ECG fiducial points detection through wavelet transform

Abstract

The wavelet transform is introduced as a novel method to analyze signals in both time and frequency domains, and therefore it is suitable for analysis of time-varying signals such as ECG. Detection of ECG signal characteristic patterns has been achieved using an algorithm based on the wavelet transform decomposition. ECG signals are represented by a pyramidal algorithm at successive scales into fine and coarse components used for the successive fiducial points detection. The presented time-scale diagrams illustrate the detection capability of fiducial points on ECG signal (QRS complexes, P and T waves).