Effective Connectivity Extracted from Resting-State Fmri Images Using Transfer Entropy
Z. Wu,X. Chen,M. Gao,M. Hong,Z. He,H. Hong,J. Shen
DOI: https://doi.org/10.1016/j.irbm.2021.02.007
IF: 5.5
2021-01-01
IRBM
Abstract:Background and objective: Based on magnetic resonance imaging (MRI), macroscopic structural and functional connectivity of human brain has been widely explored in the last decade. However, little work has been done on effective connectivity between individual brain parcels. In this preliminary study, we aim to investigate whole-brain effective connectivity networks from resting-state functional MRI (rs-fMRI) images. Material and methods: After the functional connectivity networks of 26 healthy subjects (aged from 25 to 35 years old) from Human Connectome Project database were derived from rs-fMRI images with dynamic time warping, proportional thresholding (PT) was performed on the functional connectivity matrices by retaining the PT% strongest functional connections. PT% ranges from 40% to 10% in steps of 5%. Then, effective connections corresponding to the PT% strongest functional connections, both bi-directional and unidirectional, were estimated with Renyi's 2-order transfer entropy (TE) method. Topological metrics of the built functional and effective connectivity networks were further characterized, including clustering coefficient, transitivity, and modularity. Results: It is found that the effective connectivity networks exhibit small world attributes, and that the networks contain a subset of highly interactive regions, including right frontal pole (in-degree 6), left middle frontal gyrus (in-degree 8, out-degree 1), right precentral gyrus (out-degree 9), left precentral gyrus (out-degree 7), right posterior division of supramarginal gyrus (in-degree 2, out-degree 3), left angular gyrus (out-degree 6), left inferior division of lateral occipital cortex (out-degree 6), right occipital pole (in-degree 5), right cerebellum 7b parcel (in-degree 15), and right cerebellum 8 parcel (in-degree 7, out-degree 1). Conclusions: The observations in this study provide information about the casual interactions among brain parcels in resting state, helping reveal how different subregions of large-scale distributed neural networks are coupled together in performing cognitive functions. (C) 2021 AGBM. Published by Elsevier Masson SAS. All rights reserved.