Transform-based algorithms dedicated to ECG baseline drift cancellation: A comparative study

Abstract

Nowadays, Electrocardiogram (ECG) diagnosis stills to be very useful to inform about the cardiac health status of humans. However, acquiring an ECG signal in a correct manner requires that all noise sources should be reduced proficiently. Among such noise types, we attract the attention to the baseline wander (BW) that can be instigated, usually, by unreliable electrode contact placement, patient breathing, unclean skin and so on. In this context, three strategies dedicated to diminish BW contribution that are: Discrete Wavelet Transform (DWT), Discrete Cosine Transform (DCT), Fast Walsh-Hadamard Transform (FWHT) based algorithms, were experimented comparatively. Achieved simulations demonstrate, qualitatively and quantitatively, the high-performance capacity of the DWT-based strategy compared to the DCT-based technique and FWHT-based method.