Human Walk Modeled by PCPG to Control a Lower Limb Neuroprosthesis by High-Level Commands

Current active leg prostheses do not integrate the most recent advances in Brain-Computer Interfaces (BCI) and bipedal robotics. Moreover, their actuators are seldom driven by the subject’s intention. This paper aims at showing a summary of our current results in the field of human gait rehabilitation. In a first prototype, the main focus was on people suffering from foot drop problems, i.e. people who are unable to lift their feet. However, current work is focusing on a full active ankle orthosis. The approach is threefold: a BCI system, a gait model and an orthosis. Thanks to the BCI system, patients are able to generate high-level commands. Typically, a command could represent a speed modification. Then, a gait model based on a programmable central pattern generator is used to generate the adequate kinematics. Finally, the orthosis is tracking this kinematics when the foot is in the air, whereas, the orthosis is mimicking a spring when the foot is on the ground.