Using Motor Imagery to Control Brain-Computer Interfaces for Communication

摘要

Brain-computer interfaces (BCI) as assistive devices are designed to provide access to communication, navigation, locomotion and environmental interaction to individuals with severe motor impairment. In the present paper, we discuss two approaches to communication using a non-invasive BCI via recording of neurological activity related to motor imagery. The first approach uses modulations of the sensorimotor rhythm related to limb movement imagery to continuously modify the output of an artificial speech synthesizer. The second approach detects event-related changes to neurological activity during single trial motor imagery attempts to control a commercial augmentative and alternative communication device. These two approaches represent two extremes for BCI-based communication ranging from simple, "button-click" operation of a speech generating communication device to continuous modulation of an acoustic output speech synthesizer. The goal of developing along a continuum is to facilitate adoption and use of communication BCIs to a heterogeneous target user population.