Brain-computer interfacing for flexion and extension of bio-inspired robot fingers

Abstract

Brain-computer interface (BCI) technology is a topic of study with assistive robot systems that are expanding significantly nowadays. Several advancements have been made in the BCI sector to aid disabled people. However, most studies have used more electrodes, and most design structures are different from the anatomy of the human hand. To control robot fingers like real human hands with fewer electrodes is yet to be developed. This research aims to investigate the controllability of robot fingers using motor imagery. The proposed electroencephalogram (EEG) acquisition system comprises eight EEG electrodes attached to the primary motor cortex region of the human brain. The real human hand's anatomical behavior aided in the robot's development. Initially, the performance of the robot finger model was evaluated on the computer simulation. Finally, a robot model was developed, and flexion and extension movements were examined. According to the experiment's findings, finger flexion and extension control with eight EEG electrodes showed promising results with an accuracy of 90.0±1.43% and a precision of 0.89. Furthermore, we observed that as people age, the accuracy of robot control decreases.