Resting-state functional connectivity has been found to vary considerably over a time scale of seconds to minutes. A set of methods/approaches have been proposed to study such temporal variation and its relation to dynamic brain connectivity, which led to important insight into the mechanisms of resting-state fMRI connectivity. More importantly, these emerging methods can extract a new dimension of information from the resting-state fMRI that may be critical for assessing brain functions/dysfunctions. The study of dynamic functional connectivity opens up new horizons for the resting-state fMRI research.
Outcome/Objectives
To understand potential sources of dynamic functional connectivity, to recognize differences between dynamic connectivity and stationary connectivity, and to learn methods/approaches for studying dynamic functional connectivity.Purpose
Resting-state fMRI has been widely used for studying brain functional connectivity and revolutionized our understanding about the organization of large-scale brain networks. Conventional functional connectivity analyses employ stationary metrics and thus ignore time-varying information in resting-state fMRI data. It was found recently that temporal changes of functional connectivity in the time scale of seconds to minutes contains important information about brain network dynamics. A set of methods/approaches have been proposed and used to study the dynamic brain connectivity with resting-state fMRI data, and they not only extract new information not contained in conventional metrics, but also provides new insight into the mechanisms of resting-state fMRI connectivity. It is therefore important to summarize the recent advances in this area of research.Methods
Recent studies that employ novel approaches to explore dynamic aspect of the resting-state fMRI and associated findings will be reviewed.Results
Resting-state functional connectivity measured via fMRI correlations demonstrate large variation with the time scale of seconds to minutes. A part of such temporal dynamics can be explained by high-order correlations of multiple brain regions, i.e., their simultaneous co-activation at single time points. The observation leads to methods that can temporally decompose resting-state networks resulted from stationary analyses into multiple co-activation or connectivity patterns that occur at distinct time and may reflect different brain configurations. Moreover, investigation of wide-spread, spatially non-specific functional connectivity from the perspective of dynamics also suggests that they originate from events occurring only at a small proportion of time and are related to brain vigilance. The finding is consistent with recent results about the electrophysiological correlates of the global component of resting-state fMRI signals.Discussion/Conclusion
The study of dynamic functional connectivity opens a new window for the resting-state fMRI research. It advanced our understanding about resting-state fMRI connectivity and networks. More importantly, it provides a new dimension of information that is not captured by conventional approaches but could be critical for assessing brain functions/dysfunctions. Novel methods are urgently needed for a better quantification of temporal dynamics in resting-state fMRI.