Multimodal Auditory-visual Brain Network Analysis And Brain Mechanism Research

As the focus of neuroscience shifts from a center area of the brain to the entire network of the brain,the method of statistical inference is gradually oriented towards network analysis.Brain network analysis methods are mainly divided into three kinds: anatomical connections,functional connections and effect connections.The effect connection can reflect the influence of one brain region on the other brain region and can obtain the causal relationship between brain regions.Functional magnetic resonance imaging(fMRI)with high spatial resolution and electroencephalogram(EEG)with high time resolution are two main methods in non-invasive brain imaging technology.In this paper,we use adaptive directional transfer function(ADTF)and dynamic causality model(DCM)methods to study the role of attention on auditory-visual integration in the cross-modal cue-target experimental paradigm.And we also explore the cognitive mechanism of its reflection.The research content mainly include the following three aspects:1.EEG data analysis: Based on the ADTF algorithm,time-varying network is constructed for the scalp EEG data and its network properties are calculated.The variation of network properties with time under different experimental conditions is compared and the brain mechanism is analyzed.The results showed that after the target stimulus appears,the external information is transmitted to the brain.With the brain becoming active,the connections between the nodes increase.Meanwhile,the information transmission speed increases.After a period of stimulation,the brain is in a stable state.At present,the connection and information transmission speed is relatively stable.2.fMRI data analysis: Based on the DCM algorithm,when the stimulus onset asynchrony(SOA)is long,the effects of experimental conditions of effective and ineffective cues on the information flow between the attention and auditory-visual integration brain regions are studied.The results show when the cue is invalid,the information is transferred from the auditory-visual integration area to the brain area about attention.When the cue is valid,the information flows from the attention-related brain regions into the relevant brain regions in a top-down manner.3.Combining fMRI with EEG analysis: The precise coordinate position obtained by fMRI analysis is used as the source space for EEG data analysis.Based on the ADTF algorithm,we construct a time-varying network among visual,auditory,auditory-visual integration and attention-related brain regions.The results show that attention does participate in the auditory-visual integration process and occurs after auditory-visual integration.