ECG-Waves: Analysis and Detection by Continuous Wavelet Transform

Abstract

In this work, we have developed a new algorithm for electrocardiogram (ECG) features extraction. This algorithm was based on continuous wavelet transform (CWT). The core of the process involved analyzing the signal using the CWT coefficients with a selection of scale parameter corresponding to each ECG wave. The entry point of our method was the R peak detection. The next step was the Q and S point localization, after we identified the P and T waves. We evaluated our algorithm on apnea and MIT-BIH databases recording. The algorithm achieved a good performance with the sensitivity of 99.84 % and the positive predictive value of 99.53 %.