A discriminant bispectrum feature for surface electromyogram signal classification

• Published in: Medical engineering & physics Vol. 32; no. 2; pp. 126 - 135
• Main Authors: Chen, Xinpu, Zhu, Xiangyang, Zhang, Dingguo
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 01.03.2010; Elsevier
• Subjects: Biological and medical sciences; Classification; Discriminant Analysis; Discriminant bispectrum feature; Electrodiagnosis. Electric activity recording; Electromyography - methods; Electromyography - statistics & numerical data; Female; Fundamental and applied biological sciences. Psychology; Hand - physiology; Humans; Investigative techniques, diagnostic techniques (general aspects); Male; Medical sciences; Movement; Multivariate Analysis; Nervous system; Prosthetic control; Radiology; Regression Analysis; sEMG signal; Signal Processing, Computer-Assisted; Vertebrates: body movement. Posture. Locomotion. Flight. Swimming. Physical exercise. Rest. Sports; Wrist - physiology; Prosthetic control; sEMG signal; Discriminant bispectrum feature; Classification; Human; Prosthesis; Electrophysiology; Hand; Modeling; Spectral analysis; Wrist; Electrodiagnosis; Orthopedic surgery; Body movement; Motion study; Signal processing; Electromyography; Upper limb; Bispectrum; Mathematical model; Biomedical engineering
• ISSN: 1350-4533; 1873-4030
• DOI: 10.1016/j.medengphy.2009.10.016

Abstract This paper presents a discriminant bispectrum (DBS) feature extraction approach to surface electromyogram (sEMG) signal classification for prosthetic control. The proposed feature extraction method involves two steps: (1) the bispectrum matrix integration, and (2) the Fisher linear discriminant (FLD) projection. We compare DBS with other conventional features, such as autoregressive coefficients, root mean square, power spectral distribution and time domain statistics. First, the separability of the features is investigated by the visualization of feature distribution in the FLD subspace and quantitative measurement (Davies–Boulder clustering index). Then four linear and non-linear classifiers are used to evaluate the discriminative powers of the features in terms of classification accuracy (CA). The experimental results show that DBS has better performance than other features for identifying the motion patterns of sEMG signals, and the best CA result of DBS is 99.4%.