Purpose: To investigate the mutual agreement of T2-based and susceptibility-based methods as well as their agreement with jugular catheterization, for quantifying venous oxygen saturation (Sv O2 ) at a broad range of brain oxygenation levels.
Methods: Sv O2 measurements using T2-relaxation-under-spin-tagging (TRUST) and susceptibility-based oximetry (SBO) were performed in 13 healthy subjects under room air, hypoxia, and hypercapnia conditions. Agreement between TRUST and SBO was quantitatively evaluated. In two of the subjects, TRUST and SBO were compared against the clinical gold standard, co-oximeter measurement via internal jugular vein catheterization.
Results: Absolute Sv O2 measurements using TRUST and SBO were highly correlated across a range of saturations from 45% to 84% (Pearson r = 0.91, P < .0001). Sv O2 -TRUST was significantly lower than Sv O2 -SBO under hypoxia and room air conditions, but the two were comparable under hypercapnia. TRUST demonstrated a larger Sv O2 increase under hypercapnia than SBO and had good agreement with jugular catheterization under hypercapnia but significantly underestimated Sv O2 under room air and hypoxia. The agreement between Sv O2 -SBO and the reference did not depend on the physiological state.
Conclusion: A systematic bias was observed between T2-based and susceptibility-based methods that depended on the oxygenation state. In vivo validation with jugular catheterization indicated potential underestimation of TRUST under room air and hypoxia conditions. Our findings suggested that caution should be employed in comparison of absolute Sv O2 measurements using either TRUST or SBO.
Keywords: T2-relaxation-under-spin-tagging; brain oxygenation; susceptometry-based oximetry; venous oxygen saturation.
© 2019 International Society for Magnetic Resonance in Medicine.