Title: Classification of ECG signals using cross-recurrence quantification analysis and probabilistic neural network classifier for ventricular tachycardia patients

Authors: Shipra Saraswat; Geetika Srivastava; Sachidanand Shukla

Addresses: Amity School of Engineering & Technology, Amity University, Uttar Pradesh, India ' Amity School of Engineering & Technology, Amity University, Uttar Pradesh, India ' Dr. RML Avadh University, Faizabad, India

Abstract: Ventricular Tachycardia (VT) is one of the leading causes of sudden cardiac death in the world. Prediction of VT is usually diagnosed by using Electrocardiogram (ECG) and requires expeditious treatment which reduces the mortality rate. The cross recurrence plot (CRP) toolbox is used for computing the recurrence rate values for both (healthy and unhealthy) subjects and artificial neural network (ANN) toolbox in Matlab is used for generating the accurate results. Radial basis function neural network (RBFNN) is used for designing the probabilistic neural network classifier for discriminating the normal from abnormal (VT) signals based on the recurrence rate values. This paper illustrates the cross recurrence quantification analysis (CRQA) of ECG signals followed by the decomposition method using discrete wavelet transform (DWT) for the analysis of cardiac disorders with sensitivity, specificity of 98.5% and 97.6% respectively and overall accuracy achieved is 98.7%. This paper is useful in adopting automated approach for detecting the cardiac arrhythmias efficiently.

Keywords: electrocardiogram; artificial neural network; ventricular tachycardia; cross-recurrence quantification analysis; ventricular arrhythmia; ventricular flutter; ventricular fibrillation; ventricular tachycardia; discrete wavelet transform; MIT-BIH; daubechies; probabilistic neural network.

DOI: 10.1504/IJBET.2018.089308

International Journal of Biomedical Engineering and Technology, 2018 Vol.26 No.2, pp.141 - 156

Received: 30 Apr 2016
Accepted: 27 Jul 2016

Published online: 06 Jan 2018 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article