PrIntMesh: Precise Intersection Surfaces for 3D Organ Mesh Reconstruction

2025-11-20

AI Summary

Existing methods model anatomically connected organ substructures independently, leading to topological distortions and boundary discontinuities. To address this, we propose a template-driven 3D organ surface mesh reconstruction framework that treats the entire organ as a unified system. Our approach introduces, for the first time, a joint deformation mechanism that preserves internal geometric integrity of each substructure while explicitly constraining interfacial deformations—ensuring topology preservation, surface smoothness, and robustness to noise. By integrating deep learning with parametric deformation optimization, the method jointly learns both geometric morphology and spatial relational constraints among substructures. Evaluated on cardiac, hippocampal, and pulmonary datasets, our framework significantly outperforms state-of-the-art voxel- and surface-based methods, particularly under low-data and noisy conditions, demonstrating superior reconstruction accuracy and clinical applicability.

Technology Category

• Machine Learning: Learning with Manifolds
• Computer Vision: 3D Computer Vision
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Human organs are composed of interconnected substructures whose geometry and spatial relationships constrain one another. Yet, most deep-learning approaches treat these parts independently, producing anatomically implausible reconstructions. We introduce PrIntMesh, a template-based, topology-preserving framework that reconstructs organs as unified systems. Starting from a connected template, PrIntMesh jointly deforms all substructures to match patient-specific anatomy, while explicitly preserving internal boundaries and enforcing smooth, artifact-free surfaces. We demonstrate its effectiveness on the heart, hippocampus, and lungs, achieving high geometric accuracy, correct topology, and robust performance even with limited or noisy training data. Compared to voxel- and surface-based methods, PrIntMesh better reconstructs shared interfaces, maintains structural consistency, and provides a data-efficient solution suitable for clinical use.

Problem

• Reconstructs 3D organ meshes with preserved anatomical topology and boundaries
• Addresses implausible reconstructions from independent substructure processing
• Enhances shared interface reconstruction and structural consistency for clinical use

Innovation

• Template-based topology-preserving framework for organ reconstruction
• Joint deformation of all substructures to match patient anatomy
• Preserves internal boundaries and enforces smooth artifact-free surfaces