The Research On Complex Brain Network Using Cascading Failure Algorithm

From personnel to interpersonal network, from traffic network to large power network, Spread from the body’s metabolism to the current of the brain, we are living in a variety of complex networks. Cascading failures of networks are closely related to our daily life, Cascading effects of cascading failures have brought disasters to human beings. Epilepsy is a disease characterized by recurrent attacks of abnormal discharge of brain neurons, long-term abnormal discharge will damage to the brain. It is necessary to apply the complex network cascading failure theory to brain science, and explore its function characteristics and physiological mechanisms of the brain from the perspective of network. In this paper, we used the human brain function magnetic resonance imaging(Function Magnetic Resonance Imaging, f MRI) data to model a network after processing. We used complex network cascading failure algorithm to explore generalized tonic-clonic seizure(Generalized tonic-clonic seizure, GTCS) epilepsy patients’ brain functional networks. The specific research contents are as follows:The resting state magnetic resonance data were pre processed, including the head movement correction, the inter layer correction, space standardization and so on. Using pearson correlation analysis method to construct the normal control and patients of two 90 * 90 matrixes, comparing the two groups and we can found that the correlation values of Pearson between the brain regions of GTCS patients existe both increased and decreased. Here, we compared the changes in the Pearson related values of the brain areas, and it was found that there were so many brain areas with decreased Pearson correlation values, and the brain areas of GTCS were mainly decreased.Then the changed important brain areas(the right medial frontal gyrus, the left angular gyrus, the left posterior cingulate cortex, the left Superior parietal gyrus, the right triangular part of the superior frontal gyrus, the left precentral gyrus) were attacked by cascading failure algorithm. When attacking the “big degree” nodes, the results shows that found that GTCS patients were detected to have functional changes, the scale of the GTCS patients’ cascading failures was larger than normal group, and in the attack on the “small degree” nodes, the scale of the GTCS patients’ cascading failures was less than normal group. The results showed that when the "big degree" nodes were attacked, it is positively correlated with the value of the coefficient α. While the “small degree” nodes were attacked, it is negatively correlated with the coefficient α. These are both in line with the load- capacity application conclusions which were found by Wang Jianwei, etc. And at the same time, it has verified that the brains of the patients and the normal group both have the “Small-world” attributes. The study found that the significant changes in patients’ brain areas includes the right medial frontal gyrus, the right insula, The left medial frontal gyrus, The left orbital frontal gyrus, the left angular, etc, these brain areas are mainly on the default network. when the medial frontal gyrus area and the frontal lobe abnormal area were damaged, it will mainly lead to some obstacles about voluntary movement, speech, cranial nerves, autonomic nervous function and other aspects of mental activity, it has a direct relationship with the loss of consciousness and convulsions in the GTCS patients. The study found that middle temporal gyrus, middle frontal gyrus were abnormal, and these brain regions are the components of the functional attention network on the back side. The same findings with Wang et al. showed that GTCS patients with the functional attention network have the same damage and reorganization. It further demonstrates the effectiveness and feasibility of the way that use the cascading failures theory to analyze the brain network.