A New Algorithm for Removal Baseline Wander in ECG Signal Based on Empirical Mode Decomposition

摘要

Removal of baseline wander in ECG signal is a classical problem. In this paper, we propose a new method for removing the baseline wander interferences based on Empirical Mode Decomposition (EMD). The method is very sensitive to end conditions, that is, the end effect. We propose a novel method to restrict the end effect and improve the precision of EMD. EMD can adaptively decompose ECG signal into a series of Intrinsic Mode Functions. The baseline wander is mainly involved in special IMFs. Then selective reconstruction is performed to restore the "clean" ECG signal. To evaluate the performance of the method, Clinic ECG signals are used. Results indicate that the method is powerful and useful in removing the baseline wander in ECG signal and does not distort the ECG signal.