Application of Independent Component Analysis for Resolving Multiple Fibre Orientations in DT-MRI data
Papadakis N, Vigon L
The University of Sheffield
The use of standard diffusion tensor (DT) models in the brain is based on the fundamental assumption that the diffusion of water molecules is governed by a Gaussian probability density function. However, in cases where the fibers cross within a voxel, this assumption does not hold and the fiber orientation will be estimated erroneously. In this study, a new method for resolving multiple intravoxel fiber orientations is proposed based on Independent Component Analysis. The method was successfully applied in simulated two-fibre DW signals. Its robustness with respect to varying fiber separation angles as well as number of gradient orientations was also demonstrated.