Accurate Removal of Baseline Wander in ECG Using Empirical Mode Decomposition

Abstract

The removal of baseline wander (BW) is a very important step in the pre-processing stage of electrocardiogram (ECG). Thus, a method for accurate removing the baseline wander (BW) in ECG on the basis of empirical mode decomposition (EMD) is proposed in this paper. We briefly described the principles and characteristics of the EMD in Section II. To validate the proposed method, the recording from MIT/BIH database is used. In this paper, the traditional median filter is also applied to remove BW in ECG for comparison with our EMD method. Our experimental results show that the performance of EMD method is very good and accurate in removing BW in ECG.