Anatomy-Aware 3D Mesh Refinement of Pericardium Segmentations on Computed Tomography

2026-07-21

AI Summary

This study addresses the challenge of imprecise epicardial fat quantification in cardiac CT due to the poorly defined boundaries of the pericardium. To overcome this limitation, the authors propose an anatomy-aware 3D mesh iterative optimization framework that, for the first time, integrates anatomical priors into a model-agnostic post-processing pipeline. Leveraging a GPU-accelerated gradient-driven vector field, the method jointly enforces anatomical constraints and geometric deformations, iteratively refining an initial segmentation until equilibrium is achieved between anatomical plausibility and geometric fidelity. The approach consistently enhances segmentation quality across both an internal high-resolution dataset and a publicly available sparsely annotated dataset, yielding significant improvements in volumetric, surface-based, and anatomical metrics. Notably, it demonstrates robust performance in out-of-domain settings and under limited annotation scenarios.

Abstract

Accurate delineation of the pericardium in a cardiac CT scan is essential for quantifying epicardial adipose tissue, yet it remains one of the most challenging structures to segment due to its poor contrast boundaries. Instead of solely relying on image gradients, our framework leverages the anatomical context of surrounding anatomical structures to guide the segmentation. This work introduces a novel 3D iterative mesh refinement framework that balances anatomical and geometric forces derived from inherent anatomical rules to refine an initial, possibly ambiguous, segmentation into a high-precision, anatomically plausible result. Designed as a model-agnostic post-processing step, our method uses a 3D vector field to iteratively push the vertices to the correct anatomical locations. Evaluating the refinement on both a high-resolution in-house dataset and a coarse, sparsely annotated open-source dataset, our method consistently improves all volumetric, surface, and anatomical metrics. The framework demonstrates greater improvement when applied to weaker initial segmentations, highlighting its potential for improving segmentations for out-of-domain models and in limited-training-data scenarios. The method is formulated as a gradient-based, GPU-accelerated framework that can be easily extended to other anatomical use cases.

Problem

• pericardium segmentation
• computed tomography
• anatomical context
• low-contrast boundaries
• 3D mesh refinement

Innovation

• anatomy-aware
• 3D mesh refinement
• pericardium segmentation
• anatomical context
• GPU-accelerated optimization