Comparison of Baseline Wander Correction Methods for Handheld ECG with Motion Artefacts

Handheld electrocardiographic (ECG) device is seen as a significant advancement in the home-based healthcare system since it enables patients to monitor their cardiac electrical activity at their own homes. However, the signal from the device tends to be bristled with severe noises, especially baseline wander. This paper presents baseline wander correction methods by subtracting the baseline estimate, using penalised least-squares (PLS) smoother and Savitzky-Golay filter, compared with a conventional high-pass filter. The procedures were tested against various conditions of simulated baseline wander superimposed on a clean ECG signal. We evaluated their performance using wavelet-based weighted percentage root mean square difference (WWPRD) between the clean and derived signals. The results showed that the baseline correction using Savitzky-Golay filter provides a better performance for baseline estimation in severe baseline scenarios, making it more suited for motion artefact removal from handheld ECG equipment.