Quantitative susceptibility mapping‐based cerebral metabolic rate of oxygen mapping with minimum local variance

    loading  Checking for direct PDF access through Ovid


Oxygen is critical for aerobic energy metabolism in cerebral tissue to sustain normal neural function. Noninvasively and quantitatively mapping of cerebral metabolic rate of oxygen (CMRO2) can provide valuable information on tissue viability and activity in both research and clinical settings 1. Cerebral metabolic rate of oxygen is the product of cerebral blood flow (CBF) and oxygen extraction fraction (OEF). Cerebral blood flow can be mapped using arterial spin labeling (ASL), but OEF requires determination of deoxyheme (dH) in the draining veins/venules for each voxel of tissue. Several MRI techniques have been proposed to map dH: (i) quantitative imaging of extraction of oxygen and tissue consumption, which uses a velocity‐selective spin labeling to selectively map venous blood T2 and oxygenation 2; (ii) calibrated fMRI, which models the magnitude JOURNAL/mrim/04.02/01445475-201801000-00017/math_17MM1/v/2017-12-21T175206Z/r/image-png decay of the blood oxygen–level dependent (BOLD) signal as a complex function of dH concentration ([dH]), typically estimating [dH] from signal measurements at two vascular challenges such as hyperoxia and hypercapnia to obtain CMRO23; and (iii) quantitative BOLD, which uses a specific venous geometry of randomly oriented tubes to model [dH] dependence of signal generated in asymmetric spin echo 6 or 3D multi‐echo gradient echo data 8. One limitation of these approaches is that only the signal magnitude is used for estimating [dH].
Signal phase can be used through quantitative susceptibility mapping (QSM) processing to generate tissue magnetic susceptibility 9. Unlike R2 and JOURNAL/mrim/04.02/01445475-201801000-00017/math_17MM2/v/2017-12-21T175206Z/r/image-png , the relationship between [dH] and tissue susceptibility is linear in its concentration and independent of imaging parameters 11. Hence, QSM can be used to quantify oxygenation in large veins 14 and in tissue 19. The [dH] in tissue can be determined from QSM by compensating for contribution from non‐heme iron, such as those stored in ferritin. A single vascular challenge to modify blood flow, either vasoconstrictive or vasodilative, has been used to separate the susceptibility contribution from blood and nonblood tissue 19. However, even the requirement of a single challenge may limit its practical utility.
In this study, a new algorithm called the minimal local variance (MLV) is proposed to obtain CMRO2 maps without blood flow challenge. Minimal local variance assumes the constant CMRO2 and nonblood tissue susceptibility ( JOURNAL/mrim/04.02/01445475-201801000-00017/math_17MM3/v/2017-12-21T175206Z/r/image-png ) within each tissue type (gray and white matter) across small regions of the brain. This added prior information mitigates the ill‐posed nature of the optimization problem. The results in healthy subjects are compared with those obtained using a caffeine challenge.

Related Topics

    loading  Loading Related Articles