Modeling neuromuscular effects of ankle foot orthoses (AFOs) in computer simulations of gait

Ankle foot orthoses (AFOs) provide immediate changes to gait kinematics and alter EMG-recorded muscle activity during gait; yet our understanding of neuromuscular adaptations while using AFOs remains incomplete. To address this, we have developed a tunable AFO model to predict torque from ankle angle and velocity and to identify plausible changes in muscle excitation and function in a walking simulation. Using a dynamometer in passive mode, we isolated flexion/extension torque of three polypropylene spring leaf AFOs at 5°/s, 30°/s and 120°/s, from which a model of AFO torque as a function of deformation angle, velocity and size was derived with predictive ability of R2 > 0.9. The model coefficients did not vary linearly with size, illustrating the need to test AFO deformation response individually. We applied the tuned models to simulations of normal healthy gait to isolate AFO-induced neuromuscular changes. Compared to the No-AFO condition, AFO results show decreased net tibialis anterior excitation. Our results also show that net soleus excitation is not diminished with an AFO although soleus-induced ankle accelerations were reduced. With a tunable AFO model applied to walking simulations, we can quantify the contributions of muscle and orthosis to net joint torque and predict changes in neuromuscular control during walking.