Construction And Analysis Of Brain Functional Network Based On IOTA And Improved Kendall Algorithm

Complex networks can be regarded as an abstraction of the description of a complex system. Starting from the end of the twentieth Century, the theory of complex network has gradually penetrated into all fields of social science, and it has become one of the most important tools for people to solve the problem. The complex network theory is helpful in studying the interaction between different brain regions, topology structure and the dynamic information, as well as the relationship between disease and physiological function. EEG is an important tool for the epilepsy diagnosis and prediction, in view of the great significance of EEG signals analysis. This paper mainly focuses on the following work:Firstly, the thesis adopts C-C phase space reconstruction algorithm, to define the parameters of EEG time series phase space reconstruction, then reconstructes 16 brain electrical signals in phase space. Since the traditional phase diagram analysis is sensitive to the phase space reconstruction parameters and data length, this article employes the phase space vector on visualization mapping of the Euclidean distance matrix. We employ this method on the different leads EEG signals in patients with epilepsy, and the experimental results can clearly distinguish EEG signals with abnormal charging and discharging, which helps epilepsy lesions positioning analysis.Secondly, this thesis contributes to construct brain functional network from a new perspective, based on EEG(electroencephalogram) by one-way algorithm of inner composition alignment, and visualize the network topology. Then the network’s statistical characteristics the average node degree and clustering coefficient are analyzed, compared with normal brain network, the node average degree and clustering coefficient of epileptic brain network, are significantly different from those of normal. Experimental results validates the effectiveness of IOTA algorithm on analysizing brain functional network, as well as researching the epileptic brain functional network dynamics, and provides important reference basis for clinical diagnosis.Thirdly, this thesis puts forward an improved synchronization algorithm of IRC(inverse rank correlation) based on Kendall rank correlation. Kendall rank correlation is an generalized algorithm for analysizing and measuring nonlinear dynamics. We make use of IRC algorithm on EEG data to build the brain functional network, and analysize the average node degree indicators of the constructed brain functional network, in order to investigate whether epileptic brain functional network is different from normal people. Results show that we can make significantly distinction between epilepsy and normal by using the improved algorithm on brain functional network, besides this algorithm only needs to record a short EEG data, it is suitable for the distinction between epileptic and normal brain functional network via the average node degree index, it may help to further understand the brain function and epileptic neural abnormal dynamics behavior.