The Research Of Construction And Analysis Of FMRI Function Brain Network In The Task State

Visual perception has a very important role in human brain function. Visual information is more than80%of the amount of information which human beings accept. How does visual system to process visual information and ultimately complete the identification of object. To solve the problem, this article raised the issue of color and shape features of recognition and binding. The research of color and shape features of recognition and binding has some theoretical significance and application value.In order to uncover the mysteries of the brain’s visual area, fMRI and complex network provide an effective means and theoretical basis. Graph theory is a very important method to study complex, therefore, in order to reveal the mechanism of visual network, this article need the help of graph theory to study the function of brain networks.In the paper, the study needed to acquire fMRI data, preprocess fMRI data, locate and analyze the activation of brain area. Pre-processing data was used to construct functional brainnet and partitioning algorithm was applied to functional brainnet to divide network. Finally, graph theory was used to analyze the difference of brainnet in the resting and task state. The main work is as follows:1. In accordance with the purpose of the study, experimental paradigm was designed; fMRI data was acquired in the resting and task state. Finally, pre-processing data was used to locate and analyze the activation of brain area. The analysis of experimental data show that whether performing simple task or complex task, the brain areas associated with visual information processing have obvious activation. But also we found that the more number of complex tasks and features, the visual areas are active; through the analysis of experimental data, we confirmed the two pathways of visual information processing, one of them distributes along the temporal and occipital lobe, another one distributes along parietal and occipital lobe.2. Firstly, the brain area was selected as network node. And then we need to calculate the correlation coefficient between brain areas. Through selecting threshold was to determine network side to construct functional network. Partitioning algorithm was applied to brainnet and ultimately the brainnet was divided into some communities.3. Graph theory was used to analyze the difference of community structure. The results of division show that the visual areas are divided into same community in the resting state. It confirms that human brain has a visual network in the resting state. But the visual areas are divided into different communities in the task state.It confirms the two pathways of visual information processing, one of them distributes along the temporal and occipital lobe, another one distributes along parietal and occipital lobe.Through computing and analyzing Z value of correlation coefficient, we found that the links of brainnet are relatively sparse in the resting state, but it is relatively close in the task state. It shows that the brainnet has relatively sparse links to remains a potential to perform tasks in the resting state. But, in the task state, the links of brainnet become close, different brain areas together accomplish tasks.Through computing and analyzing the connection degree of node, we found that whether in the resting state or in the task state, occipital gyrus and lingual gyrus have a relatively high connection degree. It confirms that these two nodes are very important and they have certain stability. And the study also confirms that the parietal has a "spatial attention" feature and play an important role in the process of characteristic binding of color and shape. Lastly, the study confirms that fusiform gyrus and inferior temporal gyrus are responsible for processing color and shape information.