Detection of R-Peaks Using Single-Scale Wavelet Transform

The wavelet-based algorithm is suitable for the automatic detection of electrocardiogram (ECG). In this paper, we have developed and evaluated an ECG feature extraction system based on a single characteristic scale in the wavelet transform. For validation purposes, we also evaluated the proposed algorithm on MIT-BIH Database (the manually annotated database). Moreover, We have compared our threshold with the Martínez et al. threshold and tested on only a single characteristic scale (a fourth-scale) instead of multiscale. The results of our proposed algorithm and Martínez algorithm for QRS detection in terms of sensitivity and positive predictivity are 92.36%, 96.97% and 87.53%, 96.82% over the validation database respectively. Furthermore, in terms of sensitivity, we have obtained a 5% better than Martínez results.