Uterine Electromyography signals denoising using discrete wavelet transform

This paper introduces the application of the discrete wavelet transform in noise reduction of real uterine Electromyography (EMG) signals. With the appropriate choice of the wavelet function and the selection of the suitable decomposition level, it's possible to remove interference noise effectively. Signal to Noise Ratio (SNR) values are calculated to evaluate the global performance of noise reduction. Results show that wavelet function db4 at three level of decomposition performs denoising best out of the other wavelets. Furthermore, a comparative analysis using the band-pass Butterworth filter demonstrates the superior denoising performance of the proposed algorithm.