Measurement of relative cerebral blood volume using BOLD contrast and mild hypoxic hypoxia Richard G. Wise a,b, ⁎ , Kyle T.S. Pattinson a,c , Daniel P. Bulte a , Richard Rogers a,c , Irene Tracey a,c , Paul M. Matthews a,d , Peter Jezzard a a Department of Clinical Neurology, Centre for Functional Magnetic Resonance Imaging of the Brain (FMRIB), John Radcliffe Hospital, University of Oxford, OX3 9DU Oxford, UK b Cardiff University Brain Research Imaging Centre (CUBRIC), School of Psychology, Cardiff University, Park Place, CF10 3AT Cardiff, UK c Nuffield Department of Anaesthetics, John Radcliffe Hospital, University of Oxford, OX3 9DU Oxford, UK d Department of Clinical Neurosciences, Imperial College London and GSK Clinical Imaging Centre, Hammersmith Hospital, W12 0NN London, UK Received 29 October 2009; revised 17 May 2010; accepted 18 June 2010 Abstract Relative cerebral blood volume (CBV) was estimated using a mild hypoxic challenge in humans, combined with BOLD contrast gradient- echo imaging at 3 T. Subjects breathed 16% inspired oxygen, eliciting mild arterial desaturation. The fractional BOLD signal change induced by mild hypoxia is expected to be proportional to CBV under conditions in which there are negligible changes in cerebral perfusion. By comparing the regional BOLD signal changes arising with the transition between normoxia and mild hypoxia, we calculated CBV ratios of 1.5±0.2 (mean±S.D.) for cortical gray matter to white matter and 1.0±0.3 for cortical gray matter to deep gray matter. © 2010 Elsevier Inc. All rights reserved. Keywords: BOLD; Hypoxia; Cerebral blood volume; fMRI 1. Introduction Cerebral blood volume (CBV) is an important physio- logical parameter in the clinical study of cerebral pathology and also in neuroscientific research into brain function. For example, angiogenesis of brain tumors is associated with elevated blood volume. Clinically, CBV is often estimated by dynamic susceptibility-weighted contrast enhanced MRI employing an exogenous paramagnetic agent such as Gd- DTPA [1–4]. A more recent and less invasive development for noninvasive CBV measurement is known as vascular space occupancy or VASO [5]. This measures blood volume changes by observing the signal from brain tissue after nulling the signal from blood. Brain tissue signal therefore decreases as blood volume increases with increased brain activity. VASO provides a complementary measure of brain activity to BOLD contrast and is a somewhat more quantitative measure insofar as it provides a single physiological parameter, CBV, one of the components contributing to BOLD contrast. Alternative image contrasts based on alterations in the oxygenation state of hemoglobin have also been used to assess relative CBV. Van Zijl et al. [6] reported CBV measurement in the cat brain using spin-echo signal changes induced by hypoxia. The same group also measured CBV from alterations in BOLD signal as a result of varying arterial carbon dioxide tensions [7]. More recently, in humans, Bulte et al. [8] used hyperoxia-induced BOLD contrast in a T2*- weighted acquisition to estimate CBV. Hyperoxia, assuming an unaltered rate of metabolic oxygen consumption, induces an increase in cerebral venous blood oxygen saturation and hence an increase in T2*. This image contrast is heavily weighted towards the venous side of the vasculature because hyperoxia has little effect on arterial blood oxygen saturation. Hypoxic hypoxia has been applied safely in human MRI studies of cerebral physiology. Rostrup et al. [9] have demonstrated the decrease of T2* with increasing deoxyhe- moglobin concentration during hypoxia in gray matter and Available online at www.sciencedirect.com Magnetic Resonance Imaging 28 (2010) 1129 – 1134 ⁎ Corresponding author. Cardiff University Brain Research Imaging Centre (CUBRIC), School of Psychology, Cardiff University, Park Place, CF10 3AT Cardiff, UK. Tel.: +44 2920 870358; fax: +44 2920 870339. E-mail address: wiserg@cardiff.ac.uk (R.G. Wise). 0730-725X/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.mri.2010.06.002