Study Of Individual Differences Based On Multimodal Brain Regions

In recent years,with the development of neuroscience,many countries focus in exploring the mysteries of the brain with strong interests.For example,the National Institutes of Health previously launched the Human Connectome Project(HCP),which uses MRI Technology to further understand the connection and function of neural circuits in the human brain.China has also launched its own "brain plan".The analysis of individual differences in the human brain is an important research direction in neuroscience.For the current analysis of individual differences,although there are a lot of research data,but there are problems of inconsistent research methods and statistical standards.Even for experiments that use brain networks as research methods,there are problems with inconsistent data sources and different construction methods,resulting in differences in final results.In the analysis and comparison of individual differences,the group network to be used currently mainly has the problem of inconsistency between modularity and real network,and at the same time,it cannot meet the needs of current high-precision brain networks.This paper mainly aims at improving the existing problems in the analysis of individual differences,and carries out the study and analysis of individual differences from the perspective of brain network.The main work includes:(1)The current finest multimodal brain partitioning model at the University of Washington was used to construct the brain network,replacing the previous single-modal brain partitioning scheme.The constructed brain network can reflect the network state of human brain more accurately.At the same time,in view of the high data quality requirements of the partitioning scheme,the original data processing process was changed with a new scheme,so that some low-precision or lack of T2 w and B0 field map image data can also use multimodal Partitioning scheme for brain partitioning.(2)A new group network model is proposed to replace the previously used average network model and Bonferroni corrected network model.The new overlay group network model generates group networks from the perspective of connection.The human brain is an economic model and modular similarity,which is used as a reference threshold and a final threshold to limit the generation of the final group network.Therefore,it can not only ensure the rationality of group network threshold selection,but also make the modular attributes and the real network have high consistency.(3)Through the comparison of brain network parameters,comparative analysis of individual differences.In this paper,we compare and analyze the global and local indexes,and compare them with the group network model.We find out the brain regions with large individual differences in the new partition model,and find that the global indexes have certain stability and some connections across the edges have high variability.Compared with the most previous studies on individual differences,which were limited to some brain regions,this paper conducted a relatively comprehensive analysis of individual differences in the whole brain region from the perspective of brain network.