| Bipolar disorder(BD)is a mental disorder manifested as recurrent episodes of depression and mania/hypomania,which has the highest suicide rate of all psychiatric conditions.Suicide is a serious public health and social problem.It is very difficult to evaluate and predict the suicide risk only based on clinical characteristics.In recent years,the application of neuroimaging technology has shown that the brain functional of BD patients with suicide attempts(SA)was significantly changed.However,the potential physiological mechanism of suicide is still unclear.The main purpose of the paper is to explore the relationship between the neuronal signatures for BD patients with SA and clinical signatures via the brain functional network constructing with resting state functional magnetic resonance imaging(f MRI)data from different perspectives.We hope to provide objective markers for assisting clinical evaluation and predicting suicide risk.The main contents are as follows:1.The brain dysfunction in BD patients with SA was explored from the perspective of topological properties of static brain functional network.The static brain functional network was constructed using resting state f MRI data in BD patients and healthy controls(HC).The BD patients were divided into non SA group(NSA)and SA group according to whether there was a history of suicide behaviors.In the study,the graph theory was used to analyze the topological properties of the brain functional network at the level of whole brain,module and brain region between three groups.The results showed that the global attributes in SA patients did not change significantly.Compared with other two groups,the functional connectivity strength between the frontoparietal network and the sub-cortical network(Sub C)was significantly weakened in SA patients.In addition,the node strength of the raphe nucleus and the node efficiency of the right subgenual anterior cingulate cortex(sg ACC)were significantly reduced in SA patients relative to NSA and HCs,and showed a significant negative correlation with NGASR scores.These findings indicated that the networks related to emotional management and cognitive regulation were damaged in SA patients,which increased the suicide risk.2.Based on the first part of the study,the influence of the ACC on suicide was further explored,when the bilateral ACC was taken as the region of interest.The dynamic functional connectivity and K-means clustering algorithm were used to capture the dynamic brain functional states centered on the ACC.First,the functional state related to suicide was obtained by dwell time and functional connectivity.In suicide-related functional state,the functional connection strength between the ACC and the brain regions in the Sub C network of the SA patients was significantly increased.The abnormal functional interaction of the ACC destroyed its mediating effect on the overall emotion,cognition and decision-making evaluation of the cerebral cortex and sub-cortical regions.In addition,the frequent state transition makes the stability of the dynamic functional of the ACC became worse,which disturbed the regular time distribution of the emotional management ability.These findings suggested that the temporal variability of dynamic states with ACC has potential significance in marking suicide risk in BD patients.3.To explore the state mechanism of the whole brain in SA patients.First,the dynamic function network about whole brain was constructed.Then,the dynamic function states were captured by K-means clustering.The results indicated that the temporal properties of dynamic connectivity in SA patients have changed significantly.The fractional windows,dwell time of suicide-related state and number of state transitions were significantly correlated to suicide risk,which indicates that the temporal characteristics of whole brain dynamic functional states can be used to assess the suicide risk of BD patients.In addition,the dynamic dysregulation of the functional interaction between the default mode network and the Sub C may increase patients’ error congnition about suicidal behaviors,thus prompting patients to take high-risk behaviors.The findings also found that the sensory motor cortex,visual cortex and attention network of the SA patients were dysfunctional,which may be a potential physiological mechanism of suicide in BD patients.4.The part of the study aims to investigate the change mechanism of neural activities from HC to BD patients with SA.The Hopf bifurcation model was used to simulate brain dynamics based on the similarity of brain functional states in empirical data and simulated data in HC,NSA and SA group respectively.Subsequently,the average functional connectivity strength,functional differentiation(modularity)and functional integration among three groups were compared in the empirical and simulated brain functional networks.The results showed that the neural activity of the HC was closed to a Hopf bifurcation,and the simulated neural activity of the three groups were similar to the empirical signals.Furthermore,the whole brain computional model can reproduce the differences in the brain functional network constructed by the empirical data,and the bifurcation parameters show a decreasing trend.The results indicated that the global oscillation level of brain neural activity from HC to SA patients was gradually weakened,rather than a discrete state related to suicidal behavior.Combined with virtual lesion and graph theory analysis,the findings showed that the dynamics changes of brain regions within frontal-limbic system had an important impact on suicidal behavior.These findings suggested that the whole brain computional model can provide a new perspective for exploring the underlying mechanisms of suicide in BD patients.In the whole,the relationship between brain functional damage and suicide in SA patients were analyzed from the perspective of static functional network,dynamic functional network and whole brain computational model,with findings to providing help for the evaluating and predicting suicide risk in BD patients. |
PDF Full Text Request