Neural Surrogate Forward Modelling For Electrocardiology Without Explicit Intracellular Conductivity Tensor

TL;DR

This study introduces a deep learning model that predicts ECGs from atrial electrical potentials without needing explicit conductivity tensors, reducing structural assumptions in cardiac electrophysiology modeling.

Contribution

It presents a novel neural surrogate model that bypasses the need for explicit intracellular conductivity tensors in ECG forward modeling.

Findings

Achieved an R2 of 0.949 1.037 in predicting ECGs from electrical potentials.

Trained on 74 subjects, demonstrating high accuracy and potential clinical utility.

Reduces structural uncertainty in cardiac electrophysiology models.

Abstract

Accurate forward modelling is essential for non-invasive cardiac electrophysiology, particularly in atrial fibrillation, where electrical activation is highly disorganised. Conventional physics-based forward models require explicit specification of intracellular conductivity tensors, which are not directly measurable in clinical practice and introduce structural modelling errors. This proof-of-concept study presents a deep learning approach that learns a direct mapping from left atrial intracellular electrical potentials to far-field ECGs without requiring explicit intracellular conductivity inputs at inference time. Despite training only on 74 subjects, the model achieved an R2 of 0.949 \pm 0.037, highlighting potential to reduce structural uncertainty and improve non-invasive AF assessment.