Abstract
Prosthetic gait differs considerably from the unimpaired gait. Studying alterations in the gait patterns could help to understand different adaptation mechanisms adopted by these populations. This study investigated the effects of induced stiff-knee gait (SKG) on prosthetic and healthy gait patterns and the capabilities of predictive simulation. Self-selected speed gait of two participants was measured: one healthy subject and one knee disarticulation subject using a variable-damping microprocessor controlled knee prosthesis. Both performed unperturbed gait and gait with restricted knee flexion. Experimental joint angles and moments were computed using OpenSim and muscle activity was measured using surface electromyography (EMG). The differences between the conditions were analyzed using statistical parametric mapping (SPM). Predictive models based on optimal control were created to represent the participants. Additionally, a hypothetical unimpaired predictive model with the same anthropometric characteristics as the amputee was created. Some patterns observed in the experimental prosthetic gait were predicted by the models, including increased knee flexion moment on the contralateral side caused by SKG in both participants, which was statistically significant according to SPM. With the exception of the rectus femoris muscle, we also found overall good agreement between measured EMG and predicted muscle activation. We predicted more alterations in activation of the hip flexors than other muscle groups due to the amputation and in the activation of the biceps femoris short head, quadratus femoris, and tibialis anterior due to SKG. In summary, we demonstrated that the method applied in this study could predict gait alterations due to amputation of the lower limb or due to imposed SKG.