Dynamic Functional Brain Network In Schizophrenia Patients Based On Resting State Functional Magnetic Resonance Imaging

Schizophrenia(SZ)is considered as connectivity barriers between the components of brain networks,which can be caused by development and environment or genetic mechanisms.In recent years,researchers have found that the static brain functional connection research is not enough to reflect the complex and changeable time-varying characteristics of the brain.Dynamic functional brain connectivity is helpful to reveal the imaging characteristics of the disease and can provide more scientific evidence for the research of the pathological mechanism and clinical diagnosis and treatment of schizophrenia.We collected 61 patients with first episode schizophrenia(FESZ),79 patients with chronic schizophrenia(CSZ)and 73 normal controls(NC)in this study.The 90 brain regions were labeled with automatic anatomical labeling(AAL)and divided into seven functional brain networks.The rest magnetic resonance imaging data of all subjects were used to calculate the correlation matrix of time series of blood oxygen level dependent horizontal signals.Then,we calculated the dynamic functional brain network of three groups based on sliding time window method.Six main functional connectivity modes were extracted by Kmeans clustering method.The time-varying characteristics of the three groups of functional connectivity modes were calculated and compared,including time fraction,average dwell time,switching frequency between states,final flowing probability.We analyzed the statistical differences among three groups in brain network interaction modes and timevarying characteristics,and correlation between the significant indicators and clinical scores to explore their physiological significance,for further study on the pathological and physiological mechanism of schizophrenia.In order to verify the repeatability of the results and explore the impact of cluster number and window length on the results,verification analysis was supplemented.This study found that schizophrenia is related to inability to quickly entering and breaking off the default mode network(DMN)which is activated.When brain is in resting state,DMN plays a key regulated role in different functional connectivity modes.The average dwell time of schizophrenia patients in DMN activated states(state2 and state3)were significantly longer than that of NCs,the fraction time were also higher than that of NCs,but the final flowing probability was lower than that of NCs,and the transition frequency from state3 to state5(DMN weak state)was higher than that of NCs.Those results indicated that the activating probability of strong DMN states in schizophrenia was low,but once activated,it would remain active for a long time.We also found that there was a positive correlation between the frequency of state transition and negative symptom score in CSZs.It was speculated that the regulatory ability of which DMN controls other functional networks had degraded,which led to more sensitive to the surrounding environment and increase of negative symptoms.The results of this study showed that there existed abnormal dynamic functional brain connectivity and interaction disorder in SZ which was dominated by DMN.Dynamic FC could provide reliable biological markers for clinical diagnosis of SZ.