Adaptive diffusion basis functions decomposition for estimating intra-voxel myocardium fiber geometry

Diffusion basis functions decomposition for recovering intra-voxel fiber tract geometry from diffusion weighted MRI data has been proposed. In this formulation, the intra-voxel information is recovered at voxels containing fiber crossings or branching via the use of a linear combination of a discrete set of diffusion basis functions. Then, the parametric representation of the intra-voxel fiber geometry is described as a discrete mixture of Gaussians. The eigenvalues of tensor basis are estimated at the same time as the rest of the unknown parameters by employing an adaptive approach. The performance of our method is evaluated using both simulated and real diffusion-weighed MR data and compared with existing approach. The results show that our adaptive method outperforms existing method.