Feature extraction based on Wavelet Transform using ECG signal

The electrocardiographic signal (ECG) is being commonly used for diagnosis of different heart disease. A cardiac cycle consists of the P, Q, R, S and T waves [1] where the most significant ECG information is in the locations of the characteristics for these waves. The estimations of the heartbeat features are necessary in the diagnostic of several anomalies including some related arrhythmias such as tachycardia or bradycardia [2]. Different methods exist in this area [5, 6] but by now, the estimation of the ECG waves is a difficult problem because electrocardiographic signal depends on the physiological conditions of a patient to be investigated. In this paper, the feature extraction based on Wavelet Transform (WT) is applied as the first step, estimating several waveforms that can be employed for future diagnostic purposes. The proposed procedure uses different decomposition levels of WT applying the Haar wavelet function that presents the better results in the location of R-wave.