Real-time monitoring of military sentinel sleepiness using a novel SSVEP-based BCI system

Brain Computer Interface (BCI) creates a new communication channel for people who have motor disabilities. In recent years, SSVEP-based BCI has been focused widely, due to its advantages over other types of BCI. Brain sleepiness during military sentry is an indispensable and prevalent occurrence in military areas. In this study, a novel high-speed SSVEP-based BCI system for military sentinel monitoring is proposed. In order to generate the SSVEPs in brain activity of a sentinel, an experimental setup containing two sets of LEDs with constant twinkling frequency is proposed. In addition, the effects of two different twinkling frequencies of 15 Hz and 20 Hz, three segmentation lengths (sweeps), two FFT-based feature extraction methods, and three classifiers on the system accuracy is investigated. In addition, in order to make a compromise between the accuracy and the computation speed, Information Transfer Rate (ITR) was calculated. The experimental results shows that 20 Hz stimulant is more preferable than 15 Hz one with the best sweep length of 1 second using the Max classifier with ITR value of 19.3 bit/min.