Early Evaluation of sEMG-Driven Muscle Modelling for Rehabilitation and Assistive Applications based on Wearable Devices

In this paper, a preliminary identification and evaluation activity about the modeling of the muscle behavior during isometric contraction is reported. The muscle model is based on a second order dynamic system taking as input the surface skin electromyographic signals measured by means of a pair of electrodes placed on the skin of a human subject and providing as output the force generated by the muscle, applied to the limbs through the tendon. Experimental tests have been performed to identify an internal non-linear fitting between the measured sEMG signal and the impulsive contractile element force, in the case of isometric contractions of the biceps brachii. The proposed model is therefore exploited to drive a wearable elbow support device to limit the activity of the user's muscle to a maximum desired threshold value. The experimental results show that the proposed model can be effectively used to provide a direct information about the force provided by the muscle and to modulate the support action provided by the wearable device.