Associations between Functional Connectivity Dynamics and BOLD Dynamics Are Heterogeneous Across Brain Networks.

Vince D Calhoun,Xin Di,Yiheng Tu,Zhiguo Zhang,Bharat B Biswal,Zening Fu

doi:10.3389/fnhum.2017.00593

Abstract

Functional brain imaging has revealed two types of dynamic patterns of brain in the resting-state: the dynamics of spontaneous brain activities and the dynamics of functional interconnections between spontaneous brain activities. Although these two types of brain dynamics are usually investigated separately in the literature, recent functional magnetic resonance imaging (fMRI) studies have shown that they exhibit similar spatial patterns, suggesting the dynamics of spontaneous brain activities and the dynamics of their interconnections are associated with each other. In this study, we characterized the local blood oxygenation level dependent (BOLD) dynamics and the functional connectivity dynamics (FCD) in the resting-state, and then investigated their between-region associations. Time-varying FC was estimated as time-varying correlation coefficients using a sliding-window method, and the temporal variability of BOLD and time-varying FC were used to quantify the BOLD dynamics and the FCD, respectively. Our results showed that the BOLD dynamics and the FCD exhibit similar spatial patterns, and they are significantly associated across brain regions. Importantly, such associations are opposite for different types of FC (e.g., within-network FCD are negatively correlated with the BOLD dynamics but the between-network FCD are positively correlated with the BOLD dynamics), and they are spatially heterogeneous across brain networks. The identified heterogeneous associations between the BOLD dynamics and the FCD appear to convey related or even distinct information and might underscore the potential mechanism of brain coordination and co-evolution.

Highlights

The human brain is a highly dynamic system, which is characterized by non-stationary neural activities and represented by ever-changing mental states at time scales ranging from milliseconds to hours (Lehmann and Skrandies, 1980; Lehmann et al, 1987; Allen et al, 2014)
local BOLD dynamics (LBD), functional connectivity (FC) dynamics (FCD) and the most variable or stable FC are depicted in Figure 1 using a connectogram (Detailed FCD of all FC is provided in Figure S1 of Supplementary Materials)
We first validated that both dynamics of blood oxygenation level dependent (BOLD) and dynamics of FC were different across intrinsic connectivity networks (ICNs) and even brain regions

Summary

Introduction

The human brain is a highly dynamic system, which is characterized by non-stationary neural activities and represented by ever-changing mental states at time scales ranging from milliseconds to hours (Lehmann and Skrandies, 1980; Lehmann et al, 1987; Allen et al, 2014). Different brain regions exhibit different degrees of BOLD fluctuations at rest (Zang et al, 2007; Kannurpatti et al, 2008; Biswal et al, 2010), with the default mode network (DMN) showing higher amplitude of low frequency fluctuations (ALFF) in resting-state (Zang et al, 2007; Zou et al, 2008) This observation coincides with the fact that the DMN regions generally have higher neuronal activity (Shulman et al, 1997) and metabolic rate (Raichle et al, 2001) during the resting-state, suggesting an association between the amplitude of fluctuations and the level of local neuronal activity. Spatial heterogeneity of FCD has been reported in the literature: FC between DMN and frontal-parietal network has relatively larger variability, while the within-network FC of visual network, DMN and sensorimotor network (SMN) is more stable than that of orbitofrontal-limbic network (Allen et al, 2014; Zalesky et al, 2014)

Methods

Results

Discussion

Conclusion