QVAT: QRS Complex Detection based on Variance Analysis and Adaptive Threshold for Electrocardiogram Signal

Abstract

Heart disease is a disease that has a high level of danger. Somebody who has a history of heart disease must be careful in doing daily activities. Paramedics analyze Electro-cardiogram(ECG) signals to detect heart abnormalities. Some researchers propose automated methods to analyze the heart condition based on ECG signals. One parameter for assessing heart condition is the distance from R peak to R peak, R is the peak of the QRS complex wave. In this research, we proposed QVAT algorithm that automatically detects QRS complexes, then finds R peaks from an ECG signal. The algorithm that we use consists of several steps, namely: band-pass filter, analysis of variance, adaptive threshold and local maxima. Band-pass filters are used to reduce noise that can cause errors in detection QRS waves. Possible noise due to: interference due to electromagnetic wave voltage, noise of muscle movement. The average of variance is used to strengthen the QRS Complex feature at positive x-coordinates, the adaptive threshold is used to localize the QRS complex. The result of adaptive threshold is a region of interest (ROI) of QRS Complex that is used to find the position of R peak. We use the adaptive threshold since the magnitude and slope features of each subject's ECG signal are different. Evaluating the performance of our proposed algorithm, we tested it to detect the QRS complex in the MIT BIH Arrhythmia database. The proposed algorithm QVAT has a sensitivity $Se=\mathbf{99.79} \%$ and a positive predictive $+P=\mathbf{99.90} \%$ . These results indicate sensitivity of QVAT is better than Pantompkins, Garca Rivas and Xiang. The positive predictive parameter is comparable to Xiang method however it is better than Pantompkins, Garca Rivas.