FEATURE EXTRACTION FROM ECG SIGNAL USING WAVELET TRANSFORMS

Abstract

Wavelet Transforms are widely preferred for the analysis of non-stationary signals because of their property of multi-resolution analysis, thereby implying the analysis of the signal of different frequencies with different resolutions. They are in turn robust to noise signals which are mixed up with the signal of interest during acquisition. This report presents discrete wavelet transform based methods of pre-processing of the ECG signal for the removal of high frequency noise and low frequency baseline drift. It also presents a discrete wavelet transform based algorithm for the detection of QRS complex in the ECG signal. The algorithms have been tested using the CSE database