Real-time neuromusculoskeletal model for myoelectric control of ankle prosthesis

The functional impact of amputations presents significant challenges in daily living activities. These challenges—particularly maintaining balance-are escalated to significant patient safety risks due to severe falling injuries. Modern transtibial amputation procedures preserve much of the neuromuscular tissue involved in motor control of ankle joints, which can be utilized for myoelectric control of ankle prosthesis. In this study, a real-time neuromusculoskeletal(NMS) model of the lower limb with two degrees of freedom at knee and ankle joint was developed to receive EMG and limb kinematic signals as input and predict ankle motion as output during target following tasks and swing phase of gait. Our results indicated that the real-time NMS model predicts accurate ankle motion for various target speeds and positions. We also show the performance of the model in predicting ankle kinematics during the swing phase of motion and found mean peak correlation of 0.95 between predicted and measured motions across all subjects. The proposed modeling framework can be used for developing powered ankle prosthesis for transtibial amputees.