A Powered Ankle-Foot Prosthesis with a Neuromuscular Based Control Algorithm can Successfully Mimic Human Walking.
Hessel AL , Petak J, LeMoyne R, Tester J, Nishikawa KC
In the U.S., almost 40,000 people experience limb loss each year. Over the past decade, robotic advancements have allowed researchers to develop prosthetic devices that not only serve an aesthetic purpose, but also supply mechanical energy to the body to mimic healthy human functions. For a person with a trans-tibial amputation (PTTA), the iWalk BiOM is a commercially available powered ankle-foot prosthesis. The stock BiOM control algorithms use equations, not based upon muscle theory, to control the ankle torque output. In our research, we have developed a muscle model based on the winding filament hypothesis (WFH), which adds to the cross bridge theory and can explain the phenomena of force enhancement, force depression, and eccentric negative work. We developed a neuromuscular model control algorithm based on the WFH and incorporated it into the BiOM operating system to test whether human-style walking is attainable within a muscle-based control algorithm. We measured the metabolic cost of transport and walking gait characteristics of PTTA wearing a passive prosthesis, the BiOM with the original stock controller, and the BiOM with our WFH based controller. Our preliminary data indicate that the WFH-based BiOM is capable of producing a human-like ankle torque profile during walking that is lost in PTTA with a passive prosthesis, similar to the capability of the commercial BiOM. Further, while metabolic rates significantly decreased with the commercial BiOM, results were variable with the WFH BiOM, which we believe can be remedied with the next evolution of the WFH controller. Our future work will include testing on uneven terrain, ramps and stairs. We hope that this work will inspire bioengineers to create next-generation powered prostheses using control algorithms inspired by an understanding of muscle biology.