Differentiation between atrial fibrillation and atrial flutter using 1d Poincare maps based on endocardial bipolar intracardiac electrograms extracted from the right atria

This paper presents a novel approach developed to characterize or differentiate between the two most common cardiac arrhythmias, i.e. Atrial Fibrillation (AFib) and Atrial Flutter (AFL), using endocardial bipolar intracardiac electrograms (ICEMs). The tool is based on a previously developed real time algorithm which achieves automated detection among several ICEMs originating from High Right Atria (HRA) during a standard Electrophysiology (EP) study in humans. The algorithm uses Dominant Frequency (DF) and Average Power Spectral Ratio (APSR) determined from Power Spectral Density (PSD) estimated using nonparametric spectral density estimation tools. Next, a nonlinear analysis is performed and phase portraits are developed followed by the 1D Poincare maps of the preprocessed ICEMs by sampling at their respective DFs. To further augment the automated detection process, the standard deviation for the cluster of points in 1D Poincare Map is determined. The standard deviation is found of the order of 10-2, 10-3 and 10-4 in the case of AFib, AFL, and, Atrial Tachycardia respectively. The results for endocardial bipolar preprocessed ICEMs are presented along with their statistical properties.