A practical modification to a resting state fMRI protocol for improved characterization of cerebrovascular function

Abstract

Cerebrovascular reactivity (CVR), defined here as the Blood Oxygenation Level Dependent (BOLD) response to a CO 2 pressure change, is a useful metric of cerebrovascular function. Both the amplitude and the timing (hemodynamic lag) of the CVR response can bring insight into the nature of a cerebrovascular pathology and aid in understanding noise confounds when using functional Magnetic Resonance Imaging (fMRI) to study neural activity. This research assessed a practical modification to a typical resting-state fMRI protocol, to improve the characterization of cerebrovascular function. In 9 healthy subjects, we modelled CVR and lag in three resting-state data segments, and in data segments which added a 2–3 minute breathing task to the start of a resting-state segment. Two different breathing tasks were used to induce fluctuations in arterial CO 2 pressure: a breath-hold task to induce hypercapnia (CO 2 increase) and a cued deep breathing task to induce hypocapnia (CO 2 decrease). Our analysis produced voxel-wise estimates of the amplitude (CVR) and timing (lag) of the BOLD-fMRI response to CO 2 by systematically shifting the CO 2 regressor in time to optimize the model fit. This optimization inherently increases gray matter CVR values and fit statistics. The inclusion of a simple breathing task, compared to a resting-state scan only, increases the number of voxels in the brain that have a significant relationship between CO 2 and BOLD-fMRI signals, and improves our confidence in the plausibility of voxel-wise CVR and hemodynamic lag estimates. We demonstrate the clinical utility and feasibility of this protocol in an incidental finding of Moyamoya disease, and explore the possibilities and challenges of using this protocol in younger populations. This hybrid protocol has direct applications for CVR mapping in both research and clinical settings and wider applications for fMRI denoising and interpretation.

Lag-optimized BOLD cerebrovascular reactivity derived from breathing task data has a stronger relationship with baseline cerebral blood flow

“…We assessed the impact of methodological factors on the correlation between bCBF and BOLD-CVR. We present complimentary results to those in our previous manuscript (Stickland et al, 2021), suggesting that adding a simple breathing task to a resting-state BOLD fMRI scan, alongside lag-optimization of a P ET CO 2 regressor within CVR modelling, improves the relationship between BOLD-CVR and baseline CBF. We also report exploratory results of positive spatial correlations between bCBF and ALFF, mRSFA and fALFF, for each of the three resting-state segments.…”

“…Furthermore, the REST CDB data segment that was analyzed was preceded by 55 seconds of the subject breathing normally (the last 43 seconds of the ‘breathe normally’ task cue, see Figure 1, and the first 12 seconds of the REST CDB section not analyzed during CVR modelling). We previously found no clear evidence for obvious differences in common resting-state metrics (RSFA, ALFF, fALFF, LFCD) between all three resting-state data segments (Stickland et al, 2021), though further investigation of neural and vascular after-effects of breathing tasks may be warranted.…”

“…We have shown previously (Stickland et al, 2021) that adding breathing tasks to resting state fMRI paradigms makes CVR mapping more robust: if BOLD signal changes related to P ET CO 2 are small in amplitude, as may occur if natural fluctuations in P ET CO 2 are small during resting-state paradigms, it can be hard to distinguish them from other physiological, artefactual, or neuronally-driven fluctuations that occur at the same low frequencies. Our observations in this work suggest that BOLD-CVR estimates derived from resting-state data also show a much more variable relationship with bCBF.…”

“…The betaweight for the unshifted PETCO2 regressor, scaled by the fitted mean, produced CVR maps with no lag optimization (No-Opt CVR, units: %BOLD/mmHg). For CVR maps with lag optimization (Lag-Opt CVR), the model was run for each shifted PETCO2 regressor; parameter estimates were taken from the model with the largest full model R-squared (Moia et al, 2020;Stickland et al, 2021). For further details, refer to our previous work where we present the analysis of these CVR maps using the same subjects (Stickland et al, 2021) .…”

“…Therefore, we also predict that lag-optimization of BOLD-CVR estimates will lead to a stronger relationship between bCBF and BOLD-CVR. Note, this manuscript focuses on the relationship between CVR maps presented previously (Stickland et al, 2021) and newly presented bCBF maps.…”