A wireless, wearable Brain-Computer Interface for neurorehabilitation at home; A feasibility study

Despite Brain-Computer Interfaces (BCI) showing potential for motor rehabilitation following stroke, clinical adoption is lacking. BCI featuring unsupervised training at home may facilitate the clinical use of BCI. This study tested the feasibility of a wearable, 16 dry electrode wireless electroencephalography (EEG) system for BCI training with healthy participants. Twelve participants trained for 6 days using an EEG BCI at home, receiving remote instructions. A Graz style two-state motor imagery training protocol was used, requiring two different mental strategies. The BCI used common spatial patterns and linear discriminant analysis to classify mental states and provide feedback. The primary outcome was online classification accuracy of the mental state during BCI training, and secondary outcomes were the stability and distinctiveness of the EEG signal during mental states as well as the changes in alpha (8-13Hz) and beta (14-30Hz) power due to training. Six participants achieved control over the BCI. The distinctiveness of the EEG signal between the states increased over time and was associated with success. Stability increased over time but only in one mental state. Successful control was characterized by activity changes in alpha and beta power. Participants can achieve voluntary control over a BCI within 6 days using a wireless, wearable dry EEG system at home with minimal expert instruction or supervision.