OpenSim Model for Biomechanical Analysis with the Open-Source Bionic Leg

Lower-limb robotic prostheses present a potential solution for enhancing the mobility of individuals with amputation. However, comparing results amongst different research groups is a challenge as there are many differences between research devices. A recent effort from the University of Michigan created and established an open-source design for an active knee and ankle prosthesis with embedded sensors, including a shank inertial measurement unit (IMU), a 6 DOF loadcell, and joint encoders. As the usage of this design grows, it is expected that the field will require additional analyses of the locomotion biomechanics when using the open-source leg beyond what the embedded sensors can provide. In this study, we present, validate, and release a model for the software OpenSim, that serves as a solution for the full body analysis of the inverse kinematics on optical motion capture data. Our model can adjust to different pylon heights, incorporate mass & inertial properties, and provide visualizations of the prosthetic leg. We validated the model kinematically by asking four individuals with amputation to walk on a force-instrumented treadmill. Our model shows an accurate match with the encoders embedded in the leg with an RMSE of 2.34 deg for the knee and an RMSE of 2.54 deg for the ankle. This work should help facilitate the use of motion capture with the open-source leg for the bilateral analysis of the locomotion utilizing the open-source prosthetic device.