Whole-brain BOLD responses to graded hypoxic challenges at 7 T, 9.4 T, and 15.2 T: Implications for ultrahigh-field functional and dynamic susceptibility contrast MRI

Abstract

Purpose

Blood oxygen–level dependent (BOLD) functional MRI signals depend on changes in deoxyhemoglobin content, which is associated with baseline cerebral blood volume (CBV) and blood oxygen saturation change. To accurately interpret activation-induced BOLD responses and quantify perfusion values by BOLD dynamic susceptibility contrast (BOLD-DSC) with transient hypoxia, it is critical to assess Δ$R_2^{*}$ values in tissue and blood across varying levels of hypoxia and magnetic field strengths (B₀).

Methods

Whole-brain BOLD responses were examined using 5-s graded hypoxic challenges with 10%, 5%, and 0% O₂ at ultrahigh field strengths of 7 T, 9.4 T, and 15.2 T. Both tissue and blood responses were analyzed for BOLD-DSC quantification.

Results

Substantial heterogeneity in hypoxia-induced Δ$R_2^{*}$ was observed among regions under different hypoxic doses and B₀. Nonlinear Δ$R_2^{*}$ responses with increasing field strength were observed, depending on hypoxic levels: 10% O₂ condition exhibited pronounced supralinear trends, whereas 0% and 5% O₂ conditions showed nearly linear dependencies. Blood arterial and venous ${∆R}_2^{*}$ responses showed a similar dependence as tissue. However, at 15.2 T, the venous signal saturated under 5% and 0% O₂ conditions. Quantitative CBV values obtained from BOLD-DSC data showed dependency on susceptibility effects, and higher B₀ and hypoxic severity resulted in slightly higher CBV, indicating that caution is needed when comparing quantitative CBV values derived from different experimental protocols. Normalizing regional CBV values to those of white matter effectively reduced the impact of varying susceptibility contrasts.

Conclusions

Our investigations provide biophysical insights into the BOLD contrast mechanism at ultrahigh fields, and address quantification issues in susceptibility-based CBV measurements.