3D Cardiac Anatomy Generation Using Mesh Latent Diffusion Models

2025-08-18

AI Summary

Addressing the challenge of generating high-fidelity, diverse 3D cardiac anatomies from medical imaging, this paper introduces MeshLDM—the first latent diffusion model (LDM) tailored for 3D mesh generation of cardiac structures. MeshLDM innovatively integrates a differentiable 3D mesh encoder with an LDM framework and incorporates clinically relevant phase features (diastolic/systolic) to enable accurate modeling of dynamic left ventricular morphology in myocardial infarction patients. Evaluated on public datasets, generated meshes achieve a Hausdorff distance error of only 2.4% relative to population-level ground-truth mean anatomy. Both qualitative and quantitative assessments demonstrate state-of-the-art performance. This work establishes a novel paradigm for high-fidelity, controllable, and generalizable 3D anatomical generation—enabling applications in cardiovascular electromechanical simulation, in vitro testing, and medical data augmentation.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Computer Vision: Diffusion Models for Vision
• Natural Language Processing: (Large) Language Models

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Diffusion models have recently gained immense interest for their generative capabilities, specifically the high quality and diversity of the synthesized data. However, examples of their applications in 3D medical imaging are still scarce, especially in cardiology. Generating diverse realistic cardiac anatomies is crucial for applications such as in silico trials, electromechanical computer simulations, or data augmentations for machine learning models. In this work, we investigate the application of Latent Diffusion Models (LDMs) for generating 3D meshes of human cardiac anatomies. To this end, we propose a novel LDM architecture -- MeshLDM. We apply the proposed model on a dataset of 3D meshes of left ventricular cardiac anatomies from patients with acute myocardial infarction and evaluate its performance in terms of both qualitative and quantitative clinical and 3D mesh reconstruction metrics. The proposed MeshLDM successfully captures characteristics of the cardiac shapes at end-diastolic (relaxation) and end-systolic (contraction) cardiac phases, generating meshes with a 2.4% difference in population mean compared to the gold standard.

Problem

• Generating diverse realistic 3D cardiac anatomies
• Applying latent diffusion models to 3D medical imaging
• Creating cardiac meshes for simulations and data augmentation

Innovation

• MeshLDM architecture for cardiac generation
• Latent Diffusion Models for 3D meshes
• Generates realistic left ventricular anatomies