Study On Abnormal Brain Function Networks In Bipolar Depression Disorders

Bipolar disorder is a chronic,recurrent disease that affects 1%of the world’s population.Its core cognitive model is the negative self-cognitive schema.Currently,the pathogenesis of bipolar disorder is not clear,and its clinical diagnosis relies on subjective judgement,with a lack of objective imaging markers.Therefore,a non-invasive magnetic resonance imaging technique is used to explore the abnormal neural activity of the functional brain network in this disease,in order to find objective neuroimaging markers and aid the diagnosis and evaluation of the disease in clinical practice.This article combines multi-modal brain imaging indicators,starting from data in both task states and resting states,revealing the abnormal neural response patterns and dynamic brain state abnormalities of bipolar disorder from both static and dynamic dimensions.The main research ideas include the following two parts:1、Using the superior illusion experiment paradigm,the data of patients and normal groups were used to calculate the whole brain task activation and task functional connections.The activation values,functional connection values,and integrated activation and functional connection values were used as three levels of features.The support vector machine classification model was used to identify patients from the normal control group,and the relationship between the most discriminative features and clinical scales was explored.This study found that both activation and functional connections can identify patients from healthy controls,and the most discriminative features in functional connections can predict scales.In addition,the multi-level functional features that integrate activation and modulated connections further improve the performance of classification and prediction.At the same time,it was found that the most discriminative features are mainly located in the midline cortical network and the default network,demonstrating abnormal neural activity responses in self-referencing processes in individuals with bipolar disorder.2、The data of patients and the normal group were normalized using Z-scores.Then,the data of all time points of the normal group were clustered using the K-means clustering algorithm.The co-activation pattern states clustered were mapped to the data of all time points of the patients using the Pearson correlation.Relevant indicators of co-activation patterns were explored,including the state time ratio,state duration,state appearance frequency,and transition probability between states.This study found that six co-activation pattern states were clustered using the K-means clustering algorithm,which were made up of three pairs of opposite states.These states were dominated by spatial patterns in which three networks and other functional networks were co-activated or inhibited.Additionally,significant differences were found in the transition probability of patients’states,indicating abnormal dynamic changes in brain states in patients.In summary,this article explored the abnormality of patient’s brain functional network under cognitive task states using the superior illusion paradigm.Additionally,coactivation pattern analysis was used to investigate the abnormal dynamic state of the patient’s brain under resting conditions.Through the analysis of multi-modal imaging,the neuropathological mechanism of this disease was revealed,providing evidence for the early diagnosis and clinical evaluation of the disease.