Machine learning methods for locating re-entrant drivers from electrograms in a model of atrial fibrillation

TL;DR

This paper presents a machine learning approach using electrogram data to accurately locate re-entrant drivers in atrial fibrillation models, potentially improving clinical ablation procedures.

Contribution

It introduces a novel, robust machine learning method that correlates electrogram gradients with driver displacement, enhancing localization accuracy in AF models.

Findings

Locates 95.4% of drivers in single-driver tissues

Achieves 94.8% accuracy in tissues with two drivers

Applicable to tissues with multiple drivers

Abstract

Mapping resolution has recently been identified as a key limitation in successfully locating the drivers of atrial fibrillation. Using a simple cellular automata model of atrial fibrillation, we demonstrate a method by which re-entrant drivers can be located quickly and accurately using a collection of indirect electrogram measurements. The method proposed employs simple, out of the box machine learning algorithms to correlate characteristic electrogram gradients with the displacement of an electrogram recording from a re-entrant driver. Such a method is less sensitive to local fluctuations in electrical activity. As a result, the method successfully locates 95.4% of drivers in tissues containing a single driver, and 94.8% (92.5%) for the first (second) driver in tissues containing two drivers of atrial fibrillation. Additionally, we demonstrate how the technique can be applied to tissues with an arbitrary number of drivers. Extending the technique for use in clinical practice could alleviate the limitations in current ablation techniques that arise from limited mapping resolution.