ORIGINAL ARTICLE 
Annals of Nuclear Medicine Vol. 15, No. 2 111-116, 2001 

Arterial fraction of cerebral blood volume in humans measured by positron emission tomography 

Hiroshi ITO Iwao KANNO Hidehiro IIDA, Jun HATAZAWA, Eku SHIMOSEGAWA, Hajime TAMURA and Toshio OKUDERA 

Department of Radiology and Nuclear Medicine, Akita Research Institute of Brain and Blood Vessels 

In quantitative functional neuroimaging with positron emission tomography (PET) and magnetic resonance imaging (MRI), cerebral blood volume (CBV) and its three components, arterial, capillary, and venous blood volumes are important factors. The arterial fraction for systemic circulation of the whole body has been reported to be 20-30%, but there is no report of this fraction in the brain. In the present study, we estimated the arterial fraction of CBV with PET in the living human brain. C15O and dynamic H215O PET studies were performed in each of seven healthy subjects to determine the CBV and arterial blood volume (Va), respectively. A two-compartment model (influx: K1, efflux: k2) that takes Va into account was applied to describe the regional time-activity curve of dynamic H215O PET. Kl, k2 and Va were calculated by a non-linear least squares fitting procedure. The Va and CBV values were 0.011+-0.004 ml/ml and 0.031+-0.003 ml/ml (mean+-SD), respectively, for cerebral cortices. The arterial fraction of CBV was 37%. Considering the limited first-pass extraction fraction of H215O, the true arterial fraction of CBV is estimated to be about 30%. The estimated arterial fraction of CBV was quite similar to that of the systemic circulation, whereas it was greater than that (16% widely used for the measurement of cerebral metabolic rate of oxygen (CMRO2) using PET. The venous plus capillary fraction of CBV was 63-70% which is a important factor for the measurement of CMRO2 with MRI. 

Key words: cerebral blood volume, artery, vein, human, brain, PET 

INTRODUCTION 

QUANTITATIVE FUNCTIONAL NEUROIMAGING with positron emission tomography (PET) and magnetic resonance imaging (MRI) has been widely used to investigate the neurophysiology and the pathophysiology of cerebrovascular and neuropsychiatric diseases. For such quantitative imaging, cerebral blood volume (CBV) is an important factor. The CBV in a brain region is the sum of three components, i.e., the arterial, capillary, and venous blood volumes.1-3 The fractions of these component are needed 

Received September 27, 2000, revision accepted January 18,2001 . 

For reprint contact: Hiroshi Ito, M.D., Department of Radiology and Nuclear Medicine, Akita Research Institute of Brain and Blood Vessels, 6-lO Senshu-kubota-machi. Akita 010-0874, JAPAN. 
E-mail: hito@akita-noken.go.jp 

for quantitative functional imaging with PET and MRI, but there has been no report of these fractions in the brain. Since regional intravascular radioactivity in the brain contributes to the regional radioactivity measured by PET,4 the correction of intravascular radioactivity is needed in quantitative PET analysis.1 The CBV can be measured by 15O-labeled carbon monoxide (Cl5O) and PET.5 Nevertheless when a tracer that has a relatively high first-pass extraction fraction is used, the radioactivity concentration in the arterial blood differs from that in the capillary and venous blood.3 For such tracers, determination of the arterial fraction of CBV is necessary for correction of intravascular radioactivity. 2

Whereas the blood oxygenation level dependent (BOLD) contrast measured by functional magnetic resonance imaging (fMRI) has been used as an indicator of neuronal activity,6 the measurement of the cerebral metabolic rate of oxygen (CMRO2) by means of BOLD contrast has recently been reported.7 Since the concentration of oxygenated blood in venous and capillary blood is the main contributor to BOLD contrast,6,8 the venous plus capillary fraction of CBV is an important factor for the measurement of CMRO2 by means of BOLD contrast with fMRI.7-9 

15 O-labeled water (H2 15 O) is a tracer that can freely diffuse through the blood-brain barrier. After intravenous infusion of H2 15 O, the radioactivity concentration in the capillary and venous blood is same as that in the brain tissue 10 but it differs from that in the arterial blood. The radioactivity concentration in the arterial blood can be differentiated on a regional time-activity curve by kinetic analysis with a two-compartment model that takes into account the contribution of radioactivity from the arterial blood in a region of interest. Such a two-compartment model has been used to calculate the blood flow and arterial blood volume in myocardium using H2 15 O. ll Recently this two-compartment model has been applied to describe the regional time-activity curve of H2 15 O in the brain 3 With this model, the arterial blood volume in a brain region can be calculated from dynamic H2 15 O PET data. 

In the present study we estimated the arterial fraction of CBV by means of PET in the living human brain. Both C l5 O and H2 15 O PET studies were performed in each of seven healthy subjects. The CBV and arterial blood volume were determined from Cl 50 and H2 15 O PET studies, respectively. 

MATERIALS AND METHODS 

Theory 

The CBV (ml/ml) in a brain region is assumed to be the sum of the three components as follows. 1-3 

CBV = Va + Vc + Vv (Eq, l) 

where Va is the arterial blood volume (ml/ml), Vc, is the capillary blood volume (ml/ml), Vv is the venous blood volume (ml/ml). 

To describe the kinetics of H2 15 O in the brain, the twocompartment model that takes Va into account (Va model, Fig. 1 ) is used. According to this model, the radioactivity concentration in a brain region can be expressed as follows. 3-11 
 
where Cb(t) is the radioactivity concentration in a brain region, Ca(t) is the radioactivity concentration in the arterial whole blood (arterial input function), Kl is the influx rate constant (ml/ml/min) and k2 is the efflux rate constant (min-1). The influx rate constant Kl corresponds to the cerebral blood flow (CBF). The Kl/k2 ratio is defined as the distribution volume (Vd, ml/ml). 

Subjects 

The study was approved by the Ethics Committees of Akita Research Institute of Brain and Blood Vessels. Seven healthy volunteers including 4 men and 3 women (age range, 22-61 yr; average, 49.1 yr) were recruited and gave written informed consent. The subjects were determined to be healthy based on their medical history, physical examination, blood screening analysis and MRI of the brain. 
PET procedure 

The PET system used was Headtome V (Shimadzu Corp., Kyoto, Japan)12 which provides 47 sections with a center to center distance of 3.125 mm. The intrinsic spatial resolution was 4.0 mm full width at half maximum (FWHM) in-plane and 4.3 mm FWHM axially. With a Butterworth filter, the reconstructed in-plane resolution was approximately 8 mm FWHM. 

C15O study 

To measure the CBV, a C l5 O PET study was performed.5 The static PET scan was started 3 minutes after I minute of continuous inhalation of C l5 O gas (approximately 5 GBq by mouth). The scan time was 4 minutes. Three arterial blood samples were taken during PET scanning. The transmission scanning for attenuation correction was performed just before C l5 O PET scanning. 
H2 15 O study After the C l5 O PET study, 360-second dynamic scanning was performed following continuous intravenous infusion of H2 15 O over 2 minutes. The scan sequence consisted of six 5-sec frames, six 15-sec frames and eight 30-sec frames. The dose of radioactivity was 1.1 to 1.4 GBq at the start of the scanning. The arterial input function was obtained by continuous measurement of arterial whole blood radioactivity with a beta probe. Dispersion and delay occurring in the beta detector system and in the internal-arterial line were corrected according to the methods previously reported. 13,14 The dispersion was corrected by deconvolution with a single exponential function assuming the dispersion time constant to be 4 sec. The delay was corrected by means of a non-linear curve fitting to the time-activity data of the whole brain (gantry coincidence curve of PET scanner) with a singletissue compartment model that takes the delay into account. Two blood samples were taken at the beginning and at the end of scanning to measure the arterial CO2 gaseous pressure. A head fixation system with individual molds for each subject was used to minimize head movement over the period of the PET measurements. 

Regions of interest 

Regions of interest (ROls) were drawn on the PET images. ROls were defined for four neocortical regions representing the frontal, temporal, parietal, and occipital lobe. The ROls were elliptical in shape with a short axis of 16 mm and long axis of 32 mm. Each ROI was drawn in three adjacent sections, and data were pooled to obtain the average radioactivity concentration for the whole volume of interest. To obtain regional time-activity curves, regional radioactivity was calculated for each frame, corrected for decay and plotted against the time. 

Kinetic analysis 

The two-compartment model that takes Va into account (Va model. Fig. l) was used to describe the H215O time-activity curves for each region. The rate constants and Va were estimated by non-linear curve fitting to the regional time-activity curves in a least-squares manner.15 The model equation solved in the convolution integral procedure (Eq. 2) was used for this analysis. The two-compartment model assuming V* to be zero (NoVa model) which has widely been used to calculate CBF16,17 was also used to describe the time-activity curves. To compare the two models statistically, methods, the Akaike information criterion 18 and F-statisticsl9 were used. Values are shown as mean +-SD. 

RESULTS 

Typical time-activity data for a cerebral cortical region and fitted curves determined by both the Va and NoVa models are shown in Figure 2. The regional time-activity data were well described by both models. The results of kinetic analyses for H215O PET are given in Table 1 . The mean K1 values obtained from the NoVa model were significantly overestimated by 6% compared with those from the Va model. The mean Vd values in the two models were identical. The Akaike information criteria (AIC) in the Va model were significantly smaller than those in the NoVa model, indicating the Va model to be the preferred model. F-statistics also showed the Va model to be preferred for each time-activity curve. Good correlation was observed in Kl values was observed between the Va and NoVa models (Y = 0.93X + 0.06: X, Va model; Y, NoVa model; r = 0.98). The range of PaCO2 was 37.9 to 43.2 mm Hg for all subjects. 

The CBV, Va, and venous plus capillary blood volume (Vv + Vc = CBV - Va, Eq. l) values are given in Table 2. The average cerebral cortices CBV, Va and Vv + Vc were 0.031+-0.003 (ml/ml), 0.011+-0.004 (ml/ml) and 0.020+-0.002 (ml/ml), respectively. The arterial fraction of CBV (Va/CBV) was 37+-11% that of the cerebral cortical average. The venous plus capillary fraction (Vv + Vc/CBV) was 63+-11%. The CBV value of the occipital cortex was significantly higher and the CBV value of the frontal cortex was significantly lower than the CBV value in the other regions. There were no significant differences between the regions in Va Values. No correlation between CBV and Va was observed (Fig. 3). 

DISCUSSION 

Arterial fraction of cerebral blood volume 

This is the first study to estimate the arterial fraction of CBV in the living human brain. The arterial fraction of CBV was 37(~o for the cerebral cortices (Table 2). This fraction was slightly greater than the arterial blood volume fraction for the systemic circulation of the whole body, which has been reported to be 20-30%. 20 In the measurement of CMRO2 with a bolus inhalation of 15O-labeled molecular oxygen (15O2) with PET, the arterial fraction of CBV is necessary for correction of intravascular radioactivity since the first-pass extraction fraction of 15O2 is relatively high (about 40%).2 According to the literature, the arterial fraction of CBV has been assumed to be 16% for this correction 2 but this value was not for the brain. When an arterial fraction of 37% is used instead of 16%, the calculated CMRO2 will be lowered. No correlation between CBV and Va was observed (Fig. 3), indicating that V* might be independent of CBV, but variations in Va were larger than those of CBV since Va is sensitive to statistical noise. 

H215O is a tracer that can freely diffuse through the blood-brain barrier but, in fact, the capillary permeability-surface product of H215O is not infinite, indicating that the first-pass extraction fraction of H2150 (E) is less than l .21 ,22 Since radioactivity in the capillary and vein become higher than that in the brain tissue when E is less than 1 ,3 Va is greater than the true intravascular volume of the artery. The V* value obtained in the present study there-fore includes a part of the intravascular volume of the capillary and vein. The true arterial fraction of CBV (X) is expressed with the first-pass extraction fraction of H215O (E) and estimated arterial fraction (37%) as follows: 
X + (1 - X) ¥ (1 - E) = 0.37 
Assuming E to be 0.9,21,22 the true arterial fraction of CBV (X) will be 30%. 

In the present study the small- to large-vessel hematocrit ratio was assumed to be 0.85, in order to calculate the CBV according to the method reported,5 but this ratio has also been reported to be 0.69 as measured by PET in the human brain 23 When the hematocrit ratio of 0.69 is used instead of 0.85, the CBV will become greater. The hematocrit ratio might not be uniform in the brain, and individual variation might exist. These may also cause errors in the determination of CBV. In addition, changes in cerebral hematocrit with cerebrovascular disease have been reported.24 

Venous plus capillary fraction of cerebral blood volume 

The venous plus capillary fraction of CBV was 63% for the cerebral cortices (Table 2). Considering limited first-pass extraction fraction of H215O, the true venous plus capillary fraction of CBV is estimated to be about 70% as mentioned above. Since the capillary bed is smaller in volume than the noncapillary bed, this fraction represents mainly the venous fraction. 2,8 The measurement of CMRO2 using BOLD contrast with fMRI has recently been reported.7 Since the concentration of oxygenated blood in venous and capillary blood is the main contributor to BOLD contrast,6,8 the venous plus capillary fraction of CBV is a important factor for the measurement of CMRO2 using BOLD contrast with fMRI.7-9 The estimated value of venous plus capillary fraction of CBV in the present study can be used for such a measurement. 

Comparison of the V,* ond No V** models 

Although the regional time-activities were well described by both the Va and NoVa models (Fig. 2), the AICs in the Va model were significantly smaller than those in the NoVa model, indicating the Va model to be the preferred model.3 F-statistics also showed the Va model to be preferred. The Kl obtained from the NoVa model was significantly overestimated by 6% compared with K1 from the Va model, whereas the Vd was identical in the two models (Table 1 ) but a good correlation in K1 values was observed between the V* and the NoVa models. Since Va Was small, the overestimation of K1 for the NoVa model which has widely been used to calculate CBF16,17 would be small. 

CONCLUSION 

The arterial fraction of CBV was 37%. Considering a 
limited first-pass extraction fraction of H2 15O, the true arterial fraction of CBV is estimated to be about 30%. This fraction quite similar to that for the systemic circulation of the whole body but higher than that widely used for the correction of intravascular radioactivity in the measurement of CMRO2 with 15O2 and PET. The venous plus capillary fraction of CBV was 6_3-70% which is an important factor for the measurement of CMRO2 using BOLD contrast with fMRI. 

ACKNOWLEDGMENTS 

The assistance of the members of the Akita Research Institute of Brain and Blood Vessels in performing the PET experiments is gratefully acknowledged. 

REFERENCES 

l. Lammertsma AA, Jones T. Correction for the presence of intravascular oxygen-15 in the steady-state technique for measuring regional oxygen extraction ratio in the brain: 1 . Description of the method. J Cereb Blood Flow Metab 1983 ; 3: 416-424. 

2. Mintun MA, Raichle ME, Martin WR, Herscovitch P. Brain oxygen utilization measured with O-15 radiotracers and positron emission tomography. J Nucl Med 1984; 25: 177-l87. 

3. Ohta S, Meyer E, Fujita H, Reutens DC, Evans A, Gjedde A. Cerebral [15O]water clearance in humans determined by PET: I. Theory and normal values. J Cereb Blood Flow Metab 1996; 16: 765-780. 

4. Ito H, Hatazawa J, Murakami M, Miura S, Iida H, Bloomfield PM, et al. Aging effect on neutral amino acid transport at the blood-brain barrier measured with L-[2-18F]-fluoro-phenylalanine and PET. J Nucl Med 1 995; 36: 1232-1237. 

5. Martin WR, Powers WJ, Raichle ME. Cerebral blood volume measured with inhaled Cl5O and positron emission tomography. J Cereb Blood Flow Metab 1987; 7: 421-426. 

6. Jezzard P, Song AW. Technical foundations and pitfalls of clinical fMRl. Neuroimage 1996; 4: S63-75. 

7. Kim SG, Rostrup E, Larsson HB, Ogawa S, Paulson OB. Determination of relative CMRO2 from CBF and BOLD changes: Significant increase of oxygen consumption rate during visual stimulation. Magn Reson Med 1999; 41 : 1152-1161. 

8. Hathout GM, Gambhir SS, Gopi RK, Kirlew KA, Choi Y, So G, et al. A quantitative physiologic model of blood oxygenation for functional magnetic resonance imaging. Invest Radiol 1995; 30: 669-682. 

9. Davis TL, Kwong KK, Weisskoff RM, Rosen BR. Calibrated functional MRI: mapping the dynamics of oxidative metabolism. Proc Natl Acad Sci USA 1998; 95: 1834-1839. 

10. Herscovitch P, Raichle ME. What is the correct value for the 
brain-blood partition coefficient for water? J Cereb Blood 
Flow Metab 1985; 5: 65-69. 

11. Iida H. Rhodes CG, de Silva R, Yamamoto Y, Araujo LI, Maseri A, et al. Myocardial tissue fraction-Correction for partial volume effects and measure of tissue viability. J Nucl Med 1991 ; 32: 2169-2175.
 
12. Iida H. Miura S, Kanno I, Ogawa T, Uemura K. A new PET camera for noninvasive quantitation of physiological functional parametric images: Headtome-V-dual. In Quantification of Brain Function Using PET. Myers R, 
Cunningham V, Bailey D. Jones T (eds). San Diego; Academic Press, Inc., 1996: 57-61 . 

l3. Iida H, Kanno I, Miura S, Murakami M, Takahashi K, Uemura K. Error analysis of a quantitative cerebral blood flow measurement using H215O autoradiography and positron emission tomography, with respect to the dispersion of the input function. J Cereb Blood Flow Metab 1986; 6: 536-545.
14. Iida H. Higano S. Tomura N, Shishido F, Kanno I, Miura S, et al. Evaluation of regional differences of tracer appearance time in cerebral tissues using [150] water and dynamic positron emission tomography. J Cereb Blood Flow Metab 1988; 8: 285-288. 

15. Marquardt D. An algorithm for least-squares estimation of nonlinear parameters. J Soc Indust Appl Math 1963; 11 : 431-441. 

16. Herscovitch P, Markham J, Raichle ME. Brain blood flow measured with intravenous H215O. I. Theory and error analysis. J Nucl Med 1983; 24: 782-789. 

17. Kanno I, Iida H, Miura S, Murakami M, Takahashi K, Sasaki H, et al. A system for cerebral blood flow measurement using an H215O autoradiographic method and positron emission tomography. J Cereb Blood Flow Metab 1987; 7: l43-153 . 

18. Akaike H. A new look at the statistical model identification. IEEE Trans Automat Contr 1974; 19: 716-723. 

19. Hawkins RA, Phelps ME, Huang SC. Effects of temporal sampling, glucose metabolic rates, and disruptions of the blood-brain barrier on the FDG model with and without a vascular compartment: Studies in human brain tumors with PET. J Cereb Blood Flow Metab 1986; 6: 170-183. 

20. Mellander S. Johansson B. Control of resistance, exchange, and capacitance functions in the peripheral circulation. Pharmacol Rev 1968; 20: 117-196. 

21. Eichling JO, Raichle ME, Grubb RL Jr, Ter-Pogossian MM. Evidence of the limitations of water as a freely diffusible tracer in brain of the rhesus monkey. Circ Res 1974; 35: 358-364. 

22. Herscovitch P, Raichle ME, Kilbourn MR, Welch MJ. Positron emission tomographic measurement of cerebral blood flow and permeability-surface area product of water using [15O]water and [11C]butanol. J Cereb Blood Flow Metab 1987; 7: 527-542. 

23. Lammertsma AA, Brooks DJ, Beaney RP, Turton DR, Kensett MJ, Heather JD, et al. In vivo measurement of regional cerebral haematocrit using positron emission tomography. J Cereb Blood Flow Metab 1984; 4: 317-322. 

24. Yamauchi H, Fukuyama H, Nagahama Y, Katsumi Y, Okazawa H. Cerebral hematocrit decreases with hemodynamic compromise in carotid artery occlusion: a PET study. Stroke 1998; 29: 98-103.