Assessment of QRS Detection on Noisy ECG Signal Using Threshold-based Method

Abstract

The threshold-based method has been used at the last stage of most QRS detection for monitoring many cardiac diseases including arrhythmias detection. The method also been proven to achieve good detection accuracy in clean and high-quality ECG signal. However, the threshold-based method is sensitive to noise and the performance of detectors will be affected by low-quality ECG signal data. The numerous types of noise and different abnormal morphological of ECG signal will produce low-quality of signal. Because of that, the performance of the threshold-based method to detect the QRS complex in low-quality ECG signals should be investigated. In this paper, to assess the QRS detection on noisy ECG signal, the performance of threshold-based method is evaluated. Two experiments were performed using MIT-BIH Noise Stress Test data and MIT-BIH Arrhythmia data to assess the algorithm performance. The performance of the algorithm was then analysed and presented. This paper shows the capability of the threshold-based detection method to identify QRS complex in noisy ECG signals.