Document Type : Full Research Paper


1 Assistant Professor, Brain Behavioral Science Group, University of Birmingham

2 Professor, School of Electrical and Computer Engineering Tarbiat Modares University

3 Associate Professor, Biophysics Group, School of Medicine, Tarbiat Modares University



A novel approach to surface electromyogram (sEMG) signal classification using its higher order statistics (HOS) is presented in this study. As the probability density function of the sEMG during isometric contraction in some cases is very close to the Gaussian distribution, it is frequently assumed to be Gaussian. As this assumption is not valid when the force is small, in this paper, we consider the non-Gaussian characteristics of the sEMG, and compute the second-, the third- and the fourth order statistics of the sEMG as its features. These features are used to classify four upper limb primitive motions, i.e., elbow flexion (EF), elbow extension (EE), forearm supination (FS), and forearm pronation (FP). We used the sequential forward selection (SFS) method to reduce the number of HOS features to a sufficient minimum while retaining their discriminatory information, and apply the Knearest neighbor method for classification. Our approach is robust against statistical variations in noise, and does not require additional computations compared to existing methods for providing high rates of correct classification of the sEMG, which makes it useful in devising real-time sEMG controlled prostheses.


Main Subjects

[1]     De Luca C.J.; Physiology and mathematics of myoelectric signal; IEEE Transactions on Biomedical Engineering 1979; 26:313-325.
[2]     Saridis G.N., Gootee T.P.; EMG pattern analysis and classification for a prosthetic arm; IEEE Transactions on Biomedical Engineering 1982; 29:403-412.
[3]     Kermani M.Z., Wheeler B C, Badie K., Hashemi R.M.; EMG feature evaluation for movement control of upper extremity prostheses; IEEE Transactions on Rehabilitation Engineering 1995; 3(4):324-333.
[4]     Kelly M., Parker P.; Application of neural network to myoelectric signal analysis: a preliminary study; IEEE Trans. Biomedical Engineering 1990; 37(3):221-230.
[5]     Karlik B., Tokhi M. O., Alci M.; A fuzzy clustering neural network architecture for multifunctional upperlimb prosthesis; IEEE Transactions on Biomedical Engineering 2003; 50 (11):1255-1261.
[6]     Englehart K., Hudgins B., Parker P.A..; A waveletbased continuous classification scheme for multifunctional myoelectric control; IEEE TBME, 2001; 48:302-311.
[7]     Englehart K., Hudgins B, Parker P A, Stevenson M.; Classification scheme of the myoelectric signal using time-frequency based representation; Med. Eng. Phys. (Special Issue: Intel. Data Anal. Electromyogr. Electroneurogr.) 1999; 21:431-438.
[8]     Englehart K., Signal Representation for Classification of the Transient Myoelectric Signal, Ph.D. Dissertation, Univ. New Brunswick, Fredericton, NB, Canada, 1998.
[9]     Erfanian A., Chizeck H.J. , Hashemi R.M.; Chaotic activity during electrical stimulation of paralyzed muscle; 18th Annual IEEE/EMBS Conf.: Bridging Disciplines for Biomedicine. 1997; 4:1756 -1757.
[10] Roesler H., Statistical analysis and evaluation of myoelectric signals for proportional control, in: The Control of Upper Extremity Prostheses and Orthoses; Springfield, IL: C. C. Thomas; 1974:44-53.
[11] Hunter I.W., Kearney R.E., Jones L.A.; Estimation of the conduction velocity of muscle action potential using phase and impulse response function techniques; Med. Biol. Eng. Comput. 1987; 25:141-126.
[12] Bilodeau M., Cincera M., Arsenault A.B., Gravel D.; Normality and stationarity of EMG signals of elbow flexor muscles during ramp and step isometric contractions; J. Electromyogr. Kinesiol. 1997; 7: 87- 96.
[13] Clancy E.A., Hogan N.; Probability density of the surface electromyogram and its relation to amplitude detectors; IEEE Trans. BME., 1999; 46 (6):730-739.
[14] Lindstrom L., Magnusson R.; Interpretation of myoelectric power spectra: a model and its application; in Proc. of the IEEE 1977; 65:653-660.
[15] Plévin E, Zazula D; Decomposition of surface EMG signals using non-linear LMS optimization of higherorder cumulants; in Proc. of 15th IEEE CBMS 2002: 149-154, Slovenia.
[16] García G A, Nishitani R, Okuno R, Akazawa K; Independent component analysis as a pre-processing tool for decomposition of surface electrode-array electromyogram; in Proc. ICA 2003: 191-196, Nara, Japan.
[17] Mendel J.M.; Tutorial on higher-order statistics (spectra) in signal processing and system theory: theoretical results and some applications; in Proceedings of the IEEE 1991; 49 (30):278-305.
[18] Nazarpour K., Sharafat A.R., Firoozabadi S.M.P; A novel feature extraction scheme for myoelectric signals classification using higher order statistics; in Proceedings of the 2nd Int. IEEE/EMBS Conference on Neural Engineering NER2005: 293-296, Virginia, USA.
[19] Nazarpour K., Sharafat A.R., Firoozabadi S.M.P.; Negentropy analysis of electromyogram signal; in Proceedings of the IEEE Statistical Signal Processing, SSP 2005: 974-977, Bordeaux, France.
[20] Nazarpour K., Sharafat A.R., Firoozabadi S.M.P.; Surface EMG signal classification using a selective mix of higher order statistics; in Proc. IEEE/EMBS 27th EMBC 2005: 4208-4211, Shanghai, China.
[21] Dembele D., Favier G.; Recursive estimation of fourthorder cumulants with application to identification,” Signal Processing, 1998; 68:127-139.
[22] Theodoridis S., Koutroumbas K., Pattern Recognition, Academic Press, 1999.
[23] Hudgins B., Parker P.A., Scott R.; A new strategy for multifunction myoelectric control; IEEE Transactions on Biomedical Engineering, 1993; 40:82-94.