An Efficient Method of ECG Signals Denoising Based on an Adaptive Algorithm Using Mean Filter and an Adaptive Dual Threshold Filter

Abstract

This paper is proposing an efficient denoising method of baseline wandering and high frequency noise for ECG signals. The proposed method starts by extracting baseline wandering from ECG signal. This technique has been developed using an adaptive algorithm based on mean filter. Next, an adaptive dual threshold filter is proposed to deal with high frequency noises. This filter is inspired from an adaptive dual threshold median filter recently developed in image processing. This paper presents certain applications of this algorithm on some of the MIT-BIH Arrhythmia database’s signals. Interestingly, results of this method are very promising compared to recently published methods. In the case of an input SNR level of 5dB to 20dB, the values of the PRD parameter are varied between 25 % and 7 %. For the same input SNR level, the obtained values of the MSE parameter are varied between 0.0087 and 0.00062. The aim of this work, in addition to improving denosing, is to implement this method in real-time systems, which is the next step of this work.