[Show abstract][Hide abstract] ABSTRACT: In this paper, an automatic method to segment the blood vessel for 3D MRA (Magnetic Resonance Angiography) is presented. The segmentation process classifies MRA data into two parts: background and blood vessels. The process includes statistical model based on the voxel intensity and MRF model based on the context information of voxels. Both the models were built on 3D voxel, rather than on 2D. The proposed method is tested on the 3D Time-Of-Flight (TOF) -MRA data. The segmentation results give a good performance in extracting blood vessels.
International Conference on Machine Learning and Cybernetics, ICMLC 2011, Guilin, China, July 10-13, 2011, Proceedings; 01/2011