HeartVolMesh: Cardiac Volumetric Mesh Reconstruction via Covariance-Guided Graph Deformation

2026-07-05

AI Summary

Existing segmentation-to-mesh approaches struggle to accurately reconstruct thin cardiac walls and lack consistent anatomical correspondence across subjects. To address these limitations, this work proposes a template-driven, high-fidelity tetrahedral mesh reconstruction method. The approach models template vertices as anisotropic Gaussian kernels and employs a joint 3D CNN–GNN architecture to predict vertex displacements and covariances—parameterized via Cholesky decomposition. A covariance-guided graph deformation mechanism is introduced to refine mesh geometry while preserving both topological integrity and cross-subject correspondence, substantially improving thin-wall reconstruction accuracy. By integrating staged alignment, non-rigid registration, and deformation propagation strategies, the method outperforms existing deformation-based models in both surface and volumetric mesh quality, enabling more precise patient-specific cardiac modeling.

Abstract

Accurate patient-specific tetrahedral cardiac meshes are essential for in-silico trials, yet common segmentation-then-modelling pipelines can blur thin-wall anatomy and offer limited cross-case correspondence. We propose HeartVolMesh, which lifts each template vertex to an anisotropic Gaussian kernel and uses a 3D CNN-GNN to predict per-vertex displacements and Cholesky-parameterized covariances from volumetric images. Training is guided by a covariance-aware negative log-likelihood loss with lightweight mesh regularization. For volumetric meshing, we warp a fixed tetrahedral template to the reconstructed surface via staged alignment, non-rigid registration, and deformation propagation, preserving connectivity and correspondence by construction, with resolution controlled by template density. Experiments show consistent gains over deformation-based baselines in surface mesh accuracy and volumetric mesh fidelity.

Problem

• cardiac mesh reconstruction
• patient-specific modeling
• volumetric meshing
• thin-wall anatomy
• cross-case correspondence

Innovation

• covariance-guided deformation
• tetrahedral mesh reconstruction
• 3D CNN-GNN
• anisotropic Gaussian kernel
• patient-specific cardiac modeling