Denoising of ECG based on Wavelet Energy Entropy and an Improved-thresholding Function

摘要

Electrocardiogram (ECG) signal's noise reduction is critical for ECG automatic diagnosis and analysis. In this work, an improved de-noising method is proposed based on the wavelet energy entropy (WEE) and a new wavelet thresholding function for ECG de-noising. To overcome the disadvantages of hard and soft thresholding function, a new wavelet thresholding function without any experience parameter is proposed. Together with the WEE based method for choosing the optimal threshold, the improved wavelet threshold de-noising method has been demonstrated to achieve satisfactory performance for ECG de-noising, especially for ECG with non-stationary white noise.