An improved EMD based ECG denoising method using adaptive switching mean filter

Abstract

Electrocardiogram (ECG) conveys numerous clinical information on cardiac ailments. For proper diagnosis of cardiac disorders, high-quality ECG signals are always desired. However, in reality, pathological ECG signals are corrupted with several noises. In this work, empirical mode decomposition (EMD) along with adaptive switching mean filtering (ASMF) based ECG denoising technique has been proposed. Initially, an EMD based approach is utilized for eliminating high-frequency noises and enhancing QRS complexes in the ECG signal. Then, an ASMF operation is performed for further improvement of the signal quality. The validity of the performance of the described technique is evaluated on standard MIT-BIH arrhythmia database. Gaussian noise at different signal to noise ratio (SNR) levels are added to the original signals. A close study of the simulation and performance parameters indicate that the described technique outperforms the existing methods for denoising of real ECG signals.