Detection of ventricular fibrillation using Hilbert transforms, phase-space reconstruction, and time-domain analysis

Abstract

This study proposes feature extraction using Hilbert transforms, phase-space reconstruction, and time-domain analysis to detect ventricular fibrillation and normal sinus rhythm from electrocardiogram (ECG) episodes. We implemented three preprocessing steps to extract features from ECG episodes. In the first step, we use Hilbert transforms to extract peaks. In the second step, we use statistical methods and extract four features from the peaks. In the final step, we extract four features using statistical methods based on the Euclidean distance between the origin (0, 0) and the peaks after the peaks are plotted in a two-dimensional phase-space diagram. By applying time-domain analysis directly to the series of successive peak-to-peak interval values, we extract seven additional features. Using a neural network with weighted fuzzy membership functions (NEWFM), we applied the nonoverlap area distribution measurement method, and from 15 initial features, we selected 11 minimum features exhibiting the highest accuracy. Then, we applied the 11 minimum features as inputs to the NEWFM and recorded sensitivity, specificity, and accuracy values of 79.12, 89.58, and 87.51 %, respectively. In addition, McNemar's test revealed a significant difference between the performances of NEWFM with and without feature selection (p < 0.05).