Learning Shape Priors for Robust Cardiac MR Segmentation from Multi-view Images
This addresses robust segmentation for cardiac analysis, offering incremental improvements in accuracy and data efficiency over baseline models.
The paper tackles cardiac MR segmentation by learning shape priors from multi-view images to improve left ventricular myocardium segmentation, achieving reductions in mean Hausdorff distance from 3.24 to 2.49 on apical slices, 2.34 to 2.09 on middle slices, and 3.62 to 2.76 on basal slices with only 10% training data.
Cardiac MR image segmentation is essential for the morphological and functional analysis of the heart. Inspired by how experienced clinicians assess the cardiac morphology and function across multiple standard views (i.e. long- and short-axis views), we propose a novel approach which learns anatomical shape priors across different 2D standard views and leverages these priors to segment the left ventricular (LV) myocardium from short-axis MR image stacks. The proposed segmentation method has the advantage of being a 2D network but at the same time incorporates spatial context from multiple, complementary views that span a 3D space. Our method achieves accurate and robust segmentation of the myocardium across different short-axis slices (from apex to base), outperforming baseline models (e.g. 2D U-Net, 3D U-Net) while achieving higher data efficiency. Compared to the 2D U-Net, the proposed method reduces the mean Hausdorff distance (mm) from 3.24 to 2.49 on the apical slices, from 2.34 to 2.09 on the middle slices and from 3.62 to 2.76 on the basal slices on the test set, when only 10% of the training data was used.