A brain-computer interface based on abstract visual and auditory imagery: Evidence for an effect of artistic training

Various kinds of mental imagery have been employed in controlling a brain-computer interface (BCI). BCIs based on mental imagery are typically designed for certain kinds of mental imagery, e.g., motor imagery, which have known neurophysiological correlates. This is a sensible approach because it is much simpler to extract relevant features for classifying brain signals if the expected neurophysiological correlates are known beforehand. However, there is significant variance across individuals in the ability to control different neurophysiological signals, and insufficient empirical data is available in order to determine whether different individuals have better BCI performance with different types of mental imagery. Moreover, there is growing interest in the use of new kinds of mental imagery which might be more suitable for different kinds of applications, including in the arts.