Efficient Filtering Framework for Electrocardiogram Denoising

Abstract

A simple and efficient method to remove white Gaussian noises and physiological noises from electrocardiogram (ECG) signals is presented. It is based on simple tools usually used in digital signal processing like moving average filter, median filter, baseline drift removal and peak detection. We show by several simulations that the proposed algorithm outperforms significantly conventional median filter and moving average filter and can be considered as a valid concurrent to the standard wavelet-based method.

Keywords

• ecg denoising
• median filter
• moving average filter
• statistical estimation
• discrete cosine transform (dct)
• base wandering removal