Geometry-Free Conditional Diffusion Modeling for Solving the Inverse Electrocardiography Problem

2026-01-26

AI Summary

This work addresses the ill-posed and non-unique inverse problem in electrocardiographic imaging (ECGI), which traditionally relies on patient-specific anatomical modeling. The authors propose, for the first time, a geometry-free conditional diffusion model that learns a probabilistic mapping from noisy body-surface potentials to epicardial potentials in a purely data-driven manner. By leveraging the diffusion process, the method enables sampling of multiple plausible solutions without requiring individualized anatomical meshes. Trained entirely with deep neural networks, the approach significantly outperforms strong baselines—including CNNs, LSTMs, and Transformers—on real-world ECGI datasets, achieving markedly improved reconstruction accuracy. These results demonstrate the effectiveness and potential of diffusion models for non-invasive cardiac electrophysiological imaging.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems

Abstract

This paper proposes a data-driven model for solving the inverse problem of electrocardiography, the mathematical problem that forms the basis of electrocardiographic imaging (ECGI). We present a conditional diffusion framework that learns a probabilistic mapping from noisy body surface signals to heart surface electric potentials. The proposed approach leverages the generative nature of diffusion models to capture the non-unique and underdetermined nature of the ECGI inverse problem, enabling probabilistic sampling of multiple reconstructions rather than a single deterministic estimate. Unlike traditional methods, the proposed framework is geometry-free and purely data-driven, alleviating the need for patient-specific mesh construction. We evaluate the method on a real ECGI dataset and compare it against strong deterministic baselines, including a convolutional neural network, long short-term memory network, and transformer-based model. The results demonstrate that the proposed diffusion approach achieves improved reconstruction accuracy, highlighting the potential of diffusion models as a robust tool for noninvasive cardiac electrophysiology imaging.

Problem

• inverse electrocardiography
• electrocardiographic imaging
• geometry-free
• underdetermined problem
• noninvasive cardiac electrophysiology

Innovation

• geometry-free
• conditional diffusion model
• inverse electrocardiography
• probabilistic reconstruction
• data-driven ECGI