Diffeomorphic Shape Matching by Operator Splitting in 3D Cardiology Imaging

Abstract

We develop an operator splitting approach to solve diffeomorphic matching problems for sequences of surfaces in three-dimensional space. The goal is to smoothly match, at a very fast rate, finite sequences of observed 3D-snapshots extracted from movies recording the smooth dynamic deformations of "soft" surfaces. We have implemented our algorithms in a proprietary software installed at The Methodist Hospital (Cardiology) to monitor mitral valve strain through computer analysis of noninvasive patients' echocardiographies.

arXiv comment18 pages; 8 figures

References in corpus (7)

• VoxelMorph: A Learning Framework for Deformable Medical Image Registration
• Unsupervised Learning of Probabilistic Diffeomorphic Registration for Images and Surfaces
• Learning a Probabilistic Model for Diffeomorphic Registration
• An inexact Newton-Krylov algorithm for constrained diffeomorphic image registration
• CLAIRE: A distributed-memory solver for constrained large deformation diffeomorphic image registration
• Fast GPU 3D Diffeomorphic Image Registration
• Distributed-memory large deformation diffeomorphic 3D image registration