Multi-scale segmentation of carotid artery wall in MRI images

Abstract:

The segmentation of carotid artery wall is essential for subsequent plaque analysis. Involving with attempt of different techniques, this work will focus on obtaining the boundaries of lumen and external wall based on a comparatively accurate and efficient approach. In particular, a method described as prior-constrained carotid artery segmentation in DOG scale-space (multi-scale PCS) is proposed. Moreover, due to the significance, lumen segmentation is testified through several feasible algorithms in order to find a better option, by means of self-defined computing method to assess the quality. © 2010 IEEE.