Ankle Prosthesis With Brain Computer Interface Commands Based on Electroencephalograph for Transtibial Amputees

Abstract

Amputation is the removal or severing of a body part, such as a finger, arm, or leg. Amputations are usually performed to treat a condition or disease. Lower limb amputations account for 85-90% of all amputations, and transtibial amputations are the most common type of amputation surgery. To live a normal life, transtibial amputees choose to use assistive devices such as crutches or passive prostheses. The use of passive equipment is still limited due to the user’s functional requirements which need to be made as close to natural as possible. Motivated by the above points, an active Ankle Prosthesis for transtibial amputation patients using EEG signals has been designed. The 2-channel C3 and C4 EEG signals are recorded using a self-designed EEG Instrumentation. Ankle prosthesis modified based on an open-source 3D design adapted to the subject of anthropometry. The 3D design is printed with a 3D printer made from PLA. The resulting Ankle Prosthesis has a total mass of 635 grams, an extension length of 19 cm, and a footing dimension of the prosthesis ($25 \times 10$) cm. The Ankle Prosthesis can move at an angle of 15 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sup> for Dorsiflexion and 65 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sup> for Plantar flexion. EEG signal features in the form of Event Related Synchronization (ERS) and Event Related Desynchronization (ERD) are extracted from EEG signals in real-time during Motor Execution and Motor Imagery tasking. In the experimental test with 2 normal subjects, the highest success was 33.33% in motor execution mode for subject 1 and 38.89% in motor imagery mode for subject 2. This shows that subject2 can command the Ankle Prosthesis better than subject 1. From the success data, the most detections during motor execution were obtained by subject 1 for left-hand movements with an average ERD/ERS value of −56.9727± 42.4894% on channel C4. As for the motor imagery, the most detection was obtained by subject 2 for right-hand movement with an average value of ERD/ERS point C3 of −55.7915± 25.75024%.