PROJECT TITLE :

Influence of Joint Angle on EMG-Torque Model During Constant-Posture, Torque-Varying Contractions

ABSTRACT:

Relating the electromyogram (EMG) to joint torque is helpful in various application areas, as well as prosthesis management, ergonomics and clinical biomechanics. Restricted study has related EMG to torque across varied joint angles, notably when subjects performed force-varying contractions or when optimized modeling ways were utilized. We have a tendency to related the biceps-triceps surface EMG of twenty-two subjects to elbow torque at six joint angles (spanning 60 to a hundred thirty five ) during constant-posture, torque-varying contractions. Three nonlinear EMG -torque models, advanced EMG amplitude (EMG ) estimation processors (i.e., whitened, multiple-channel) and therefore the length of knowledge used to train models were investigated. When EMG-torque models were formed separately for each of the six distinct joint angles, a minimum “gold commonplace” error of MVC resulted (i.e., error relative to most voluntary contraction at 90 flexion). This model structure, but, did not directly facilitate interpolation across angles. The best model that did therefore achieved a statistically equivalent error of MVC . Results demonstrated that advanced EMG processors cause improved joint torque estimation as do longer model training durations.