Classification of Cardiac Arrhythmias Using PNN Neural Network Based on ECG Feature Extraction

Abstract

This paper addresses the problem of automatic discrimination of rhythms in ECG signals. In particular the main focus of the paper is at discriminating normal sinus rhythm (NSR), ventricular tachycardia (VT), and ventricular fibrillation (VF). After de-noising ECG signal using db6 wavelet, some features of ECG signal are extracted. Then matrix feature is produced that is given to PNN neural network to classify NSR, VF and VT. Results show classification for 98.15% of NSR, 90% of VT and 93.33% of VF events.