Abstract—Biomedical signals are commonly used as a convenient solution of Human Computer Interface (HCI) for the disabled persons. Myoelectric control system is the fundamental component of modern prostheses, which uses the myoelectric signals from an individual’s muscles to control the prosthesis movements. For this purpose, surface electromyogram (SEMG) data collected from thirty participants using eight electrodes located on the human forearm is used. Various feature sets were extracted and projected in a manner that ensured maximum separation between different movements of hand and then fed to the four different classifiers. We have used Sparse Principal component analysis as feature projection which very profoundly discriminated the feature sets. The second contribution is the use of a majority voting algorithm post processing approach to maximize the probability of correct classification of the myoelectric data for different movements of forearms. Practical results and ANOVA tests proved the feasibility of the proposed approach with an average classification accuracy > 98% for different subjects forearm movements. The focus of this work is to optimize the configuration of the classification scheme. The SVM ensemble based limb motion classification system demonstrates exceptional classification accuracy and results in a robust method of motion classification with low computational load.

Index Terms—Discriminant Locality Preserving Projections (DLPP), myoelectric control, Myoelectric Signal (MES), pattern classification, prosthesis, Sparse Principal Component Analysis (SPCA)

Cite: Anjana Goen and D. C. Tiwari, "Classification of the Myoelectric Signals of Movement of Forearms for Prosthesis Control," Journal of Medical and Bioengineering, Vol. 5, No. 2, pp. 76-84, April 2016. Doi: 10.18178/jomb.5.2.76-84