Manifold Learning of Dynamic Functional Connectivity Reliably Identifies Functionally Consistent Coupling Patterns in Human Brains.

Yuyuan Yang,Yuyang Zhu,Hui Shen,Yu Lei,Lubin Wang

doi:10.3390/brainsci9110309

Yuyuan Yang, Yuyang Zhu + Show 3 more

Open Access

https://doi.org/10.3390/brainsci9110309

Copy DOI

Abstract

Most previous work on dynamic functional connectivity (dFC) has focused on analyzing temporal traits of functional connectivity (similar coupling patterns at different timepoints), dividing them into functional connectivity states and detecting their between-group differences. However, the coherent functional connectivity of brain activity among the temporal dynamics of functional connectivity remains unknown. In the study, we applied manifold learning of local linear embedding to explore the consistent coupling patterns (CCPs) that reflect functionally homogeneous regions underlying dFC throughout the entire scanning period. By embedding the whole-brain functional connectivity in a low-dimensional manifold space based on the Human Connectome Project (HCP) resting-state data, we identified ten stable patterns of functional coupling across regions that underpin the temporal evolution of dFC. Moreover, some of these CCPs exhibited significant neurophysiological meaning. Furthermore, we apply this method to HCP rsfMR and tfMRI data as well as sleep-deprivation data and found that the topological organization of these low-dimensional structures has high potential for predicting sleep-deprivation states (classification accuracy of 92.3%) and task types (100% identification for all seven tasks).In summary, this work provides a methodology for distilling coherent low-dimensional functional connectivity structures in complex brain dynamics that play an important role in performing tasks or characterizing specific states of the brain.

Highlights

The human brain is a hierarchically organized complex system that can be empirically parsed into functionally specialized units commonly referred to as functional brain networks
We found that local linear embedding of dynamic functional connectivity (dFC) based on the Human Connectome Project (HCP) resting-state dataset leads to a prominent manifold structure of a “specific shape with extended branches” (Figure 2), in which each branch represents the consistent coupling patterns (CCPs) of the resting-state dFC
Taking all matrix-mapped connectivity pairs’ coreness values into account together, we found that the core-quality value matrices showed naturally task-specific states: the resting-state connectivity pairs’ importance where the default node network inherently showed a more important impact to adapt to a task-free environment, which might occur during passive rest and mind-wandering [42]; the CCPs between the cingulo-opercular network (CON) and whole brain show lower coreness values reflecting the weakness of the CON activity, which may be associated with the decrease in self-psychological activity during tasks

Summary

Introduction

The human brain is a hierarchically organized complex system that can be empirically parsed into functionally specialized units commonly referred to as functional brain networks. A large body of neuroimaging studies have made substantial progress in delineating this functional architecture mainly based on resting-state static functional connectivity (sFC). These stable spatiotemporal patterns of resting-state functional activity in a population closely resemble patterns of evoked task-based brain activity [1] and have a significant biological and genetic basis [2,3]. These distributed functional networks cooperate with one another to respond to internal and external stimuli, which underpin various cognitive tasks in the brain. Identification of the brain’s functional architecture has important significance to understanding information processing procedures in the brain and the relationship between brain functions and individual behavior.

Methods

Results

Discussion

Conclusion