WE-FG-206-02: Brief Variations of BOLD Signal in Resting State FMRI Leading to Functional Connectivity Pattern Identification.

Med Phys. 2016 Jun;43(6):3831

Authors: Ball N, Chen N

Abstract
PURPOSE: To assess the role of spontaneous brain activity in identifying functional connectivity patterns in the brain during resting-state using both high and low blood-oxygen level dependent (BOLD) signals from functional magnetic resonance imaging (fMRI).
METHODS: fMRI BOLD signals within seed regions from known functional connectivity networks were extracted to find an average seed-region time profile. A threshold was applied to the BOLD signal and and time points were selected where the BOLD signal satisfied the threshold requirements. The same threshold was also applied to the time profile of each voxel in the brain, i.e., including voxels outside of the seed region. The signal profile of seed region time points above the threshold was compared to the signal profile of whole brain voxel time points over the threshold. Any time point that occurred on both signal profiles was considered a match. We then evaluated whether or not a voxel was part of the functional connectivity network based on the percentage of matching time points.
RESULTS: These findings suggest that functional connectivity patterns can be identified only using a small number of time points. Using only 10% of the time points during a functional MRI functional connectivity patterns can be reproduced with varying degrees of accuracy. Using this same method, with a 33% threshold a histogram of numbers of matches per matches possible within the region where we expect to see the functional connectivity pattern, shows no significant separation between using the top third, the middle third, or the bottom third of data.
CONCLUSION: This work indicates that functional connectivity patterns can be reproduced not only with spontaneous activity of high signals, but also from brief instances of low BOLD signal. As the potential role of resting-state fMRI for clinical diagnosis evolves, understanding the fundamentals of functional connectivity patterns is critical. NIH R01-NS074045.

PMID: 28048928 [PubMed - in process]