چكيده به لاتين
Abstract
Brain computer interfaces (BCI) are systems that enable communication between the brain and machines without using nerve channels and muscular tissues. These systems use EEGs, electrical signals produced in neurons, to manage processes of other systems. Proper conditions should be provided so that with an appropriate stimulation, an electrical signal containing a specific message is produced and recorded. After decoding theses messages, a due operational code for the target system is produced. One way of acquiring signals is to stimulate the visual cortex and record the resulting EEG. In order to excite this part of the brain, one method would be to display flashing colored lights and colored geometric shapes on a monitor. The contrast between the stimulus and screen background plays an important role in the transmission of information, while performing the stimulation on a computer screen. In order to analyze the role of contrast in this thesis, the signals produced by the visual system, excited by colored squares with different colors and flashing frequency in different steps, are recorded. These signals are called steady-state visually evoked potentials (SSVEPs). Then, using multi-way canonical correlation analysis (MCCA), frequency characteristic of the recorded signals are extracted. Results of this study show that red and gray color have the best quality of information transmission in different frequency levels from the reference frequency of the stimuli to the recorded SSVEP. Furthermore, it is noted that after prolonged staring at a, first red and then gray, display participants become more exhausted, which results in a significant decrease in functionalities after numerous tests. Results of this study might be used for choosing the frequency limits, color and contrast for the applications of BCI.
Keywords: Brain-computer interface (BCI), Effect of fatigue, Color contrast effects in classification, Steady-state visual evoked potential (SSVEP), canonical correlation analysis (CCA), Multi-way canonical correlation analysis (MCCA), the Chalder fatigue scale (CFS).
