A unified framework for geometry-independent operator learning in cardiac electrophysiology simulations

2025-12-01

AI Summary

Biophysical models for left atrial electrophysiological simulation are computationally expensive, hindering real-time clinical decision-making and large-scale population analysis. Method: We propose a geometry-agnostic neural operator framework that enables rapid prediction of local activation time (LAT) fields across diverse anatomical geometries. Our approach introduces a universal atrial coordinate system to decouple electrophysiological dynamics from mesh topology, establishing a domain-invariant biophysical mapping; it employs a vision Transformer–based neural operator trained on large-scale synthetic data generated by a GPU-accelerated solver. Contribution/Results: The method achieves a mean prediction error of only 5.1 ms within a maximum activation time of 455 ms, with inference latency of just 0.12 ms per sample—surpassing state-of-the-art operator learning methods in both accuracy and efficiency. This scalable framework advances personalized arrhythmia diagnosis and therapy, and enables high-throughput electrophysiological analysis.

Technology Category

• Machine Learning: Hardware-aware ML
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Accurate maps of atrial electrical activation are essential for personalised treatment of arrhythmias, yet biophysically detailed simulations remain computationally intensive for real-time clinical use or population-scale analyses. Here we introduce a geometry-independent operator-learning framework that predicts local activation time (LAT) fields across diverse left atrial anatomies with near-instantaneous inference. We generated a dataset of 308,700 simulations using a GPU-accelerated electrophysiology solver, systematically varying multiple pacing sites and physiologically varied conduction properties across 147 patient-specific geometries derived from two independent clinical cohorts. All anatomical and functional data are expressed in a Universal Atrium Coordinate system, providing a consistent representation that decouples electrophysiological patterns from mesh topology. Within this coordinate space, we designed a neural operator with a vision-transformer backbone to learn the mapping from structural and electrophysiological inputs to LAT fields. With a mean prediction error of 5.1 ms over a 455 ms maximum simulation time, the model outperforms established operator-learning approaches and performs inference in 0.12 ms per sample. Our framework establishes a general strategy for learning domain-invariant biophysical mappings across variable anatomical domains and enables integration of computational electrophysiology into real-time and large-scale clinical workflows.

Problem

• Predicts local activation times across diverse atrial anatomies instantly
• Learns geometry-independent biophysical mappings for electrophysiology simulations
• Enables real-time clinical integration of computational electrophysiology workflows

Innovation

• Geometry-independent operator-learning framework for cardiac simulations
• Universal Atrium Coordinate system decouples patterns from mesh
• Vision-transformer neural operator predicts activation times across anatomies