A New Neural Network System for Arrhythmia's Classification

A new neural network system for classification of the cardiac rhythm is presented in this paper. The system is composed of two neural network classifiers : a morphological classifier cascaded to a timing classifier. While the morphological classifier classify the P and QRS complexes into normal and/or abnormal beats, the timing classifier takes as inputs the information of the morphological classifier and the duration of the PP, PR and RR intervals and output the following arrhythmias: sinus tachycardia, sinus bradycardia, sinus arrhythmia, atrial extrasystoles, atrial tachycardia, atrial fibrillation, atrial flutter, ventricular tachycardia, ventricular extrasystoles, ventricular flutter and supraventricular tachycardia in addition to the normal sinus rhythm.