BCI Use and Its Relation to Adaptation in Cortical Networks

IEEE Transactions on Neural Systems and Rehabilitation Engineering | , Vol 25: pp. 1697-1704

Publication

Brain-computer interfaces (BCIs) carry great potential in the treatment of motor impairments. As a new motor output, BCIs interface with the native motor system, but acquisition of BCI proficiency requires a degree of learning to integrate this new function. In this review, we discuss how BCI designs often take advantage of the brain’s motor system infrastructure as sources of command signals. We highlight a growing body of literature examining how this approach leads to changes in activity across cortex, including beyond motor regions, as a result of learning the new skill of BCI control. We discuss the previous research identifying patterns of neural activity associated with BCI skill acquisition and use that closely resembles those associated with learning traditional native motor tasks. We then discuss recent work in animals probing changes in connectivity of the BCI control site, which were linked to BCI skill acquisition, and use this as a foundation for our original work in humans. We present our novel work showing changes in resting state connectivity across cortex following the BCI learning process. We find substantial, heterogeneous changes in connectivity across regions and frequencies, including interactions that do not involve the BCI control site. We conclude from our review and original work that BCI skill acquisition may potentially lead to significant changes in evoked and resting state connectivity across multiple cortical regions. We recommend that future studies of BCIs look beyond motor regions to fully describe the cortical networks involved and long-term adaptations resulting from BCI skill acquisition.