Predicting the effects of walker height and weight support on assisted gait using physics-based predictive simulations

Walker-assisted gait is widely used in clinical rehabilitation for individuals with muscle weakness and balance impairments. This study presents a predictive simulation framework that integrates a 3D full-body musculoskeletal model driven by muscle torque generators within an optimal control problem. We calibrated the muscle torque generators model using experimental isometric and isokinetic data from a healthy participant, obtained from a BIODEX dynamometer equipment. To assess the predictive capability of the framework, we evaluated the effects of walker height and percentage of body weight support on walker-assisted gait patterns, by running nine predictive simulations across varying walker configurations. Results showed that effects of walker height were well predicted, while effects of weight support were only partially predicted. Moreover, we explored if the model could predict assisted gait in the presence of muscle weakness, specifically at the ankle muscles. Results showed some compensatory movements with increased weakness, although they did not exhibit the expected impairment in propulsion. This study provides an important step toward optimizing walker-assisted gait through simulation-based design and personalization.