Construction And Analysis Of Resting State Whole Brain Network Model And Brain Functional Network

The neural mass model(NMM)is an important tool for modelling brain dynamics modelling at the level of neural mass,and is used to study the abnormal electroencephalogram signals generated by the brain in patients with epilepsy and cognitive impairment.Compared with other statistical mathematical models,the most important advantage of neural mass model is that the state variables and parameters in the model have clearly physiological meaning.The exploration of the physiological parameters of NMM is helpful to discover the physiological mechanism of specific electroencephalogram signals generated by the brain.Firstly,for single-region NMM,thalamic model,J&R model and Wendling model are constructed to analyze the impact of changes in physiological parameters on the output signal of the model.The study of thalamic model and J&R model found that changes in parameters can lead to the decrease of output dominant frequency,and the dominant frequency is the largest when all parameters are taken as standard values.The study of Wendling model found that the output of the model can be divided into six types.Secondly,for double-region coupled NMM,cortico-cortical coupled NMM is constructed to analyze the impact of changes in coupling coefficients on the output signals of each brain region.It was found that as the coupling coefficients increased,the output signals of the two brain regions tended to synchronize.The thalamo-cortical coupled NMM is constructed to analyze the impact of changes in coupling coefficients on the output signals of each brain region.It was found that increasing the coupling coefficient from thalamic to cortical,the output signals of the two brain regions tended to synchronize,while increasing the coupling coefficient from cortical to the thalamic,the output signal of the thalamic model gradually decreased to the theta band.Analyzing the impact of changes in coupling coefficient and other synapse parameters in thalamo-cortical coupled NMM on the relative high-frequency power of the output signal of the cortical module.It was found that except for the fast inhibitory parameters,all other parameters can lead to a significantly decrease in relative high frequency power.Finally,the resting state whole-brain network model is constructed to analyze the correlation between the resting state simulated functional connectivity matrix and the empirical functional connectivity matrix under different global coupling coefficients,it was found that when the global coupling coefficient was set to 20.3,the model fitted to a maximum value of 67.6%.Further analysis of the common graph theory measures and small world network properties of simulated brain network and real brain network under different thresholds shows that the common graph theory measures of simulated values are similar to real values.When the threshold is set to 0.22,both simulated and real brain networks have small world properties,indicating that the constructed resting state whole brain network model can be used to study the neural mechanisms of complex brain networks to a certain extent.