Development of a concept demonstrator for QRS complex detection using combined algorithms

This paper describes a study on combined algorithms used for classification of QRS Complex in ECG signals. The proposed algorithm/detector uses Hillbert transform on a Wavelet base for the pre-processing stage. Both Hillbert transform and Wavelet base known to be superior in reducing unwanted noise resembles in ECG signal such as baseline wander and muscle noise. In addition, the Pan-Tompkins algorithm was employed as the QRS peak detection. A testing against MIT-BIH Arrhythmias Database results in a reliable detection error rate (DER) of 98.7%. It can be concluded that, the proposed method offers significant noise reduction in pre-processing stage and still produce a reliable results even with the contaminated ECG signal.