Sparse Deformable Models with Application to Cardiac Motion Analysis

Deformable models have been widely used with success in medical image analysis. They combine bottom-up information derived from image appearance cues, with top-down shape-based constraints within a physics-based formulation. However, in many real world problems the observations extracted from the image data often contain gross errors, which adversely affect the deformation accuracy. To alleviate this issue, we introduce a new family of deformable models that are inspired from compressed sensing, a technique for efficiently reconstructing a signal based on its sparseness in some domain. In this problem, we employ sparsity to represent the outliers or gross errors, and combine it seamlessly with deformable models. The proposed new formulation is applied to the analysis of cardiac motion, using tagged magnetic resonance imaging (tMRI), where the automated tagging line tracking results are very noisy due to the poor image quality. Our new deformable models track the heart motion robustly, and the resulting strains are consistent with those calculated from manual labels.