Noise Removal from Surface Respiratory EMG Signal

Authors: Slim Yacoub, Kosai Raoof

Abstract:

The aim of this study was to remove the two principal noises which disturb the surface electromyography signal (Diaphragm). These signals are the electrocardiogram ECG artefact and the power line interference artefact. The algorithm proposed focuses on a new Lean Mean Square (LMS) Widrow adaptive structure. These structures require a reference signal that is correlated with the noise contaminating the signal. The noise references are then extracted : first with a noise reference mathematically constructed using two different cosine functions; 50Hz (the fundamental) function and 150Hz (the first harmonic) function for the power line interference and second with a matching pursuit technique combined to an LMS structure for the ECG artefact estimation. The two removal procedures are attained without the use of supplementary electrodes. These techniques of filtering are validated on real records of surface diaphragm electromyography signal. The performance of the proposed methods was compared with already conducted research results.

Keywords: Surface EMG, Adaptive, Matching Pursuit, Powerline interference.

Digital Object Identifier (DOI): doi.org/10.5281/zenodo.1334610

References:

[1] Deng.Y, Wolf W, Schnell R, Appel U. New aspects to eventsynchronous cancellation of ECG interference: an application of the method in diaphragmatic EMG signals. IEEE Trans Biomed Eng. 2000; 47(9):1177-1184.
[2] Sinderby. C, Lindström. L, Grassino. A.E. Automatic assessment of electromyogram quality. Automatic EMG analysis. Journal of Applied Physiology. 1995; (79): 1803-1815.
[3] Marque C, Bisch C, Dantas R, Elayoubi S, Brosse V, Perot C. Adaptive filtering for ECG rejection from surface EMG recordings. Journal of electromyography and kinesiology 2005; (15):310-315.
[4] Levkov C, Mihov G, Ivanov R, Dascalov I, Christov I, Dotsinsky I. Removal of power-line interference from the ECG: a review of the subtraction procedure. Biomedical Engineering Online 2005, 4:50 doi:10.1186/1475-925X-4-50.
[5] Gideon.F, Inbar & Antoine.E Noujaim . On surface EMG spectral characterization and its application to diagnostic classification. IEEE Transactions on Biomedical Engineering, 1984; 31(9):597-604.
[6] Hualou Liang, Zhiyue Lin, Fuliang Yin. Removal of ECG contamination from Non linear Analysis 2005; (63): 745-753.
[7] Drake J D M, Callaghan J P. Elimination of electromyogram contamination from electromyogram signals: An evaluation of currently used removal techniques. Journal of electromyography and kinesiology 2006; (16):175-187.
[8] Huhta J C, Wbester J G. 60 Hz interference in electrocardiography. IEEE Trans Biomed Eng. 1973; 20:91-100.
[9] Metting Van Rijn, Peper A, Grimbergen CA. High-quality recordings of bioelectrical events, Part 1: Interference reduction, theory and practice." Med Biol Eng Comput. 1990; 28:389-397.
[10] Redfern. M.S, Hughes.R.E, Chaffin. D.B. High-Pass filtering to remove electrocardiographic interference from torso EMG recordings. J Clin Biomech 1993; (8) : 44-48.
[11] Widrow B, John R, Glover J.R, McCool & al. Adaptive noise cancelling principles and applications. Proceeding of the IEEE. 1975; 63 (12):1692- 1716.
[12] Widrow B, McCool .J.M, Larimore.M.G, Richard Johnson.G. Stationary and nonstationary learning characteristics of the LMS adaptive Filter. Proceeding of the IEEE. 1976; 64 (8): 1151-1162.
[13] Deluca.C.J, Fever.R.S, and Stulen.F.B. Pasteless electrode for clinical use. Med Biol Eng Comput .1979; (17) : 387-390.
[14] Levkov.CH. Fast integer coefficient FIR filter to remove the A.C interference and the high frequency noise component in biological signals.Med Biol Eng Comput. 1989; (27) :330-332.
[15] Kunt. M, Rey.H, Ligtenberg.A. Pre-processing of electrocardiograms by digital technics. Signal Processing. 1982; (4) : 215-222.
[16] Furns. G.S, Tompkins. W.J. A learning filter for removing noise interference. IEEE Trans Biomed Eng. 1983; (30) : 234-235.
[17] Baratta RV, Solomonow M, Zhou B-H, Zhu M. Method to reduce the variability of EMG power spectrum estimates, Journal of electromyography and kinesiology 1998; 8 :279-285.
[18] Bensaadoun Y, Raoof K, Novakov E. Elemination du 50Hz du signal ECG par filtrage adaptatif multidimensionnel. Innov. Techn .Biol .Med, 1994;15 (6) :750-759.
[19] Bahoura M. Hassani S.G. Lee et M. Hubin. Modification de la méthode de Widrow pour l'élimination de l'interférence 50 Hz du signal ECG. Innov. Tech. Biol. Med. 1997; 18 (2): 119-127.
[20] Lindström. L.H and Magnusson. H.N. Interpretation of myoelectric power spectral model and its application. IEEE Trans Biomed Eng. 1976; 65 (5): 653-662.
[21] Raoof K. Traitement du signal electromyographique des muscles respiratoires et estimation des paramètres en temps réel. Joseph fourier University of Grenoble France Phd thesis 1993.
[22] Yacoub S, Ben Brahim J, Ketata R, Gumery P Y, Raoof K. Real time Multidimensional treatments of surface electromyographic signals by electrodes array. Innov. Techn .Biol .Med. 1997;18 (4) :267-275.
[23] Mallat S and Zhang. Z. Matching pursuit with time frequency dictionaries. IEEE Trans Signal Processing. 1993 ; (41) 3397-3415.
[24] Mallat S. Une exploration des signaux en ondelette . Ecole polytechnique 2000.
[25] Davis G, Mallat S and Zhang Z. Adaptive time-frequency decompositions. Optical Engineering, 1994 ; 33 (7): 2183-2191.
[26] Cherkassky V, Kilts S. Myopotential denoising of ECG signals using wavelet thresholding methods. Neural Networks 2001; 14 (8): 1129- 1137.
[27] Ping Zhou, Todd A Kuiken. Eliminating cardiac contamination from myoelectric control signals developed by targeted muscle reinnervation. Physiological Measurement 2006; 27: 1311-1327.