Geometry-Free Conditional Diffusion Modeling for Solving the Inverse Electrocardiography Problem
This work addresses the need for noninvasive cardiac imaging by providing a geometry-free, data-driven method that handles the non-unique nature of the problem, though it is incremental as it applies diffusion models to a specific domain.
The paper tackles the inverse electrocardiography problem by proposing a conditional diffusion model that learns a probabilistic mapping from body surface signals to heart surface potentials, achieving improved reconstruction accuracy compared to deterministic baselines.
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.