Improving QRS detection for artifacts reduction

Abstract

Since the QRS complex in electrocardiogram signals is one of the most important tasks to describe the operation of heart, high accuracy detection for this complex should be considered. In this study one of the newest methods of QRS complex detection combined with several artifact sources reduction methods has been performed. QRS detection algorithm includes baseline drift removal, Butterworth filtering, notch filtering and extracting five special features from ECG to identify QRS complex. In order to validate the robustness of this method, four important artifact sources such as power line interference, electrode contact noise, motion artifact and muscle contraction (EMG) have been produced and combined with ECG signal. The performance of this approach against these noises based on three MIT-BIH recording classes (Normal, LQT and TWA) has been discussed with ROC (Receiver Operating Characteristics). With proposed QRS detection algorithm 100% and 94.88% accuracy has been achieved in best and worst case respectively. Thus this method has the ability to cancel respiration modulation and reduce EMG noise, motion and power line artifacts effectively.