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

摘要

Mapping resolution has recently been identified as a key limitation insuccessfully locating the drivers of atrial fibrillation. Using a simplecellular automata model of atrial fibrillation, we demonstrate a method bywhich re-entrant drivers can be located quickly and accurately using acollection of indirect electrogram measurements. The method proposed employssimple, out of the box machine learning algorithms to correlate characteristicelectrogram gradients with the displacement of an electrogram recording from are-entrant driver. Such a method is less sensitive to local fluctuations inelectrical activity. As a result, the method successfully locates 95.4% ofdrivers 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 withan arbitrary number of drivers. Extending the technique for use in clinicalpractice could alleviate the limitations in current ablation techniques thatarise from limited mapping resolution.