Integrated segmentation and deformation analysis of 4-D cardiac MR images

Abstract

Segmentation and motion estimation from cardiac images are usually considered separately, yet they can obviously benefit from each other. In this paper, we propose a joint segmentation and motion estimation algorithm for the purposes of myocardial deformation analysis and strain estimation. We use segmentation as a guide for selecting feature points with significant shape characteristics, and invoke a Generalized Robust Point Matching (GRPM) strategy with Boundary Element Method (BEM)-based regularization model to estimate the dense displacement field and strain map from 3-D cardiac sequences. Quantitative analysis of the results is performed in comparison with the displacements found using implanted markers, taken to be gold standards.