Comparizon between Wavelet Packets and Adaptive Filtering for the EHG Signal Denoising

摘要

The electrohysterogram (EHG) does represent the electrical activity of the uterus during the pregnancy. The use of such a signal to characterize the pathological pregnancy contractions would be of great interest to help the preterm delivery diagnosis. The EHG signal is corrupted by environmental noises and by the mother's electrocardiogram (ECG). Wavelets are a powerful denoising tool and have already proved their efficiency on the EHG denoising. We compare the denoising method based on the wavelet packets, to adaptive filtering. We evidence that the wavelet packets provide better results than the adaptive filtering, because they remove all kind of noise, whereas adaptive filtering is only adapted to the ECG rejection.