Roll-Over Shape-Based Design of Novel Biomimetic Ankle-Foot Prosthesis

Introduction: Past research suggests that able-bodied individuals actively control their gait kinematics to produce one invariant roll-over shape (ROS). The ROS is a spatial representation of the center of pressure in a shank-based coordinate system. Because individuals with an amputation do not have adaptive control over their center of pressure during gait, the ankle-foot prosthesis determines the resultant ROS. This study aimed to design the mechanical properties of ankle-foot prostheses to achieve specific ROSs during gait. Methods: The authors developed a mathematical model that, when used in combination with ground reaction force and shank angle experimental data, predicts the ROS created by the novel Compliant and Articulating Prosthetic Ankle (CAPA) foot. The CAPA consists of four articulating components connected by torsion springs. Able-bodied individuals walked wearing a prosthetic simulator with the solid ankle cushioned heel foot, a dynamic response foot (the Renegade AT), and versions of the CAPA with distinctly different predicted ROSs. The experimental ROS of each trial was presented, and quantitative ROS characteristics were evaluated for statistical significance. Predicted and experimental shapes for each version of the CAPA were compared by computing the distance between the shapes and ROS characteristics. Finally, we evaluated whether our method predicts the same ROS regardless of the experimental data used for the prediction. Results: Changes to the mechanical properties of the CAPA resulted in statistically significant differences in ROS characteristics, notably forward position, forward length, and orientation (P < 0.05). The mean distance between points on the predicted and experimental ROSs is 1.54% of the length of the CAPA. We demonstrated that experimental ROSs can be predicted using gait data from a different version of the CAPA and a different individual. Conclusions: This study presents and experimentally validates a method to predict the ROSs created by eight versions of the CAPA during gait. The method enables prosthetists and researchers to design ankle-foot prostheses based on the ROS created during gait.