Study On The Mechanism Of Schizophrenia Based On Large-scale Brain Network Modeling

Schizophrenia is a common mental disorder characterized by widespread abnormalities in reality perception,thinking processes,and cognitive functions.To date,the pathogenesis of schizophrenia remains unclear,and the treatment and intervention measures for patients need to be further studied.In recent years,the large-scale brain simulation technology has been widely used in exploring the pathogenesis of brain diseases and developing effective regulatory strategies.Based on the whole-brain model,we cannot only explore the pathogenesis of schizophrenia,but also take advantage of the high dimensional parameter space to develop new methods of regulation and intervention.In this thesis,we constructed a high-precision whole-brain computational model for patients with schizophrenia and healthy controls by systematically integrating multimodal brain imaging data from the two groups.The model was used to explore the neural mechanisms and regulatory strategies of patients with schizophrenia,and to analyze the static and dynamic abnormal functional networks and regulatory effects of patients.The results obtained were as follows:Firstly,the analysis of brain functional connectivity(FC)revealed two distinct features in patients with schizophrenia: "hypo-connectivity" and "hyper-connectivity".To explore the neural mechanisms of schizophrenia,we constructed whole-brain computational models for two types of patients and healthy controls,and used the "structure-function" optimization strategy to significantly improve the performance of model.The model revealed that the abnormal excitation and inhibition in the whole-brain of patients.To restore the abnormal excitation and inhibition levels in patients,the model controlled the current flow through the cortical loop to diffusely regulate the strength of synapses between whole-brain networks.As the excitation and inhibition levels returned to normal,the functional connectivity distance between the two groups decreased gradually,and the FC of patients’ brain was significantly improved.Secondly,experience-based and simulation-based static FC analysis revealed significant abnormal connections in most brain regions,cerebral lobes,and sub-networks in patients with schizophrenia.In addition,the brain network properties of patients,including global efficiency,local efficiency,clustering coefficient,and characteristic path length,also showed significant abnormalities.After restoring the abnormal excitation and inhibition levels in the two groups,the abnormal functional network connections and network topology abnormalities were partially restored.Finally,dynamic FC clustering analysis through simulation revealed that patients with schizophrenia showed different regional connections in various sub-states.In addition,the time properties of patients in some sub-states,including average residence time,state probability,and state transition probability,also showed significant abnormalities.After regulatory intervention,the abnormal regional connections and time properties in patients were partially restored.Overall,the static and dynamic FC analyses suggested that the abnormal functional network in patients with schizophrenia was significant,and restoration of the abnormal excitation and inhibition levels could significantly improve the abnormal functional network.This result indicates that local loop abnormal excitation-inhibition may be a key neural mechanism of schizophrenia.This thesis fully utilizes the advantages of highdimensional parameter space of computational models and develops a regulatory method based on loop current,which provides a model reference for exploring the neural mechanisms and regulatory strategies of schizophrenia.