Synergy-based feedforward with minimal feedback control predicts walking over multiple cycles.

Neural feedback is important for the control of movement, and multiple neurological disorders (e.g., stroke, cerebral palsy, Parkinson's disease, incomplete spinal cord injury) are characterized by altered neural feedback. Researchers have created numerous computational neuromusculoskeletal models controlled by simulated neural feedback mechanisms, but these models rarely represent actual human subjects and thus have not found practical clinical application. As a step toward designing patient-specific treatments for individuals with neurological disorders, this study used the Neuromusculoskeletal Modeling Pipeline to develop and evaluate a novel synergy-based feedforward (FF)+feedback (FB) control model using a personalized three-dimensional neuromusculoskeletal walking model of an actual human subject post-stroke. Experimental walking data collected from the subject were used to create the subject's personalized walking model. Then for five calibration walking cycles, personalized synergy-based FF + FB control models were created. First, the personalized model was used to estimate lower body muscle activations, consistent with the subject's electromyographic, joint motion, and joint moment data. Second, five synergy activations per leg with associated synergy vectors were calculated that closely reconstructed the subject's muscle activations and joint moments simultaneously. Third, nominal FF synergy activation controls were calculated by averaging the synergy activations for each leg. Fourth, the nominal FF synergy controls were scaled by 0%, 25%, 50%, 75%, 100%, and 125%, and the gap in reproducing the subject's muscle activations was filled by fitting FB synergy activation controls as a function of joint positions, velocities, and moments as surrogates for muscle lengths, muscle velocities, and tendon forces. Next for three testing walking cycles, six synergy-based FF + FB models were used to control the subject's personalized walking model in predictive simulations. The 100% FF model (which still had minimal FB) reproduced the testing walking cycles the most closely, and only the 75%, 100%, and 125% FF models predicted near-periodic walking motions using initial conditions consistent with experimental values. The 0%, 25%, and 50% FF models could generate near-periodic walking motions only when the initial conditions were allowed to diverge substantially from experimental values. Our findings suggest that predictive simulations of walking may require substantial feedforward control when modeling an actual human subject.