Construction And Application Of Brain Network Of Based On Event Related Potential

Event Related Potential(ERP)is an important evoked potential that has been widely employed in cognitive research and psychological diseases.It has high temporal resolution,and a specific structure on latency,topological distribution,reaction pattern and repeatability.It is believed that ERP can reflect the course of information processing during a cognitive task.Traditional research on ERP mostly employs signal averaging technique to reveal the underlying neural mechanism.However,cognitive process usually requires the brain to function in a unified manner through information integration and separation.Unfortunately,the important aspects of information integration and separation occurred during cognitive process are largely ignored by the traditional method.Therefore,this thesis put forward an appropriate method to construct brain network of ERP,and apply the theory of complex brain network to investigate brain network in different cognitive tasks to reveal the mechanism of brain information processing on network level.The proposed method for constructing ERP network was applied on study the auditory P3 in schizophrenia(SZ)and error relative component in healthy subjects.The main content of this thesis includes:1.We proposed a network formation method that is suitable for the analysis of ERP.The method selects 60 electrodes according to the international 10-20 system as nodes of the network.Considering ERP has certain repeatability and response pattern,signal is segmented into epochs according to the component of interest.The connection strength between a pair of electrodes is calculated by averaging strength of correspond stimulus of each channel.2.With the developed network construction method,we analyzed the network characteristics of auditory P3 evoked by auditory oddball paradigm in SZ.After obtaining the brain functional network of auditory P3,the differences on electrical physiological and graph theory was studied between the SZs and normal controls.We further analyzed the correlation between electrophysiological parameters and ERP network properties.Our results showed that auditory P3 of SZ has decreased amplitude and prolonged latency.From a network perspective,SZ had lower local and global efficiency,robustness and coupling degree.Ourfindings are well in line with ‘dysconnection' hypothesis in SZ pathology.Further,a significant correlation between the amplitude and the network properties of auditory P3 in SZ was identified(p < 0.05).This result implicates that underlying brain network abnormality maybe the cause of the abnormal auditory P3 that is observed in SZ group.3.We analyzed the network characteristics of different stages during error processing.This thesis constructed ERP network for different stages during error processing includes the baseline,ERN,Pe and the post-action 400~550ms stage.Various network measures were selected in order to study the network separation and integration.The results showed that the temporal region was less involved in error processing,and the brain network was more efficient before error related action.ERN had a higher pattern of information processing than Pe.This thesis provided the new view on the mechanism of error processing from the perspective of brain network.In conclusion,this thesis was an important attempt to explore the ERP brain network and obtained some meaningful results.The proposed method of constructing ERP network and the findings of this thesis provided new views and ideas for the research of brain networks.