Automatic QRS-complex peak detector based on moving average and thresholding

Abstract

In this article, we present QRS-complex peak detection using moving average and thresholding process. In the QRS-complex detection, electrocardiogram (ECG) signal is detrended to remove the baseline shift of the signal. Detrended ECG signal is passed through band pass filter that consist of low pass and high pass filter to remove various noises that present in the signal. The output of ECG signal from filtering process is differentiated to provide the slope of QRS-complex. To make all ECG data point positive, derived ECG signal is processed by squaring process. The final process to detect QRS-complex is passing the ECG signal to moving average and thresholding process. We use ECG data that taken directly from patient with AD8232. The signal that is obtained from AD8232 is processed in Arduino UNO. Processed signal is sent to Android Smartphone and recorded in this device. The accuracy, positive predictive, and sensitivity of QRS-complex detection are 98.28%, 99.01%, and 99.26%, respectively.