Effects of Illumination and Noise on the Performance of a P300 Brain-Computer Interface for Assistive Vehicles

Disabled people often have difficulties in conveying their intentions to assistive vehicles, which can transport them to desired destinations, and a P300 brain-computer interface (BCI) system may help them use intelligent assistive vehicles by selecting a desired destination from predefined ones. However, real world driving often exposes the P300 BCI system to various illumination and noise environments and their effects on the performance of the system remain a question. The goal of this paper was to investigate this question by examining the effects of three levels of illumination and two levels of background noise on the accuracy of the system and user preference rating. Experimental results from twelve participants show that illumination does not have a significant effect on accuracy, although users prefer the low illumination level to others. However, noise significantly affects the system accuracy and user preference rating, with lower noise leading to higher accuracy and user preference rating. The study can not only provide new insights into the application of the P300 BCI for assistive vehicles in practice, but also facilitate the research on the effects of environmental factors on the performance of a P300 BCI system.