Investigation of Functional Brain Connectivity by Electroencephalogram Signals using Data Mining Technique

Abstract

Human brain is one of the most complex and the most vital human body organs with different parts of it being interconnected even if these parts are anatomically separate. It is essential to consider the brain function as an integrated system in order to get insight into the complex structure and function of the cerebral network as a key concept in neuroscience. The patterns obtained from the function of different brain areas and their processing techniques yield a complete set of information about the available relationship between brain areas and make it possible to analyze the function of the brain system correctly using new modeling tools. In this regard, in the present study, brain function was analyzed in 6 subjects using the picture-naming test (148 images of animals D249 and tools D250) and EEG signal recording method through 256 channels. In this two-session test (with 74 stimuli), 12 signals related to the brain function of these subjects were recorded and analyzed in delta, theta, beta, alpha and non-filter state. Furthermore, the pattern of relationship between the channels and the brain communication network in different areas was calculated and elicited in two modes of D249 and D250 (the test stimuli included animals and tools pictures) using the available tools and calculation methods (correlation coefficient, t-test and association rule mining). The obtained results showed that the frontal and temporal areas had the highest activity in comparison with other areas. The brain behavioral patterns in these subjects were very similar in the three bands of theta, beta and alpha.