Recently, the relationship between muscles' electrical activity and body movements has been considered in many medical applications. In these applications, uses of non-expensive and portable of electromyography (EMG) electrodes have advantageous compared to the use of force sensors and cameras which are often very expensive and require massive structures. In this paper, we evaluate the ability of the Fast Orthogonal Search (FOS) methodology to predict jaw motion using Electromyography (EMG) signals recorded from two masticatory muscles, namely masseter and temporalis. Results show the efficiency of FOS in predicting the kinematic parameters (position and orientation) based on EMG signals. Additionally, the proposed model can be utilized to control masticatory robots employing recorded EMG signals.

Keywords