An EOG-based wheelchair robotic arm system for assisting patients with severe spinal cord injuries.

OBJECTIVE: In this study, we combine a wheelchair and an intelligent robotic arm based on electrooculogram (EOG) signal to help patients with spinal cord injuries (SCIs) accomplish a self-drinking task. The main challenge is to accurately control the wheelchair to ensure that the randomly located object is within a limited reachable space of the robotic arm (length: 0.8 m; width: 0.4 m; height: 0.6 m), which requires decimeter-level precision and is still unproved for EOG-based systems as well as EEG-based systems.

APPROACH: A novel high-precision EOG-based human machine interface (HMI) is proposed, which can effectively translate two kinds of eye movements (i.e., blinking and raising eyebrows) into various commands. For the wheelchair, the positional precision can reach decimeter-level and the minimal steering angle is 5°. For the intelligent robotic arm, a shared control is implemented based on the EOG-based HMI, two cameras and the arm's own intelligence.

MAIN RESULTS: After a short training, five healthy subjects and five paralyzed patients with severe SCIs successfully completed three experiments. For healthy subjects/patients with SCIs, the system achieved an average accuracy of 99.3%/97.3%, an average response time of 1.91 s/2.02 s per command and an average stop-response time of 1.30 s/1.36 s, with a 0 false operation rate.

SIGNIFICANCE: The EOG-based HMI can provide sufficient control precisions to integrate a wheelchair and a robotic arm into a system, which can help patients with SCIs to accomplish the self-drinking task. (ChiCTR1800019764).