Assessment of the macrovascular contribution to resting-state fMRI functional connectivity at 3 Tesla

Abstract

In resting-state functional magnetic resonance imaging (rs-fMRI) functional connectivity (FC) mapping, temporal correlation are widely assumed to reflect synchronized neural-related activaty. Although a large number of studies have demonstrated the potential vascular effects on FC, little research has been conducted on FC resulting from macrovascular signal fluctuations. Previously, Tong et al. found (Tong et al., 2019b) a robust anti-correlation between the fMRI signals in the internal carotid artery and the internal jugular vein (and the sagittal sinus). The present study extends the previous study to include all detectable major veins and arteries in the brain in a systematic analysis of the macrovascular contribution to the functional connectivity of the whole-GM. This study demonstrates that: 1) The macrovasculature consistently exhibited strong correlational connectivity among itself, with the sign of the correlations varying between arterial and venous connectivity; 2) GM connectivity was found to have a strong macrovascular contribution, stronger from veins than arteries; 3) functional connectivity originating from the macrovasculature displayed disproportionately high spatial variability compared to across all GM voxels; 4) macrovascular contributions to connectivity were still evident well beyond the confines of the macrovascular space. These findings highlight the extensive contribution to rs-fMRI BOLD and FC predominantly by large veins, but also by large arteries. These findings pave the way for future studies aimed at more comprehensively modeling and thereby removing these macrovascular contributions.