Temporal Changes in Local Functional Connectivity Density Reflect the Temporal Variability of the Amplitude of Low Frequency Fluctuations in Gray Matter.

• Main Authors: D Tomasi, E Shokri-Kojori, N D Volkow
• Format: Article
• Language: English
• Published: Public Library of Science (PLoS) 2016-01-01
• Series: PLoS ONE
• Online Access: http://europepmc.org/articles/PMC4846007?pdf=render

Data-driven functional connectivity density (FCD) mapping is being increasingly utilized to assess brain connectomics at rest in the healthy brain and its disruption in neuropsychiatric diseases with the underlying assumption that the spatiotemporal hub distribution is stationary. However, recent studies show that functional connectivity is highly dynamic. Here we study the temporal variability of the local FCD (lFCD) at high spatiotemporal resolution (2-mm isotropic; 0.72s) using a sliding-window approach and 'resting-state' datasets from 40 healthy subjects collected under the Human Connectome Project. Prominent functional connectivity hubs in visual and posterior parietal cortices had pronounced temporal changes in local FCD. These dynamic patterns in the strength of the lFCD hubs occurred in cortical gray matter with high sensitivity (up to 85%) and specificity (> 85%) and showed high reproducibility (up to 72%) across sessions and high test-retest reliability (ICC(3,1) > 0.5). The temporal changes in lFCD predominantly occurred in medial occipitoparietal regions and were proportional to the strength of the connectivity hubs. The temporal variability of the lFCD was associated with the amplitude of the low frequency fluctuations (ALFF). Pure randomness did not account for the probability distribution of lFCD. Shannon entropy increased in proportion to the strength of the lFCD hubs suggesting high average flow of information per unit of time in the lFCD hubs, particularly in medial occipitoparietal regions. Thus, the higher dynamic range of the lFCD hubs is consistent with their role in the complex orchestration of interacting brain networks.