Semiasynchronous BCI Using Wearable Two-Channel EEG

Motor imagery (MI)-based electroencephalogram can be modulated voluntarily by users through feedback provided by a brain-computer interface (BCI). In MI-based BCIs, feedback design plays an important role in MI training, especially in BCIs that use a limited number of electrodes. This paper presents a semiasynchronous MI-based BCI system that uses discrete and continuous feedback together, in order to improve practicability and training efficiency. Results show that the proposed method can reduce command generation time to 3.77 s with the accuracy of 77%, using our proposed continuous feedback with a two-channel MI-based BCI system. Our proposed system is noninvasive and wearable, featuring improvement in the practical utility, and operational convenience of a BCI system.