Phase Gradient Divergence for the Quantitative Detection of Focal Activation Events During Cardiac Fibrillation

Abstract

PurposeAtrial fibrillation is the most common arrhythmia. Spiral wave and focal activation (FA) are known to play an important mechanistic role in the generation, sustenance, and termination of tachyarrhythmia. However, to date no quantitative method of detecting FA under complex excitations has yet been established.MethodsIn this study, we propose such a detection method of FA by calculating the divergence of the spatial gradient of a phase map, which identifies the phase of the excitation cycle at each point of the heart as derived from measurement signals. Next, to verify the accuracy of the proposed method, we conducted a membrane potential measurement experiment using an excised porcine heart (n = 1).ResultsBy comparing the conventional and proposed methods for 126 instances of FA, we found that the proposed method showed improved detection accuracy, with precision, sensitivity, and F-measure values of 0.45, 0.84, and 0.58, while conventional method showed 0.04, 0.26, and 0.08, respectively.ConclusionThe proposed method, which uses the divergence of the phase gradient to predict FA, shows potential to suppress false positives that are observed in the conventional method, and it can more accurately detect FA than conventional methods.