Pattern recognition of cardiac arrhythmias using two intracardiac channels

Abstract

New techniques of arrhythmia detection using morphology of the waveform have shown promise in correctly detecting fatal arrhythmias. Morphologic analysis of two channels, atrial and ventricular, has the potential of improving the ability of the algorithm to distinguish arrhythmias. This study uses correlation waveform analysis of intracardiac electrograms from the right ventricular apex and the high right atrium to define a two dimensional feature space with linear decision boundaries created from a training set by a least squares minimum distance classifier. The total set represented 16 patients. This method correctly discriminated 48/48 cycles of sinus rhythm, 47/48 of ventricular tachycardia (VT), 39/48 of atrial tachycardia and 40/48 of VT with retrograde, and supraventricular tachycardia with aberrant conduction. Performing this technique on two intracardiac electrograms appears to reliably differentiate various arrhythmias and warrants further study.<>