Making BCIs colorful: the influence of different chromatic combinations on the performance of the noise-tagging BCI

dc.contributor.advisorDesain, P.W.M.
dc.contributor.advisorThielen, J.
dc.contributor.authorHeer, R. de
dc.description.abstractA brain computer interface (BCI) allows a user to operate a computer using brain activity only (i.e., no muscles). Objects that represent functions are placed on a screen, and each object flashes, usually alternating between black and white in a unique and random way. Such a unique and random flash pattern is called a noise-tag. When the user focuses on a specific object, a corresponding signal is evoked in the brain of the user because of the unique noise-tag. This technique has only been used with black and white or with blue and green alternating colors. In the current research, the usage of chromatic combinations in the noise-tagging BCI was studied. We investigated the performance of the BCI when using different chromatic combinations, the pleasantness of those chromatic combinations, the correlation between the pleasantness of the chromatic combination and the performance of the BCI and the differences the brain responses evoked by the stimuli when using different chromatic combinations. All of the chromatic combinations used proved feasible by not performing significantly less than traditional black and white. The participants of the experiment liked all chromatic combinations equally. The normalized likeness ratings did not significantly correlate with the performance of the BCI in terms of accuracy. Visual inspection showed that the transient responses were clear for the chromatic combination black and white, but less clear in other conditions. Visual inspection also showed that the spatial filters generated by Canonical Correlation Analysis (CCA) differed mostly per participant, and not per condition. This information has the potential to make communication easier for paralyzed patients and offers new options for colorful games.en_US
dc.thesis.facultyFaculteit der Sociale Wetenschappenen_US
dc.thesis.specialisationBachelor Artificial Intelligenceen_US
dc.thesis.studyprogrammeArtificial Intelligenceen_US
dc.titleMaking BCIs colorful: the influence of different chromatic combinations on the performance of the noise-tagging BCIen_US
Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Heer, de R._BSc_Thesis_2017.pdf
898.31 KB
Adobe Portable Document Format
Thesis text