The influence of stimulus shape and movement on noise tagging based BCI

Abstract

Brain-Computer Interfaces (BCI) are a means to communicate to a digital device without the use of the peripheral nervous system. Noise tagging is a promising and relatively new technique that could provide high speed communication, but so far it has mainly been tested using common static squares. In this paper, I have explored the effects of stimulus shape and movement on the performance of a BCI. Animated figures could be used as targets in BCI controlled gaming, and be used to integrate BCI into the environment in an intuitive way. There was no significant effect found of either shape or movement on the performance, and the success of cross-condition classification opens up the possibility of using targets of different shapes with different animations at the same time without having to train different classifiers for each target type.