Electromyography-assisted optimization (EAO) is limited in that all Electromyography (EMG) information of main model muscles is needed, although it can predict proper muscle forces. This report proposes how to use muscle forces predicted by classical optimization (OPT) which does not use EMG information as a substitute for EMG activity, and investigates validity of lower muscle forces predicted by this method during gait. Five male subjects walked freely five times. The model we used had three joints and eleven muscles of the lower leg in the sagittal plane. EMG electrodes were placed on eight muscles. Iliopsoas muscle force predicted by OPT was used to substitute for iliopsoas muscle activity. Pearson's correlation coefficients between predicted muscle forces and EMG were used as indices of validity. Results showed that muscle forces predicted by EAO correlated well with EMG (
r = 0.74∼0.84). It seems that iliopsoas muscle force predicted by EAO was affected by force of other muscles which our musculoskeletal model did not include. This study suggests, however, that iliopsoas muscle force predicted by EAO was suitable, considering that the eleven muscle forces satisfied joint moment equation and that the muscle forces except iliopsoas muscle correlated with EMG well.
View full abstract