Modifying Gait: A Strategy for Joint Reaction Forces Reduction in Below-Knee Amputees

Compared to healthy people, people with transtibial amputation (TTA) are more prone to arthritis in their intact legs, especially in hip and knee joints. The main reason is that people with TTA typically depend more on their intact legs for generating forward movements, resulting in high joint reaction forces in the intact leg. Therefore, a new gait pattern with lower joint loads is essential to people with TTA to prevent arthritis. In this paper, we use predictive simulation to generate a new gait pattern that can decrease the joint reaction forces for people with TTA. The reflex-based controller is used to control muscle excitation, which is fed into Hill muscle-based forward dynamics to generate muscle forces and thus gait movement based on the motion of the equation. To find the optimal parameters in the reflex-based controller, we then propose a cost function to minimise effort, ground reaction force and joint reaction force together by covariance matrix adaptation evolution strategy. Experimental results have indicated the new gait pattern generated by our approach can reduce the joint reaction forces for people with TTA, while the energy consumption doesn't have a significant change, which demonstrates the effectiveness of this method.