Developing a Brain-Computer Interface Based on Visual Imagery

  • Justin Kilmarx | The University of Texas at Austin

A brain-computer interface (BCI) is a technology to provide direct communication between the brain and an external device. In this project, we have utilized noninvasive electroencephalography (EEG) to record and decode neural activity during the observation and mental imagery of visual stimuli. Our platform has demonstrated successful discrimination between face and scene image categories during visual observation and imagery periods. Additionally, we have shown above chance decoding accuracy during real-time prediction of face and scene imagery and resting state. This platform provides further insight into the use of visual imagery, a protocol that has not yet been much tested for BCI applications. Unlocking visual imagery as a BCI control strategy can provide a more intuitive association between the mental task and intended action compared to more popular protocols such as motor imagery.

Watch Next