Recognition of ST segment of electrocardiogram based on wavelet transform

Objective. The aim is to research the extraction of R waves and ST segment feature points based on Wavelet Transform (WT), and propose an algorithm used to recognize the shapes of ST segment. Method. First, electrocardiogram (ECG) signals are decomposed by WT algorithm using a dyadic spline wavelet. Based on the relation between the feature points of ECG signals and the maximum pairs of the signals’ WT, a scheme is developed to identify feature points at different wavelet decomposition scales. Then, an identification algorithm is presented to recognize the shapes of ST segments by lines fitting. At last, the proposed scheme is demonstrated by the datum from the standard MIT/BIH ECG database. Results. The automatic ECG analysis system based on the proposed scheme can extract the ECG feature points correctly and recognize the shapes of ST segment successfully. Conclusion. The proposed algorithm can be used to analyze ST segment of ECG correctly and reliably, and the result is greatly useful for clinical diagnosis of coronary heart disease (CHD). [Life Science Journal. 2007; 4(2): 90 – 93] (ISSN: 1097 – 8135).