Abstract:Macrovascular biases have been a long-standing challenge for fMRI, limiting its ability to detect spatially specific neural activity. Recent experimental studies, including our own ( ; ), found substantial resting-state macrovascular BOLD fMRI contributions from large veins and arteries, extending into the perivascular tissue at 3 T and 7 T. The objective of this study is to demonstrate the feasibility of predicting, using a biophysical model, the experimental resting-state BOLD fluctuation amplitude (RSFA) and associated functional connectivity (FC) values at 3 Tesla. We investigated the feasibility of both 2D and 3D infinite-cylinder models as well as macrovascular anatomical networks (mVANs) derived from angiograms. Our results demonstrate that: 1) with the availability of mVANs, it is feasible to model macrovascular BOLD FC using both the mVAN-based model and 3D infinite-cylinder models, though the former performed better; 2) biophysical modelling can accurately predict the BOLD pairwise correlation near to large veins (with R ranging from 0.53 to 0.93 across different subjects), but not near to large arteries; 3) compared with FC, biophysical modelling provided less accurate predictions for RSFA; 4) modelling of perivascular BOLD connectivity was feasible at close distances from veins (with R ranging from 0.08 to 0.57), but not arteries, with performance deteriorating with increasing distance. While our current study demonstrates the feasibility of simulating macrovascular BOLD in the resting state, our methodology may also apply to understanding task-based BOLD. Furthermore, these results suggest the possibility of correcting for macrovascular bias in resting-state fMRI and other types of fMRI using biophysical modelling based on vascular anatomy.

Predicting the macrovascular contribution to resting-state fMRI functional connectivity at 3 Tesla: A model-informed approach