Classification of surface electromyogram signals based on directed acyclic graphs and support vector machines

This paper presents a novel classification approach for surface electromyogram (sEMG) signals. The proposedclassification approach involves two steps: (1) feature extraction from an sEMG, in which a 7-dimensional featurevector is extracted from 27 types of features of the sEMG by linear discriminant analysis (LDA), and (2) a novelclassifier, DAGSVMerr, based on a directed acyclic graph (DAG) and support vector machine (SVM), in which aseparability measure function based on erroneous recognition rates (ERRs) is defined to determine the initial operationlist. The proposed approach takes advantage of the feedback idea to improve the performance of the classification. Theexperimental results show that the proposed approach has a better performance than traditional methods, and it achievesan average classification accuracy rate of 99.4% ± 1.3% with an error rate of 0.6%. Correct classification rates of theproposed approach are very high, and the approach can be utilized to recognize gesture instructions by analyzing sEMGsignals in gesture equipment control studies.

PDF

___

[1] Purushothaman G, Ray KK. EMG based man–machine interaction—A pattern recognition research platform. Robot Auton Syst 2014; 62: 864-870.
[2] Cho YJ, Kim JY. The effects of load, flexion, twisting and window size on the stationarity of trunk muscle EMG signals. Int J Ind Ergonom 2012; 42: 287-292.
[3] Alpaydin E. Introduction to Machine Learning. 3rd ed. Cambridge, MA, USA: MIT Press, 2014.
[4] Hajibanden N, Faghihi F, Ranjbar H, Kazari H. Classifications of disturbances using wavelet transform and support vector machine. Turk J Elec Eng & Comp Sci 2017; 25: 832-843.
[5] Abel EW, Zacharia PC, Forster A, Farrow TL. Neural network analysis of the EMG interference pattern. Med Eng Phys 1996; 18: 12-17.
[6] Phinyomark A, Quaine F, Charbonnier S, Serviere C, Tarpin-Bernard F, Laurillau Y. Feature extraction of the first difference of EMG time series for EMG pattern recognition. Comput Meth Prog Bio 2014; 117: 247-256.
[7] Al-Shrouf L, Saadawia MS, S¨offker D. Improved process monitoring and supervision based on a reliable multi-stage feature-based pattern recognition technique. Inform Sciences 2014; 259: 282-294.
[8] Shao Y, Lunetta RS. Comparison of support vector machine, neural network, and CART algorithms for the landcover classification using limited training data points. Isprs J Photogramm 2012; 70: 78-87.
[9] Chen X, Wang ZJ. Pattern recognition of number gestures based on a wireless surface EMG system. Biomed Signal Proces 2013; 8: 184-192.
[10] Khushaba RN, Takruri M, Miro JV, Kodagoda S. Towards limb position invariant myoelectric pattern recognition using time-dependent spectral features. Neural Networks 2014; 55: 42-58.
[11] Wen C, Zhang J, Rebelo A, Cheng F. A Directed Acyclic Graph-Large Margin Distribution Machine Model for Music Symbol Classification. Plos One 2016; 11: e0149688.
[12] Hu X, Li W, Kan J. Gesture control technology based on surface electromyography for automatic pruning machine. Journal of Beijing Forestry University 2017; 117-124 (article in Chinese with an abstract in English).
[13] Hudgins B, Parker P, Scott RN. A new strategy for multifunction myoelectric control. IEEE T Bio-Med Eng 1993; 40: 82-94.
[14] Zardoshti-Kermani M, Wheeler BC, Badie K, Hashemi RM. EMG Feature evaluation for movement control of upper extremity prostheses. IEEE T Rehabil Eng 1996; 3: 324-333.
[15] Graupe D, Cline WK. Functional separation of EMG signals via ARMA identification methods for rrosthesis control purposes. IEEE T Syst Man Cyb 1975; SMC-5: 252-259.
[16] Stulen FB, Deluca CJ. Frequency parameters of the myoelectric signal as a measure of muscle conduction velocity. IEEE T Bio-Med Eng 1981; 28: 515-523.
[17] Oskoei MA, Hu H. Support vector machine-based classification scheme for myoelectric control applied to upper limb. IEEE T Bio-Med Eng 2008; 55: 1956-1965.
[18] Rønager J, Christensen H, Fuglsang-Frederiksen A. Power spectrum analysis of the EMG pattern in normal and diseased muscles. J Neurol Sci 1989; 94: 283-294.
[19] Potluri C, Anugolu M, Schoen MP, Subbaram ND, Urfer A, Chiu S. Hybrid fusion of linear, non-linear and spectral models for the dynamic modeling of sEMG and skeletal muscle force: an application to upper extremity amputation. Comput Biol Med 2013; 43: 1815-1826.
[20] Klema VC, Laub AJ. The singular value decomposition: its computation and some applications. IEEE T Automat Contr 1980; 25: 164-176.
[21] Barmpakos D, Strimpakos N, Karkanis SA, Pattichis C. Towards a versatile surface electromyography classification system. In: Proceedings of XIV Mediterranean Conference on Medical and Biological Engineering and Computing; 2016; Paphos, Cyprus: Springer. pp. 33-36.
[22] Jiang MW, Wang RC, Wang JZ, Jin DW. A method of recognizing finger motion using wavelet transform of surface EMG signal. IEEE Eng Med Biol Soc 2005; 3: 2672-2674.
[23] Lucas MF, Gaufriau A, Pascual S, Doncarli C, Farina D. Multi-channel surface EMG classification using support vector machines and signal-based wavelet optimization. Biomed Signal Proces 2008; 3: 169-174.
[24] Chu JU, Moon I, Lee YJ, Kim SK. A supervised feature-projection-based real-time EMG pattern recognition for multifunction myoelectric hand control. IEEE-ASME T Mech 2007; 12: 282-290.
[25] Rong Y, Hao D, Han X, Zhang Y, Zhang J, Zeng Y. Classification of surface EMGs using wavelet packet energy analysis and a genetic algorithm-based support vector machine. Neurophysiol 2013; 45: 39-48.
[26] Kiatpanichagij K, Afzulpurkar N. Use of supervised discretization with PCA in wavelet packet transformation-based surface electromyogram classification. Biomed Signal Proces 2009; 4: 127-138.
[27] Rosso OA, Blanco S, Yordanova J, Kolev V, Figliola A, Sch¨urmann M, Ba¸sar E. Wavelet entropy: a new tool for analysis of short duration brain electrical signals. J Neurosci Meth 2001; 105: 65-75.
[28] Zhang X, Zhou P. Sample entropy analysis of surface EMG for improved muscle activity onset detection against spurious background spikes. J Electromyogr Kines 2012; 22: 901-907.
[29] Guo Y, Naik GR, Huang S, Abraham A, Nguyen HT. Nonlinear multiscale Maximal Lyapunov Exponent for accurate myoelectric signal classification. Appl Soft Comput 2015; 36: 633-640.
[30] Zhou ZH. Machine Learning. Beijing, China: Tsinghua University Press, 2016.
[31] Platt JC, Cristianini N, Shawe-Taylor J. Large margin DAGs for multiclass classification. In: IEEE Conference on Neural Information Processing Systems; 2000; Cambridge, MA, USA: MIT Press. pp. 547-553.
[32] Gonzalez-Abril L, Velasco F, Angulo C, Ortega JA. A study on output normalization in multiclass SVMs. Pattern Recogn Lett 2013; 34: 344-348.
[33] Li W, Xing X, Fu L, Yu Z. Application of improved grid search algorithm on SVM for classification of tumor gene. International Journal of Multimedia and Ubiquitous Engineering 2014; 9: 181-188.
[34] Takahashi F, Abe S. Decision-tree-based multiclass support vector machines. In: 9th International Conference on Neural Information Processing; 2002; Singapore. pp. 1418-1422.
[35] Wang Y, Chen HH, Shen Y. Multi-class support vector machine based on directed acyclic graph. Electric Machines and Control 2011; 85-89 (article in Chinese with an abstract in English).
[36] Chen X, Zhu X, Zhang D. A discriminant bispectrum feature for surface electromyogram signal classification. Med Eng Phys 2010; 126-135.
[37] Liu J. Adaptive myoelectric pattern recognition toward improved multifunctional prosthesis control. Med Eng Phys 2015; 424-430.