Alberto L Vazquez¹, Mitsuhiro Fukuda¹, and Seong-Gi Kim^1
1Radiology, University of Pittsburgh, Pittsburgh, PA, United States

The accuracy of blood oxygenation data (i.e. BOLD fMRI) to calculate the changes in CMRO2 was examined. A systematic error is introduced in CMRO2 calculations when a fully oxygenated arterial input is assumed; however, this error did not affect the temporal estimates of CMRO2. The changes in tissue oxygen tension and blood oxygenation were in near equilibrium over a time-scale of 1 to 2 s, indicating that steady-state models are reasonable. CBV measurements are important for BOLD fMRI calibration in order to obtain reliable CMRO2 estimates.

Feng Xu¹, Peiying Liu¹, and Hanzhang Lu^1
1University of Texas Southwestern Medical Center, Dallas, TX, United States

Calibrated fMRI relies on an iso-CMRO2 challenge to obtain the calibration factor. O2 challenge has been proposed to be a potential method for the calibration experiment. However, the assumption that O2 challenge is iso-CMRO2 has not been validated. The present study will investigate whether physiologic manipulation of O2 content in the arterial blood will change brain metabolism. We used a recently developed TRUST MRI technique to monitor the subject’s CMRO2 while altering the O2 concentration in the inspired air. Our data suggest that O2 content changes brain metabolism and an inverse relationship was observed between CMRO2 and arterial O2 content.

Clarisse Ildiko Mark¹, and Gilbert Bruce Pike^1
1McConnell Brain Imaging Center, Montreal Neurological Institute, McGill University, Montreal, Quebec, Canada

Estimates of changes in cerebral metabolic rate of oxygen (ΔCMRO₂) and coupling relationship (n) to blood flow (ΔCBF) under neuronal activation, crucial in interpreting BOLD results, are highly sensitive to variability in individual subjects calibration (M)-values and brain regions. We thereby sought to acquire precise calibration data under robust control of hypercapnic (HC) and hyperoxic (HO) levels, together with visual stimulation and motor tasks in the same set of subjects. Based on low-variability M-values, our findings demonstrate significantly decreased variability in CMRO₂- and n-estimates under HC, and even better under HO.

Zachary Myles Smith¹, John S Hunt, Jr.¹, Ethan Li¹, Jia Guo¹, David D Shin¹, Richard B Buxton¹, and David J Dubowitz^1
1Radiology, University of California San Diego, La Jolla, CA, United States

We previously reported increased CMRO2 during sustained hypoxia, despite limited O2 availability. The biological rationale for this paradoxical response remains unclear. To investigate the possible influence of PaCO2 on CMRO2 and cerebral tissue oxygenation (PtO2), we made MRI measurements of CBF and oxygen extraction fraction during acute hypoxic conditions. Subjects’ CO2 levels were either unrestrained (“low CO2”), or clamped at normoxic levels (“high CO2”). Maintaining a “high CO2” partially mitigated the paradoxical increase in CMRO2, and also improved cerebral tissue oxygentation during acute hypoxic conditions. CO2 is thus an important covariable in the cerebral response to hypoxia.

Nicholas P Blockley¹, Valerie E M Griffeth¹, and Richard B Buxton^1
1Center for fMRI, Department of Radiology, University of California San Diego, La Jolla, California, United States

The calibrated BOLD technique typically uses a respiratory challenge to measure the calibration constant M. This is both time consuming and uncomfortable for the participant. Here we consider the possibility of determining M using measurements of R₂′ made using an asymmetric spin echo experiment without administering any gases. The key question in this hypothesis is what fraction of the R₂^* change on activation can be captured by R₂′ measurements in the baseline state?

Basavaraju G Sanganahalli^1,2, Peter Herman^1,2, Douglas L Rothman^2,3, Hal Blumenfeld^2,4, and Fahmeed Hyder^2,3
1Diagnostic Radiology, Yale University, New Haven, CT, United States, ²Quantitative Neuroscience with Magnetic Resonance in Medicine (QNMR), Yale University, New Haven, CT, United States, ³Diagnostic Radiology and Biomedical Engineering, Yale University, New Haven, CT, United States, ⁴Neurology, Neurosurgery, Neuroscience, Yale University, New Haven, CT, United States

Because oxidative demands in cortex and subcortex are largely unknown, we evaluated regional energetics with high field calibrated fMRI in rat brain. During somatosensory stimulation we measured BOLD, CBV, and CBF to calculate CMR_O2 in cortex and subcortex and compared these with neural recordings. We find that while neural-BOLD, neural-CBV, and neural-CBF relationships differ significantly between cortex and subcortex, CMR_O2 values are quite similar in these regions. These regional energetic estimates from calibrated fMRI are in agreement with neural recordings. Thus these results suggest that neurometabolic couplings are similar in cortex and subcortex, but neurovascular couplings are quite different.

Valerie Griffeth¹, and Richard Buxton^2
1Department of Bioengineering, UC San Diego, La Jolla, CA, United States, ²Department of Radiology, UC San Diego, La Jolla, CA, United States

We applied a calibrated-BOLD methodology to compare the effects of a simple flickering checkerboard stimulus to that of a complex movie stimulus on the coupling of CBF and CMRO₂ responses in the human visual cortex. We found no significant difference in the coupling between the movie stimulus and two flickering checkerboard contrast levels of 10% and 40%. Furthermore, the BOLD, CBF and CMRO₂ responses to these two contrast levels appear to bracket the physiologic response to the complex movie stimulus.

Hanzhang Lu¹, Feng Xu¹, Ksenija Grgac^2,3, Peiying Liu¹, Qin Qin^2,3, and Peter van Zijl^2,3
1Advanced Imaging Research Center, University of Texas Southwestern Medical Center, Dallas, TX, United States, ²Department of Radiology, Johns Hopkins University, Baltimore, MD, United States, ³F.M. Kirby Center, Kennedy Krieger Institute, Baltimore, MD, United States

Recently, a T2-Relaxation-Under-Spin-Tagging (TRUST) MRI technique was developed to quantitatively estimate blood oxygenation (Y) via the measurement of pure blood T2. However, a human validation study has not been conducted. Here we used in vitro blood experiments to determine a 3-dimentional calibration plot in which blood T2 is a function of both Y and Hct. We further showed that, during hypoxia, Ya measured with PulsOx was 84.0±3.6% (N=7), while Ya derived from TRUST blood T2 and in vitro calibration plot was 83.7±3.6%. Furthermore, a significant correlation was observed between these two measures across subjects (P=0.05).

Claudine Joëlle Gauthier^1,2, and Richard D Hoge^1,2
1Physiology/Biomedical Engineering, Université de Montréal, Montreal, Quebec, Canada, ²CRIUGM, Montreal, Quebec, Canada

Calibrated MRI techniques estimate changes in cerebral metabolic rate of O2 consumption (CMRO2) from BOLD task measurements. Different calibration techniques involve estimation of M, equivalent to the maximum possible BOLD signal change, by extrapolating from smaller changes obtained during hypercapnia or hyperoxia. We present a generalization of previous BOLD signal models which can be applied to data acquired during hypercapnia, hyperoxia, or both hypercapnia and hyperoxia simultaneously (HO-HC). We demonstrate the application of this generalized model during all three manipulations. While comparable group average results were achieved, the HO-HC method yielded more robust estimates of M and CMRO2.

Alberto L Vazquez¹, Mitsuhiro Fukuda¹, and Seong-Gi Kim^1
1Radiology, University of Pittsburgh, Pittsburgh, PA, United States

The temporal evolution of the changes in cellular oxidative metabolism, tissue oxygen tension and blood oxygenation were investigated using flavoprotein auto-fluorescence imaging (FAI), oxygen microelectrodes and deoxyhemoglobin-sensitive optical imaging of intrinsic signal (OIS, a BOLD fMRI surrogate), respectively. Fast increases in cellular oxidative metabolism were observed with increases in neural activity using FAI. The results showed that the increase in CMRO2 prompted the need for oxygen from the surrounding tissue and blood shortly after, indicating that, temporally, the changes in blood oxygenation closely reflect the average changes in tissue oxygen but less so the changes in cellular CMRO2.