Mixture Model for estimating fiber ODF and multi-directional Tractography

Introduction: Diffusion Tensor Imaging (DTI) is now a well-established scheme for analyzing neural pathways in the brain by means of streamline or probabilistic tractography. But DTI models the diffusion of water molecules by a Gaussian process, while the data captured by the diffusionweighted MRI (DW-MRI) could very well be non-Gaussian. To overcome this limitation, a High Angular Resolution Diffusion Imaging (HARDI) scheme was proposed [1]. Several methods have been proposed to estimate the fiber orientation distribution function (fODF) from HARDI signals. Other methods estimate multiple tensors at each voxel and subsequently perform tractography [2]. In this work we propose to use a mixture of Watson orientation functions to model the fODF and subsequently perform multi-directional streamline tractography on real datasets.