A hill-based EMG-driven model to estimate elbow torque during flexion and extention

Modeling of muscle force for a particular motion consists of wide range of approaches including Hill-based models. In this regard, An EMG-driven Hill-based model of the elbow during flexion-extension movement was developed using MATLAB software. Musculoskeletal model was composed of the linkage dynamics of the forearm and hand, EMG activation and musculoskeletal geometry. Six men and women participated in the current study, performing fully elbow flexion and extension for no and maximum load. Surface EMG electrodes were placed to detect the activity of muscles in order to be used in the musculoskeletal model. Besides that, capturing motion via cameras and analyzing it were performed in order to gain kinematic data. Model parameters of three flexors and one extensor were estimated by minimizing the sum of squared differences between the measured and estimated torques. Fmincon in the optimization Toolbox of MATLAB software was used to find the best possible and constricted parameters for model. Maximum force, tendon slack length and a subject-specific parameter, i.e. d, were the parameters which were calculated for all the muscles. Consequently, Reduction of RMS showed that considering a subject-specific parameter would help to have a better estimation. Furthermore, influence of parameter d for maximum load had a greater enhancement in the model estimation. Model couldn't estimate both flexion and extension parts in one movement for a particular subject. It should be noticed that, increasing in activity of Triceps at the middle of the movement, made us limit the whole action to only flexion. Finally, we concluded that flexion parameters couldn't predict the extension and would alter in according to variability of parameters.