Deterministic tracking of contrast enhanced MR-Angiography (CE-MRA) data by using established methods of Diffusion-Tensor-Imaging (DTI)

Abstract

We propose a method to generate tractograms from CE-MRA data by using existing tools designed to work on diffusion tensor data. Pseudo diffusion tensor data were generated by converting the Hessian matrix derived from a multi-scale approach in order to obtain a highly anisotropic tensor representation along the vessels. Deterministic tracking was then applied on this pseudo tensor data set. The results derived from a high resolution 3D MR angiographic data set of the lower extremities demonstrate that it is possible to reconstruct the vascular system with diffusion tensor fiber tracking tools. So far, only deterministic tracking has been applied which may cause to a certain extent fragmentary reconstruction. This current limitation, however, may be overcome in future investigations by using probabilistic tracking.