MRI-based Grey Matter And White Matter Brain Network Method And Application In Schizophrenia

Schizophrenia,a common mental disease which occurrs in young adults,has a variety of pathological characteristics.The pathogenesis and effective treatment of schizophrenia is an important scientific problem.Magnetic resonance imaging(MRI)has been one of the most powerful technology to explore brain network and pathological mechanism of schizophrenia.Based on multimodal MRI technology,this dissertation developed corresponding brain network analysis methods for specific difficult problems in schizophrenia.From two perspectives(brain structure and function)and two levels(gray matter and white matter),the dissertation verified,supplemented and revised the hypothesis of brain ‘dysconnectivity' in schizophrenia.Besides,this dissertation investigated the schizophrenia brain network changes through longitudinal follow–up intervention studies,to provide a theoretical basis for clinical treatment and intervention ideas.This dissertation can be summarized into three parts,including seven main contents.Part 1: As one of neural development/degeneration disorders,it often needs a long–term follow–up for patients to describe the brain changes of schizophrenia.Aim to the problem,this part proposed and developed two methods of brain network,using the cross–sectional MRI data of patients to depict the dynamic changes of brain gray matter and white matter with progress of schizophrenia development,which confirmed,complemented and combined the hypothesis of the neurodevelopment/degeneration and brain ‘dysconnectivity' in schizophrenia.This part mainly includes the following two contents:1.Brain imaging technology based on MRI can well characterize the structure of brain gray matter.However,traditional methods of gray matter structural analysis,such as voxel–base morphometry(VBM)and Freesurfer's gray matter cortical analysis,can only delineate the structural characteristics of individual regions and ignore the relationship between regions.In addition,the structural features of the brain are not constant,which will change with the development,aging,disease and other factors.VBM cannot describe the causal relationship between different regions when brain changes.This paper developed a population–based covariant causal network method for gray matter analysis.By sorting the gray matter morphological images of patients with schizophrenia according to the duration of the disease,a population–level "time series" of gray matter structure with the disease development was constructed.Using Granger causality analysis,this chapter delineated the sequence of gray matter atrophy and confirmed the progressive pattern of gray matter atrophy in schizophrenia patients,which provided brain imaging evidence from gray matter for the theoretical combination of neurodevelopment/degeneration hypothesis and ‘dysconnectivity'hypothesis in schizophrenia.2.To further confirm the progressive pattern of brain impairment in schizophrenia,this chapter proposed a brain image analysis method – "PRogressive br Ain Dynamic Analysis based on ranked images,PRADA".Diffusion tensor imaging(DTI)data of schizophrenia patients were sorted according to the length of the illness duration to generate white matter "time series".Using sliding window strategy,"time series" were divided into several windows to describe dynamics of white matter development.The onset map was proposed to reveal abnormalities of white matter structure at different stages of schizophrenia,which provided brain imaging evidence from white matter for the theoretical combination of neurodevelopment/degeneration hypothesis and‘dysconnectivity' hypothesis in schizophrenia.Part 2: For the blood oxygen level–dependent(BOLD)–functional MRI(f MRI)signal of white matter,the second part mainly developed the analytical methods of white matter functional network and tried to solve the problems in processing the BOLD–f MRI signal of white matter.These methods can validate and extend the potential physiological basis and clinical use of BOLD–f MRI signal in white matter from multiple levels.Applications in schizophrenia provided the evidences of white matter functional signal for the ‘dysconnectivity' hypothesis in schizophrenia.This part mainly includes three contents:1.Although DTI is an effective method for the characterization of white matter structure,it cannot uncover the potential functional activations in white matter.To investigate the relationship between the structural impairment and dysfunction in white matter,this chapter proposed a DTI–f MRI fusion method – "Skeleton–based White matter Functional Analysis,SWAF".To implement the skeleton–based analysis of low frequency oscillations(LFO)in white matter and reflect the spontaneous activity of white matter,this method projected the white matter signal of DTI and white matter functional signal of f MRI to a common skeleton.This chapter also systematically evaluated the reproducibility of white matter LFO,the coupling between white matter LFO and structure,the relationship between white matter LFO and behaviors.Taken together,this part demonstrated the physiological significance and clinical value of BOLD–f MRI signal in white matter from various perspectives.2.Based on the above,this content used the BOLD–f MRI signal to build the large–scale networks of white matter and gray matter respectively,and analyzed the network properties(including clustering coefficient,shortest path length,small world metric,local efficiency,hierarchy,synchronization)and node properties(including nodal efficiency,nodal degree and nodal centrality)based on graph theory.This part illustrated the topological changes of large–scale functional networks of white matter and gray matter from network level in patients with schizophrenia.It provided the evidences from brain network for the ‘dysconnectivity' hypothesis in schizophrenia and supplemented the explanation mechanism of white matter function.3.Although the previous studies explored the gray matter network and white matter network respectively,the relationship between gray matter and white matter was not directly evaluated.In this chapter,cluster analysis is used to identify different white matter functional networks,and functional connectivity is used to characterize the functional relationship between white matter networks and gray matter networks.By comparing the differences of white matter functional networks between schizophrenic patients and healthy controls,the relationship between gray–white matter functional connectivity and the pathophysiology of schizophrenia was clarified.From the perspective of gray–white matter connectivity,it provided a supplement to the hypothesis of brain ‘dysconnectivity' in schizophrenia.Part 3: According to the clinical symptoms of patients with schizophrenia,two ways(drug intervention and physical intervention)were used for intervention.The effects of drug therapy and electroconvulsive therapy(ECT)on the brain of patients with schizophrenia and the potential regulatory mechanisms involved were explored through hypothesis driven studies to integrate the brain ‘dysconnectivity' hypothesis of schizophrenia with clinical treatment.This part includes the following two main contents:1.Although previous studies have confirmed the hypothesis of brain‘dysconnectivity' in schizophrenia from the structural and functional levels,these evidences were all from cross–sectional data of patients with schizophrenia.In this part,the effects of antipsychotic drug intervention on cerebral cortex in schizophrenia were investigated by longitudinal follow up.Using surface–based morphology analysis and structure covariance analysis,this chapter comprehensively evaluated cerebral morphology before and after treatment and the change of the covariance structure network,explored the short–term effect of antipsychotic drugs on the cortical thickness and cortical–cortical connectivity.By calculating the global efficiency of the network,the integration of the structural covariant network was characterized to reveal the potential relationship beten cortical morphological changes and brain network reconfiguration and to combine the brain ‘dysconnectivity' hypothesis of schizophrenia with clinical treatment.2.Although the previous part explored the effect of antipsychotic drug intervention on the cerebral cortex in schizophrenia through longitudinal follow up.However,schizophrenia is a very heterogeneous population.Some patients with schizophrenia have drug resistance,while some patients with schizophrenia have specific symptoms that need to be treated,which requires specific interventions,such as ECT.This study aimed to explore the effects of ECT intervention on the function and structure of insula and hippocampus in the special population of schizophrenia through longitudinal follow up,in an attempt to provide clinical significance and therapeutic targets for the hypothesis of brain ‘dysconnectivity' in schizophrenia from the special population and special therapy.In general,this dissertation,based on multimodal MRI including the BOLD–f MRI,DTI and structural MRI,developed the corresponding brain network analysis methods from the two perspectives of brain structure and brain function,verified and supplemented the brain ‘dysconnectivity' hypothesis of schizophrenia from gray matter and white matter respectively,and provided clinical significance for the hypothesis of brain ‘dysconnectivity' in schizophrenia through intervention studies.