Electromyogram (EMG) amplitude estimation and joint torque model performance

The relationship between surface EMG and torque about a joint has been the focus of many studies. Some of these studies however, have utilized conventional processors to obtain the EMG amplitudes. Recent advances (multiple channel combination and whitening) have demonstrated significant improvement in EMG amplitude estimation accuracy. The purpose of this study was to investigate the influence of these EMG amplitude advances into EMG-torque estimation. EMG from biceps/triceps muscles and torque about the elbow joint were collected from fifteen subjects producing constant-posture, nonfatiguing, force-varying contractions. EMG amplitudes were obtained using processors with and without the advances, and then they were related to torque using a linear FIR model. Results demonstrated that both whitening and multiple-channel combination reduced EMG-torque errors and their combination provided an additive benefit Specifically, the EMG-torque prediction errors were reduced to an average of 8% of maximum voluntary contraction of flexion (MVC/sub F/) when incorporating a four-channel, whitened processor.