Efficient automatic detection of QRS complexes in ECG signal based on reverse biorthogonal wavelet decomposition and nonlinear filtering

The Electrocardiogram (ECG) is one of the most used signals in the diagnosis of heart diseases. It refers to the electrophysiological activity of the heart. The ECG contains different waves and intervals that directly correlate with heart activity. In order to diagnose heart diseases, such as waves and intervals must be correctly detected and measured. Therefore, different methods are proposed and used in the research literature in order to detect, measure and analyze these waves and consequently help to diagnose heart activity. In this paper, we are interested more particularly in the detection of the QRS complex. Such QRS complex represents the depolarization state of the heart activity. To detect QRS complex, the proposed algorithm procedures use wavelet transforms. In fact, the proposed method is based on Wavelet transform decomposition of ECG signal using reverse Biorthogonal mother wavelet. The resulting coefficients are filtered using non-linear filters (averaging and median filter. The resulting filtered coefficients are then passed through the dynamic threshold to detect all the QRS complexes. The proposed algorithm is implemented and tested on a set of ECG recordings (near one million beats) taken from respectively the EUROPEAN STT, MITBIH and MITBIHST databases. The obtained results are very interesting in terms of sensitivity, positive predictive value and error (respectively 99.73%, 99.90%, and 0.37%) compared to other works in the field.