Frequency specific contribution of intrinsic connectivity networks to the integration in brain networks

Yeong-Hun Park,Viktoriya Bourakova,Jungho Cha,Jong-Min Lee

doi:10.1038/s41598-019-40699-z

Abstract

Brain networks are integrated and segregated into several intrinsic connectivity networks (ICNs). Frequency specificity of ICNs have been studied to show that different ICNs have a unqiue contribution to brain network integration along frequencies. The purpose of this study was to evaluate the contribution of individual ICN to brain network integration along their frequency. We used 14 ICNs and determined 2 frequency bands (LF1, 0.03~0.08 Hz and LF2, 0.009~0.012 Hz) from the hierarchical clustering of 101 frequency bins. We proposed a novel measure, called ICN efficiency, representing the difference between the global efficiencies of the whole brain network with and without the ICN to evaluate the contribution of the ICN to brain network integration. We found that each ICN had a different ICN efficiency at 2 frequency bands. We also found that the distinct subregions of the same ICN had a frequency specific contribution to brain network integration. Futhermore, the integration with other ICNs of the distinct subregions of the same ICN were different at 2 frequency bands. In conclusion, the contribution of each ICN to brain network integration is frequency specific and distinct subregions of the same ICN have functionally distinct roles with other ICNs at 2 frequency bands.

Highlights

Functional connectivity (FC) is defined as the correlation of low frequency (0.009~0.08 Hz) fluctuations among anatomically distinct brain areas on resting-state functional magnetic resonance imaging[1]
EICN was calculated to determine whether the contribution of intrinsic connectivity networks (ICNs) to brain network integration differs along the frequency bands (Fig. 2 shows significant differences in t-values with Bonferroni correction (p < 0.005))
Note that the statistically significant difference in EICN and the difference in the number of betweenness centrality (BC) hubs were consistent in the Auditory network (AN), Default mode network (DMN), Salience network (SN) and Precuneus network (PCN)

Summary

Introduction

Functional connectivity (FC) is defined as the correlation of low frequency (0.009~0.08 Hz) fluctuations among anatomically distinct brain areas on resting-state functional magnetic resonance imaging (fMRI)[1]. Unlike Sasai, et al, Thompson and Fransson[13] evaluated the contribution of an ICN to the integration of a brain network composed of 10 ICNs using the strength contribution and detected hubs by considering the betweenness centrality (BC) based on the concept of a path. This may not be sufficient for evaluating the contribution of an ICN to the integration of the brain network because it is based on each edge’s value over the sum of all edges and only considers the first degree of each node. This could result in evaluating the contribution of an ICN to whole brain network integration

Objectives

Methods

Results

Conclusion