropy (FA) along these tracts.