Measurement and modeling of xenon gas transfer in the human brain with 1H and hyperpolarized 129Xe MRI

Abstract

BackgroundThe feasibility of imaging hyperpolarized 129Xe dissolved in brain tissue following inhalation of xenon gas in the lungs has recently been demonstrated in humans. The image contrast in 129Xe brain MRI represents a combination of factors, including regional perfusion, polarization decay and gas transfer rate across the blood-brain barrier. PurposeTo investigate the repeatability of hyperpolarized 129Xe brain MRI in healthy normal individuals and to identify the dominant mechanisms of image contrast by assessing voxel-wise correlation between HP 129Xe brain MRI and models of 129Xe brain uptake derived from 1H arterial spin labeling (ASL) perfusion mapping. Materials and MethodsTo assess repeatability, 3 sets of hyperpolarized 129Xe brain images were acquired from 5 healthy volunteers. Quantitative maps of the human brain, including cerebral blood flow, volume and predicted xenon uptake, were derived from 1H arterial spin labeling and T2-weighted MRI. These maps were then spatially cross-correlated with hyperpolarized 129Xe brain MRI. ResultsSignal to noise ratios of 8.7–17.7 were observed across volunteers for a voxel size of 8 × 8 × 50 mm3 with intra-subject repeatability of between 6 and 29 %. Hyperpolarized 129Xe brain images showed voxel-wise correlations with cerebral blood flow (R = 0.32 to 0.62), volume (R = 0.33 to 0.63) and predicted xenon uptake (R = 0.34 to 0.63), but did not correlate with arterial transit time (R = 0.05 to 0.26). ConclusionVoxel-wise cross correlation between 129Xe and 1H ASL suggests that the regional quantity of dissolved xenon delivered by cerebral blood flow is the dominant mechanism of image contrast in HP 129Xe brain MRI, assuming normal blood-brain barrier function. Combining 1H and 129Xe brain MRI provides new opportunities to quantitatively investigate brain pathophysiology and function.