Algorithms for detecting atrial arrhythmias from discriminatory signatures of ventricular cycle lengths

摘要

Detection of arrhythmias is facilitated using irregularity of ventricular beats measured by delta-RR (ΔRR) intervals that exhibit discriminatory signatures when plotted in a Lorenz scatter-plot. An “AF signature metric” is established characteristic of episodes of AF that exhibit highly scattered (sparse) distributions or formations of 2-D data points. An “AFL signature metric” is established characteristic of episodes of AFL that exhibit a highly concentrated (clustered) distribution or formation of 2-D data points. A set of heart beat interval data is quantified to generate highly scattered (sparse) formations as a first discrimination metric and highly concentrated (clustered) distributions or formations as a second discrimination metric. The first discrimination metric is compared to the AF signature metric, and/or the second discrimination metric is compared to the AFL signature metric. AF or HFL is declared if the first discrimination metric satisfies either one of the AF signature metric.