EEG-based Brain-Computer Interface System via Time-Locked Visual Attention for Assistive Device Control

This work proposes a brain-computer interface system using visual attention paradigm. We used count-down numbers for different periods of attention to command creation from EEG signals. Electrode positions and feature extraction algorithm were verified. By using four number and the proposed simple classification method can produce four commands. The results showed that EEG from the occipital region using an alpha-beta ratio achieved an average classification accuracy of 80. 56% and 62.06% for two and four commands, respectively. The proposed system can be used for severely physically disabled people for control and communication in the future. However, the proposed needs to improve for higher efficiency by session training and using complex algorithms and classifiers.