ECG signal conditioning by morphological filtering.

Clinically obtained electrocardiographic (ECG) signals are often contaminated with different types of noise and baseline drifting commonly occurs. In order to facilitate automated ECG analysis, signal conditioning is undoubtedly a necessity. In this paper, a modified morphological filtering (MMF) technique is used for signal conditioning in order to accomplish baseline correction and noise suppression with minimum signal distortion. Compared with existing methods for ECG signal conditioning, MMF performs well in terms of the filtering characteristics, low signal distortion ratio, low computational burden as well as good noise suppression ratio and baseline correction ratio.