A Hybrid Approach for Enhancement of the ECG signal

Electrocardiogram is used to observe the functioning of the heart and the obtained ECG waveform is analyzed to find the traces of any abnormalities in the function of heart which can be indications for various cardiovascular diseases. Various types of noises corrupt the signal and make it difficult to extract the actual features of the ECG signal. It is essential to decrease the intensity of noise in the acquired signal to get an accurate estimation of the signal properties. In this paper, a hybrid approach is proposed based on Empirical mode decomposition (EMD) combined with Savitzky - Golay filter (S Golay) and wavelet transform (EMD-S Golay-WT). The hybrid denoising algorithm namely EMD – S Golay – WT have been implemented along with the denoising techniques namely EMD – WT and EMD - S Golay. The performance of these techniques is simulated on several records available in MIT BIH Arrhythmia data base. The proposed hybrid method seems to surpass the existing techniques in the case of signals affected by Electromyogram and motion artifacts noise. Our proposed hybrid method gives better SNR of 2 dB when compared with the other implemented methods