Performance Study of Neural Network Unscented Kalman Filter for Denoising ECG Signal

The aim of this paper is a removal of noise from electrocardiogram (ECG) signals. A dual unscented Kalman filter based on multilayer perceptron (MLP) has been proposed for removing the artificial white, colored Gaussian noises and non-stationary muscle artifact from ECG signals. The (MLP) is used as the nonlinear functional form of the unknown model. Dual Unscented Kalman filter (UKF) is used for the part of the algorithm that estimates the clean state and the weights of the network. The obtained results are compared with other enhancement conventional filters, such as, normalized least mean square (NLMS) and Butterworth filter (BF). The quantitative study of output of the different methods has been presented based on mean squared error (MSE), signal to noise ratio (SNR) and peak signal to noise ratio (PSNR). By considering these parameters, the experimental comparative analysis has shown that the UKF-MLP had optimal performance and capability than conventional filters for denoising ECG signal.

[1]  N. Sboui,et al.  Design of SIW iris-coupled-cavity band-pass filter circuit using wave concept iterative process method , 2017, 2017 International Conference on Information and Digital Technologies (IDT).

[2]  Jozue Vieira Filho,et al.  A wavelet-based method for power-line interference removal in ECG signals , 2018 .

[3]  J.M. Ferrero,et al.  Spectrum estimation and adaptive denoising of eElectrocardiographic signals using Kalman filters , 2006, 2006 Computers in Cardiology.

[5]  Ayad Almryad,et al.  An application of artificial neural networks to assessment of the wind energy potential in Libya , 2016 .

[6]  Maamar Bettayeb,et al.  Multi-sensor Data Fusion for Wheelchair Position Estimation with Unscented Kalman Filter , 2017, International Journal of Automation and Computing.

[7]  Christian Jutten,et al.  A Nonlinear Bayesian Filtering Framework for ECG Denoising , 2007, IEEE Transactions on Biomedical Engineering.

[8]  Ludger Klinkenbusch,et al.  Comparison of the Extended Kalman Filter and the Unscented Kalman Filter for Magnetocardiography activation time imaging , 2013 .

[9]  Sara Moein,et al.  An MLP neural network for ECG noise removal based on Kalman filter. , 2010, Advances in experimental medicine and biology.

[10]  Ben Jdira Makrem,et al.  Study of speaker recognition system based on Feed Forward deep neural networks exploring text-dependent mode , 2016, 2016 7th International Conference on Sciences of Electronics, Technologies of Information and Telecommunications (SETIT).