EMG processing for classification of hand gestures and regression of wrist torque

This paper investigates the use of myoelectric signals to identify hand gesture as well as predict wrist torque in healthy volunteers. Surface electromyography (sEMG) signals from four forearm muscles were recorded while the volunteers were exerting wrist torque on a custom-made rig. Multi class support vector machines (SVM) were used for classification and regression. The obtained experimental results proved that the proposed sEMG processing scheme enabled classifying six different hand gestures with 95.51% accuracy and estimate wrist torque intensity for each of those classes with a normalized root mean square error (NRMSE) of 0.057 for regression.