In this study we aim to investigate the sensitivity of streamlined-qBOLD¹ for detecting changes in baseline brain oxygenation. The ability to detect changes in baseline brain oxygenation is important for tracking disease progression over multiple time points. Hypocapnia (a reduction in blood CO₂) offers a convenient means to assess the ability of streamlined-qBOLD to detect changes in oxygen extraction fraction (OEF) in a healthy subject. Hypocapnia causes vasoconstriction leading to globally reduced CBF² resulting in an increased OEF. Here, streamlined-qBOLD is used to measure baseline brain oxygenation in these two conditions (normocapnia and hypocapnia) where OEF is expected to globally increase when moving from normo- to hypocapnia, compensating for vasoconstriction in order to support tissue oxygenation³.

Streamlined-qBOLD¹ is a recent refinement of the quantitative BOLD (qBOLD) technique⁴ providing a simplified approach to mapping baseline brain oxygenation. A FLAIR-GASE acquisition combines FLAIR (removal of CSF contamination), GESEPI (inherent reduction in magnetic field inhomgeneties) and ASE (direct measure of R₂′ independent of underlying R₂) to reduce confounding effects in the qBOLD model and simplify the analysis.Five healthy participants (aged 30-36; mean age 32.5 ± 2.7; 1 female) were scanned using a 3T Siemens Prisma system with a 32-channel receive coil. For each subject, FLAIR-GASE scans were acquired during normocapnia and hypocapnia. Imaging parameters were FOV 220 mm², 96x96 matrix, nine 5mm slabs (4 sub-slices), TR/TE = 3s/80ms, BW 2004 Hz/px, TI_FLAIR=1210 ms. ASE images were acquired with the following 𝜏 sampling scheme (𝜏_start:Δ𝜏:𝜏_finish=-16:8:64ms) for a total acquisition time of 4.5 mins. A prospective end-tidal gas targeting system (RespirAct^TM) was used to modulate end-tidal CO₂ between the normocapnic and hypocapnic conditions in a controlled and repeatable manner, whilst maintaining constant end-tidal O₂ (~110mmHg), Figure 1. For each FLAIR-GASE acquisition the qBOLD model was applied to quantify brain oxygenation. R₂′ was calculated using a log-linear fit to the mono-exponential regime⁵ (𝜏>15ms) of the ASE data. The intercept of this fit and the spin-echo signal (𝜏=0ms) were subtracted in order to provide a measure of Deoxygenated Blood Volume (DBV). OEF was then calculated using the following equation, $$OEF=\frac{R_2^\prime}{DBV\;\gamma\frac{4}{3}\pi\;\Delta\chi_0\;Hct\;B_0}\tag{1}$$ where DBV and R₂′ were measured and other parameters are known or assumed constants (Δ𝜒0=0.264x10^-6, Hct=0.34)¹. ROI analysis was performed for grey matter ROI’s in each participant defined by an automated segmentation of previously acquired T₁-weighted structural images (FAST, FSL) registered to the parameter maps.From Table 1, End Tidal (ET)-CO₂ was successfully decreased between the normo- and hypocapnic conditions for all participants. Table 1 displays R₂′, DBV and OEF for each subject, averaged over a grey matter ROI in the slice displayed in Figures 2-4. With the application of hypocapnia, significant (p<0.05) increases in grey matter R₂′ and OEF were detected across the group. No clear trend in DBV with hypocapnia was observed.Using the Fick principle and normocapnic values from Table 1, we can make predictions of the changes in OEF, DBV and R₂′ due to a hypocapnia induced reduction in CBF. CBF has been reported to decrease by 2-5% per mmHg decrease in CO₂^3,6, which equates to a 16-40% reduction for an 8mmHg decrease in CO₂. Assuming oxygen metabolism (CMRO₂) remains constant between normocapnia and hypocapnia, we predict an increase in OEF from 27% at normocapnia to 32-45% during hypocapnia. Assuming DBV is predominantly venous in origin and that changes in DBV are related to CBF by a power-law⁷ (exponent α=0.2)⁸, we predict a 3-10% decrease in DBV from normocapnic levels. Since R₂′ is a function of the product of these quantities, we predict an increase in R₂′ from 3.0 s^-1 at normocapnia to 3.5-4.5 s^-1 during hypocapnia. Significant increases in R₂′ and OEF with hypocapnia were detected, but changes in DBV were insignificant (Table 1). Changes in R₂′ and OEF were lower than predicted, whilst the predicted small change in DBV is likely beyond the sensitivity of this method. However, these predictions were based on the assumption that CMRO₂ is constant between normocapnia and hypocapnia, which may not be the case^2,3,9. Furthermore, CBF is not directly measured in this study (being inferred from ET-CO₂) and physiological responses to an 8mmHg decrease in ET-CO₂ may vary across the group (Table 1). Streamlined-qBOLD is sensitive to detecting changes in brain oxygenation (increases in R₂′ and OEF) caused by reduced CBF in the healthy brain. This suggests the technique could be used to track metabolic changes that underlie the progression of vascular diseases such as ischaemic stroke. However, further work is required to assess the accuracy, reproducibility and repeatability of this technique.