ValIdation of Tomographlc Measurement of Cerebral

Blood Volume with C-il-Labeled Carboxyhemoglobin

M. E. Phelps, S. C. Huang, E. J. Hoffman, and D. E. KuhI

University of California, Los Angeles, California

Red blood cells, tagged with C-il by administration of 11('Ogas, have been used toportray the distribution of blood in the brain. To date, however, the accuracy of thisapproach has not been validated. We have performed in vitro measurements of regionalcerebral blood volume (CBV) with red blood cells labeled with C-li and Cr-Si in fourdogs and two rhesus monkeys. These studies yielded a mtio of CBV@_11to CBV@r_5iof 1.02 ±0.03 (s.d.) from 92 samples. A least-squares fit to these data showedCBV@11 = 1.01 CBV@r_5i + 0.037; P@ 0.001. The ratio of CBV in gray matter tothat in white matter was 2.8 ±0.4 (n = 12) and 3.1 ±0.6 (n 8). In vivo studieswith emission computed tomography (ECT) and 11C0-RBC gave coefficients of vanation of ±2.8% and ±4.8% for cross-sectional CDV and regional (-4 cm 2) CBVover an 80-mm period. The average human CBV wasfound to be 4.2 ±0.4 cc bloodper 100 g tissue. Clear tomographic delineation of the distribution of CBV in humansubjects is achieved with ECT, which provides a “live―measurement ofthis parameterof cerebral hemodynamics. These data demonstrate that 1I1@Oadministered by singlebreath inhalation is a reliable and accurate blood tracerfor measurement ofCBV withECT.

J Nucl Med 20: 328—3M, 1979

Cerebral blood volume (CBV) is an importantfactor in the regulation and control of cerebral hemodynamics, and has been characterized in termsof normal values (1—9),its role in autoregulation (6,10—12)and changes with PaCO2 and cerebral bloodflow (2, 3, 5, 6, 10, 13). Changes in CBV have alsobeen observed during generalized seizure (14), mental activity (15), and sleep (16). The tomographicstudy of CBV with emission computed tomography(ECT) has been used to demonstrate ischemia inand vasodilation around infarcted tissue (17—19);toassess alterations in CBV with head trauma (19),vasodilation in certain tumors (6, 17) and in regionssurrounding subdural hematomas (19, 20); and tomonitor reduction of CBV in cerebral edema followed by restoration with steroid therapy (6).

Received Sept. 29, 1978; revision accepted Nov. 21, 1978.

For reprints contact: Michael Phelps, Div. of Nuclear Mcdicine, Dept. of Radiology, School of Medicine, University ofCalifornia, Los Angeles, Los Angeles, CA 90024.

The quantitative tomographic measurement ofcerebral blood volume (CBV) was first performedby Kuhi et al. (6). These authors employed Tc-99m-labeled red blood cells (RBC) and a model developed by Phelps et al. (21) for x-ray fluorescencemeasurements of regional CBV. Phelps et al.,Brownell et al. , Muehllehner et al. , and Ter-Pogossian et al. have used single-breath inhalation oflico to label carboxyhemoglobin(“CORBC)andthus image the distribution of CBV (22). Glass etal. (23) have used ‘@COto measure the whole-bodyblood volume with a reported accuracy of ±3%when rated against Cr-S 1 RBC. To date, however,no one has validated the tomographic measurementof CBV with 1'CO-RBC.

In the work reported here, in vitro studies werecarried out in dogs and rhesus monkeys to compareregional CBV as measured by 11f@@RBC and Cr-S 1RBC. Studies were also carried out in human subjects, following a single-breath inhalation of 11@@c@

328 THE JOURNAL OF NUCLEAR MEDICINE

by on April 5, 2019. For personal use only. jnm.snmjournals.org Downloaded from

BASIC SCIENCESINSTRUMENTATION

gas, to determine the accuracy and reproducibilityof the in vivo tomographic measurement of CBV.

MATERIAL AND METHODS

Labeling of red blood cells. In the animal experiments, red blood cells were labeled in vitro withCr-Si and in vivo with 11C@OThe Cr-Si labelingwas carried out by withdrawing 20 ml of venousblood into a syringe containing 2 ml of acid citratedextrose (ACD) solution and transferring this to asterile evacuated bottle containing 75 @Ciof [5'Cr}sodium chromate. This mixture was incubated forabout 30 mm with occasional swirling. Ascorbicacid, SOmg in a 1-mi volume, was added to reducethe sodium chromate to Cr3@. The solution wascentrifuged, plasma discarded, and the red bloodcells washed twice with saline. A 3-cc sample wastaken for the measurement of unbound Cr-S 1, andthe remainder was injected into the animal to measure CBV.

In the animal and human studies, red blood cellswere labeled in vivo with “COby administering anadmixture of “COand oxygen gases. The 11(@Owasprepared by the “N(p,a)―Creaction as describedelsewhere (24) and was >99% pure. In the animalstudies, 11(@Owas administered through a respiratorthat was used for controlled ventilation of the animals. In the human subjects, the individuals wereallowed to take a deep breath from an embo bagcontaining 11C0 and 02. The residual “COwasvented back into the bag and the subject returnedto normal breathing of room air.

Animal experimental procedure. Four mongreldogs weighing from 15 to 25 kg, and two rhesusmonkeys weighing from seven to nine kg, wereanesthetized with pentobarbital sodium (25 mg/Kg),paralyzed with gallium triethiodide, and passivelyventilated on 100% oxygen with a Harvard respirator. Catheters were placed in the left and rightsaphonous leg veins for the injection of Cr-Si RBCand for removal of blood samples for counting andthe measurement of hematocrit. A femoral-arterycatheter was used to monitor arterial blood pressureand for the withdrawal of blood samples for themeasurement of blood pH, PaCO@,and 1@a°2.Theend-tidal PCO2 was continuously monitored withcapnograph*. Animals were given supplementalanesthesia as needed.

While breathing a mixture of “COand 02, theanimals were injected with Cr-SI RBCs. Ten mmutes were allowed for equilibration, five 6-cc bloodsamples were taken, and the animal was then killedby a rapid injection of concentrated KCI into theheart. The brain was rapidly excised and tissueseparated into 2- to 4-g samples that were placed inweighed counting vials.

Three of the blood samples were also transferredto weighed counting vials. The tissue and bloodsamples were immediately counted for C- I I activityin a Nal(Tl) well counter. The vials were reweighedand the C-l 1 counting rate per gram of tissue (cpm/g) computed and decay-corrected to a commontime. After the C-l 1 had decayed away, the sampleswere counted again to determine cpm/g for Cr-SI.A spectrometer setting for the SI l-keV radiationfrom C-l 1 was used to discriminate against the 320-keV Cr-Si gamma. No special effort was made tosection tissue into the different anatomic components of the brain except for the samples of superficial cortex and subcortical white matter.

About i @Ciof 1-13I-labeled serum albumin(RISA) was added to two of the blood samplestaken after injection of the labeled RBCs, and alsoto a portion of the original Cr-S 1 RBC saline solution that was used for injection. These sampleswere centrifuged, RBCs and plasma separated, andeach fraction counted for 1-131, Cr-S 1, and C-l Iwith a Ge semiconductor detector to determine thedistribution of C-l 1 and Cr-S 1 between RBCs andplasma. The RISA served to determine the amountof plasma trapped in the RBC fraction. These studies showed that less than 0.001% of the Cr-S 1 activity was in the saline phase of the Cr-S 1 RBCsaline solution injected in the animals, and that lessthan 0.001% of the blood Cr-Sl activity was in theplasma fraction at the time the animal was killed.The fraction of C- I I activity in the plasma wasfound to be less than 0.03% (value from ten samples; two samples had fractions of 1. I and 0.9% butwere disregarded because significant hemolysis wasvisually apparent).

Human experimental proceduret. Five male volunteers between the ages of 22 and 27 were administered 10—ISmCi of “COby single-breath inhalation and tomographic studies were carried out witha positron transaxial tomographt (25). Approximately 4 mm were allowed for equilibration of“CO-RBCin the body blood pool. In two subjectsa single tomographic plane was repeatedly imagedover a period of 80 mm. The initial image was recorded for 2.0 mm while the subsequent imagetimes were progressively increased in proportion tothe amount of C- I 1 decay (decay-compensationscanning mode) in order to accumulate an equalnumber of counts in each image. A 2-cc venousblood sample was taken at the mid time of eachscan. These samples were transferred to weighedcounting vials, counted in a well counter, reweighed, corrected for radioactive decay, and conveiled to cpm/g.

The tomograph was calibrated with each studyby measuring the number of counts/unit time from

Volume 20, Number 4 329

by on April 5, 2019. For personal use only. jnm.snmjournals.org Downloaded from

PHELPS, HUANG, HOFFMAN, AND KUHL

formed with the tomograph to display the generalbody distribution of blood. In these studies, threeviews of the subject (anteroposterior and two 60@obliques) are simultaneously recorded (25).

CBV model. The model for the measurement ofregional cerebral blood volume (rCBV) is given inRefs. 6 and 21.

rCBV (cc per 100 g) = CBXO.85x d' (2)

where CT,@is the blood tracer tissue concentration(cpm/cc) in region i, as determined by tissue counting in the in vitro experiments or as measured tomographically in the in vivo experiments. CB is thevenous blood concentration (cpm/cc) of the bloodtracer, and d is the density of cerebral tissue (=

3@ 1.04g/cc).The factor0.85correctsfor the difference between the large-vessel and cerebral hematocrits (21), this factor being the average of thevalues from Everett et al. (26) in the rat, Larsen etal. (27) in man, Oldendorf et al. (28) in man, Sklaret al. (29) in the cat, and Studer et al. (30) in the ratnamely, 0.85 ±0.06 (1 s.d.). This value has beenused routinely by Phelps et al. (3, 10, 21), Grubb etal. (4, 5, 11, 12), Kuhl et al. (6, 19), Eichling et al.(31), and Gado et al. (14).

The model responsible for Eq. 2 was used forboth in vitro animal and in vivo human studies (inthe animal experiments d is dropped from Eq. 2since there CT,@is measured in units of cpm/g). Inthe animal studies, rCBV was measured with Cr-SiRBC and “CO-RBCto validate the use of 11(@Ofor

•1

6()

L)

I •

•S.<@ SS.. U

IU@ U•

.S ,••

,2 ••U •S.

U@.

/1U..

S

555 U

CBV (cc/IOOgms)51Cr-RBC

FIG. 1. Correlation of regional cerebral blood volumesmeasured in vitro with both “CO-RBCand Cr-51 ABC in fourdogs and two rhesus monkeys. These data exhibit one-toone correlation as described by Eq.3. Eachsampleweighedfrom 2 to 4 g and was randomlyexcised throughoutthe brainwithout regard to anatomical identity,except for the 20 sampIes representing carefully separated superficial cortex andsubcortical white matter. Circles and squares are from dogand monkeydata, respectively.Solid line is from Eq.3.

a uniform cylinder of positron activity (Ga-68). Analiquot of activity from the cylinder was alsoweighed and counted in the well counter to determine its cpm/g. The region-of-interest capabilitiesof the tomograph were then used to determine thecpm/cm2 from the image of the uniform cylinder.These results were used to determine a calibrationfactor, f, between the tomograph and the wellcounter.

f = cpm/cm2 (tomog.)cpm/cc (well c.)

The correction factor, human “CO-RBCbrainimages, and blood data allow the calculation of theregional CBV in units of cc blood per gram of tissue(see Eq. 2).

The medium-resolution mode (25) of the tomograph (1.3 cm FWHM) was used in all the studiesreported in this work. The total number of countsin each image ranged from 600,000 to 900,000.

The procedure described above was also used inthree other human subjects except that eight crosssectional images from 7 cm above to 1 cm belowthe orbitomeatal plane were taken to image the distribution of CBV in the brain. These data were alsoused to calculate the average whole-brain CBV.Two-dimensional whole-body scans were also per

6CR

uc!.@DIV O@HUCLN,O

S SSL

(i)

FIG. 2. Whole-body projections of blood distribution inhuman subject showing anteroposterior and two obliqueviews (± 60@apart). Ten mCi of 1I@Owere administered bysingle-breath inhalation; total scan time for all three viewswas 20 mm. Two-dimensional images show distribution ofblood volume in body but have limiteddetail in any particularorgan compared with tomographic images (Figs. 3, 5, and6).

330 THE JOURNAL OF NUCLEAR MEDICINE

by on April 5, 2019. For personal use only. jnm.snmjournals.org Downloaded from

BASIC SCIENCES

images over an 80-mm period in each individual.The average coefficient of variation over the 80-mmtime for eight different regions of interest (each-.-.4cmZ) in the two individuals was ±4.5%.

The decay-corrected C-I I blood concentrationwas found to decrease with time. After the 3-mmequilibration, the average biologic half-time for fourhuman subjects was 195 ±20 mm.

The CBV at cross-sectional levels of the brain,from 7 cm above to 1 cm below the orbitomeatalline (Fig. 5), was calculated for three human subjects to determine the average total cerebral bloodvolume, which was found to be 4.2 ±0.4 cc/lOO g.This value includes the large-vessel and capillaryblood. The average venous PCO2 in these subjectswas 42 ±2 mm Hg.

DISCUSSION

The data in Fig. I , Eq. 3, and Table I show aone-to-one correlation between CBV as measuredwith Cr-Si RBC administered intravenously andCBV from the administration of “COby singlebreath inhalation. These data show that the “CORBC method is reliable and accurate for the measurement of CBV.

INSTRUMENTATION

? I? I@'I@ I @‘I *@!3mm 12 22 38 58 78

FIG. 3. Single crosssection of head showing distribution of cerebral blood volume (CBV) with passageof time. Anterior istop, left is left, and level is 4 cm above orbitomeatal plane. Major features are high CBV in region of Sylvian fissure, straightsinus, superior sagittal sinus, and cortex. Numbersare mid-scantime in minutes. Imageswere selectedfrom a total of tentaken over 80 mm. Irregular dashed line shown at left is region of interest (AOl) selected for analysisof regional CBV. Notereproducibility and consistencyof images(seeFig. 4).

the measure of CBV. In the human studies, rCBVand cross-sectional CBV values were repeatedlymeasured as functions of time to determine the reproducibility of the in vivo measurements in man.

RESULTS

Animal studies. The results for rCBV as measuredwith 11CO-RBC are plotted against those measuredwith Cr-Si RBC in Fig. 1. These data were fittedwith a linear least-squares analysis to yield theequation:

y = 1.Olx + 0.037(P 40.001), (3)

where y and x are the rCBV values measured with‘@CO-RBCand 5'Cr-RBC, respectively. Table 1summarizes the data from the individual animal cxperiments.

Figure 1 includes the data from samples considered to be anatomically isolated superficial cortexand subcortical white matter, and these results wereanalyzed separately to determine the CBV ratio ofgray to white matter. The ratios were determinedfor each individual experiment to avoid differencesproduced by varying numbers of samples from individual animals that had different PaCO2 levels andtherefore different values of cerebral blood flowand CBV. The average ratios of gray matter CBVto white matter CBV were 2.8 ±0.4 (n = 12) and3.1 ±0.6 (n = 8).

Human studies. The general whole-body blooddistribution is shown in the two-dimensional imagesof Fig. 2, whereas the cross-sectional distributionsof CBV are shown in the tomographic images ofFigs. 3, 5, and 6.

The reproducibility of the in vivo tomographicmeasure of CBV in a human subject is shown inFig. 3. The same cross section was repeatedly imaged during 80 mm. The numerical data for theaverage cross sectional CBV and a selected regionof interest (Fig. 3, 3-mm image) are plotted as afunction of time in Fig. 4. The values for the average cross-sectional CBV in two individuals had acoefficient of variation of ±2.8% (1 s.d.) for 10

U ROZ on sylvia,? fissure

0 Whole cross section

.. .•. . . .

.000 00 0 0@

8

6

4

2

0

E0'

C-)U

C

>

0

_-_o 10 20 30 40 50 60 70 80

MIN. AFTER INJECTION

FIG. 4. Numerical values for CBV (from Eq. 2) in AOl shownin Fig. 3 (area of sylvian fissure), and in whole cross section,plotted as a function of time. Reproducibilities in humanstudies were ±2.8% and ±4.8%, respectively, for wholecross section and for AOl (@4 sq. cm).

Volume 20, Number 4 331

by on April 5, 2019. For personal use only. jnm.snmjournals.org Downloaded from

PHELPS, HUANG, HOFFMAN, AND KUHL

estimated by about 5% , which is consistant withsome retention in intravascular hemoglobin andmyoglobin. This percentage, however, is considerably lower than would be predicted from the studiesof Roughton and Root (32), and is lower than themeasured values of Wennesland et al. (33) usingrelatively large masses of CO. Our value of 195 ±20 mm is in good agreement with the biologic halftime of 176 ±SOreported by Glass et al. (23). Theeffect of the falling blood “COconcentration isremoved in the CBV studies by taking several bloodsamples during the study (i.e., one during midtimeofeach scan).

Our values of 2.8 ±0.4 and 3.1 ±0.6, found indog and monkey for the ratios of gray matter towhite matter CBV, are of the same order as theratio of gray to white cerebral blood flow of 3.8(34)and capillary density of about 3 (35). Our in vitroratios are considerably higher than the I .4 and 1.8found, respectively, by Greenberg et al. (13) andKuhl et ai. (19) in man using emission computedtomography. Because of spatial averaging in ECT,however, it would be expected that our in vitrovalues would be larger than those of the groupsmentioned.

The average whole-brain CBV found in this work(4.2 ±0.3 cc/I00 g) is well within the range ofvalues reported in the literature (Table 2). In spiteof the diverse techniques used to measure CBV,we believe that the average basal CBV is about 4cc/lOO g.

The model for CBV is incorporated directly intotomograph software in such a manner that after theblood samples are counted and entered into thesystem, the images are converted to units of ccblood/g tissue. Our tomograph's region-of-interest(ROl) capability allows convenient extraction of

FIG. 6. Illustrationof AOl extraction of regional CBV. Modelrepresented in Eq. 2 is incorporated directly into software oftomograph, and once blood data are entered, regional valuesof CBV can be extracted for any region of interest (AOl)controlled by joy stick. Tomograph display provides crosssectional image of CBV, selected AOl (regions outlined),numerical value of CBV, coefficient of variation within AOl,and cross-sectional area of AOl. As AOl is moved aroundimage, these parameters are updated in real time.

FIG. 5. Cerebral blood-volume distributions by tomographin cross sections from human subject. Levels are from 7 cmabove, to the orbitomeatal plane in one-cm increments, andproceed from left to right and top to bottom.

It was originally felt that because a small amountof CO is dissolved in plasma and is in equilibriumwith red blood cells, CBV would be overestimateddue to equilibration between plasma CO and tissue.This, however, was not found experimentally to bethe case (Fig. I , Eq. 3, Table 1). This is supportedby the fact that the plasma fraction of C-l 1 activitywas measured to be less than 0.03%, and if “COwas equilibrated with the extravascular space ( ‘@25times the vascular) this would still only be about1% of the “CO-RBCactivity. The value of 1% issimilar to the average overestimation of CBV byabout 2% when “COwas compared with Cr-S 1RBC (Table I). However, the average 2% overestimation shown in Table I is not statistically significant. Glass et al. (23) reported that “COoverestimated whole-body blood volume by 5% because ofbinding to myoglobin. Since myoglobin does notoccur in significant quantities in the brain, the errorwould be inconsequential.

The decrease in blood C- 11 activity with time(biologic t@= 195 ±20 mm) is probably due to theslow equilibration with the low-flow compartmentsof the body, to slow expiration of “CO.and tosome extraction by extravascular hemoglobin andmyoglobin. Roughton and Root (32) report that nomore than 60—70%of the CO that disappears fromblood in the first hour is found in expired air. Theremainder of the CO is reported by Wennesland etal. (33) to be extracted into the hemoglobin of redskeletal muscle and myoglobin of heart muscle.However, these studies were carried out at toxiclevels (32) or at least with significant masses of CO(33), as opposed to the tracer levels of “COusedin our work, and therefore direct comparisons areof limited value. Glass et al. (23) have found, using“CO.that the whole-body blood volume is over

332 THE JOURNAL OF NUCLEAR MEDICINE

by on April 5, 2019. For personal use only. jnm.snmjournals.org Downloaded from

BASIC SCIENCES

No. ofCBV(“CO-ABC)

CBV(“Cr-ABQ)P@CO,Animalsamples(average ±1 s.d.)(mm Hg) pH

AverageCBVAuthorsAnimal (cc/100 g)

INSTRUMENTATION

puted tomography and Tc-99m RBC has been re

ported by KuhI et al. (6). Note, further, that themeasurement of CBV is essentially a ‘‘live'‘rneasurement, since observed temporal changes in CBVcan produce concomitant changes in the tissue concentration of the blood tracer (i.e. , measured CBV).Thus sequential tomograms of CBV from a singleadministration of “COallow one to observe andmeasure the response of this variable to induced ornaturally occurring changes in the hemodynamicsofthe brain.

FOOTNOTES

* Beckman LB-50, Beckman Instruments, Fullerton, CA

t All human studies were carried out under approval of theUCLA School of Medicine human use committee.

:@ ECAT,― ORTEC, Inc. , Oak Ridge, TN

ACKNOWLEDGMENTS

This work was partially supported by DOE Contract EY-76-C-03-00l2 Gen-12 and NIH Grant 7Rol-GM 24839-01.

REFERENCES

I. NYLIN G, HEDLUNDS, REGNSTROM0: Studies of the cerebral circulation with labeled erythrocytes in healthy man.Circ Res 9: 664—674, 1961

2. SMITH AL, NEUFELDGR, OMINSKYAJ, et al: Effect ofarterial CO, tension on cerebral blood flow, mean transittime, and vascular volume. J Appl Physiol 31: 701—707,1971

3. PHELPS ME, GRUBB RL, TER-POGOSSIAN MM: Correlationbetween Parn, and regional cerebral blood volume by X-rayfluorescence. J App! Phvsio! 35: 274—280, 1973

4. GRUBBRL, PHELPSME, TER-POGOSSIANM: Regional cerebral blood volume in humans. Arch. Neurol. 28: 38—44,1973

5. GRUBB RL, RAICHLE ME, E1CHLING JO, et al: The effectsof changes in PaCO, on cerebral blood volume, blood flow,and vascular mean transit time. Stroke 5: 630—639,1974

6. KUHL DE, REIVICHM, ALAvI A, et al: Local cerebral bloodvolume determined by three-dimensional reconstruction ofradionuclide scan data. Circ Res 36: 610—619,1975

7. MATHEW NJ, MEYER JS, BELL RL, et al: Regional cerebralblood flow and blood volume with the gamma camera. Neuroradiology4: 133—140,1972

8. PENNRD, WALSERR, ACKERMANL: Cerebral blood volumein man. Computer analysis of computerized brain scan.JAMA 234:1154—1155,1975

9. LADURNER0, ZILKHA E, IFIFF LD, et al: Measurementofregional cerebral blood volume by computerized axial tomography. J Neuro! Neurosurg Psychiatry 39: 152—158,1976

10. PHELPSME, GRUBBRL, RAICHLEME, et al: Correlation ofregional cerebral blood volume with PaCO, and arterialblood pressure. In: Cerebra! Circu!ation and Metabolism.Langfitt T, McHenry L, Reivich M, et al, eds. New York,Springer-Verlag, 1975, pp 24—26

11. GRUBsRL, PHELPSME, RAICHLEME, et al: The effects ofarterial blood pressure on the regional cerebral blood volumeby x-ray fluorescence. Stroke 4: 390—399,1973

12. GRUBB RL, RAICHLE ME, PHELPS ME, et al: Effects ofincreased intercranial pressure on cerebral blood volume,blood flow, and oxygen utilization in monkeys. J Neurosurg43:385—398,1975

TABLE 1. COMPARISONS OF CEREBRAL BLOODVOLUME BY “CO-RBCAND Cr-51 RBC

DogDogDogDogMonkeyMonkeyAve ±

13 0.9818 1.0213 1.0514 1.0422 1.0212 0.9992 1.02

±0.07±0.08±0.04±0.05±0.08±0.07±0.03

38 7.4642 7.4370 7.2136 7.4749 7.3838 7.45

I s.d.

Average hematocrits of dogs and monkeys were 43.1±1.2 and 39.9 ±2.1.

TABLE 2. IN VIVOVALUESOF CEREBRALBLOODVOLUME (CBV)* IN THE NORMAL BRAIN

Grubb et al. (4)Phelpsetal.(3)Smith et al. (2)Grubb et al. (10);

Eichling et al.(31)

KuhI et al. (6)Greenberg et al.(11)Mathewet al. (7)Pennet al. (8)Ladurner et al. (9)

ManMonkeyGoatMonkey

3.25.4 ±0.6t4.8 (4.1)t3.6 ±0.5

Baboon 3.3±0.7Human 4.3

Human 4.2Human 3.0Human 5.7 (cortex)

5.1 (thalamus)Human 4.34 ±0.50°Human 4.2 ±0.4

KuhI et al. (19)This work

* CBV values at basal PCO2 of about 36-43 mm Hg.

Average values for whole brain unless otherwisenoted.

t Regional value in frontal lobe.@ Value in parenthesis is the authors' value cor

rected for the cerebral hematocrit which is 0.85 oflarge-vesselvalue (21).

“Averagevalue from cross-sectional levels fromabout 2 to 5 cm above orbitomeatal plane.

CBV values, as shown in Fig. 6. As the ROI ismoved around the image, the area of the ROI, theCBV, and coefficient of variation within the ROIare updated in real time.

The whole-body images (Fig. 2) provide an overall perspective of the body's blood distribution, butthey lack the spatial detail and quantitative featuresof the tomographic approach.

The quantitative regional tomographic measurement of CBV in human subjects is seen to be reproducible (Figs. 3 and 4) and allows clear delineation of the distribution of cerebral blood (Fig. 5)by a rioninvasive technique. Similar reproducibilityin the measurement of CBV with emission corn

Volume 20, Number 4 333

by on April 5, 2019. For personal use only. jnm.snmjournals.org Downloaded from

PHELPS, HUANG, HOFFMAN, AND KUHL

13. GREENBERGJ, ALAVI A, REIVICHM, et al: Local cerebralblood volume response to carbon dioxide in man. Circ Res43:324—331,1978

14. GADO MH, PHELPSME, HOFFMANEJ, et al: Changes incerebral blood volume and vascular mean transit time duringinduced cerebral seizures. Radiology 121: 105—109,1976

15. RISBERG J, INGVAR DH: Regional changes in cerebral bloodvolume during mental activity. E.xpt Brain Res 5: 72—78,1968

16. RISBERGJ, GusTAvssoN K. INGVARDH: Regional cerebralblood volume during paridoxical sleep. In: Cerebral blood

flow: Clinical and experimental results. Brock M, Fieshi C,

Ingvar DH et al, eds. Springer-Verlag, Berlin, 1969, pp 101—103

17. PHELPSME, HOFFMANEJ, COLEMANRE, et al: Tomographic images of blood pool perfusion in brain and heart.J Nuc! Med 17: 603—612, 1976

18. ACKERMAN RH, CORREIA JA, ALPERT NM, et al: Positronimaging as a clinical tool for demonstrating disturbed cerebral circulation and metabolism. J Comput Tomog 2: 664,1978 (abst)

19. KUHL DE, ALAVI A, HOFFMANE, et al:@Local cerebralblood volume in head injured patients: determination byemission computed tomography of (Tc-99m) red cells. JNuclMed 19:699—700,1978(abst)

20. KUHL DE, EDWARDSRQ, RiccI AR, et al: The MARK IVsystem for radionuclide computed tomography of the brain.Radiology 121: 405—413, 1976

21. PHELPSME, GRUBBRL, TER-POGOSSIANMM: In vivo cerebral blood volume by X-ray fluorescence: validation of themethod. J App! Physiol 35: 741—747,1973

22. PHELPS, ME: Emission computed tomography. Sem NuclMed 7:337—365,1977

23. GLASSH, BRANTA, CLARKJC, et al: Measurementof bloodvolume using red cells labeled with radioactive carbon monoxide. J Nucl Med 9: 571—575, 1968

24. CLARK,JC, BUCKINGHAMPD: Short-lived radioactive gasesfor clinical use. London and Boston, Butterworths, 1975, pp227—231

25. PHELPSME, HOFFMANEJ, HUANG SC, et al: ECAT: Anew computerized tomographic imaging system for positronemitting radiopharmaceuticals. J NucI Med 19: 635—647,1978

26. EVERETT NB, SIMMoNS B, LASHER EP: Distribution ofblood (Fe5') and plasma (I's') volumes of rats determined byliquid nitrogen freezing. Circ Res 4: 419—424, 1956

27. LARSENOA, LASSENNA: Cerebral hematocrit in normalman. J App! Physiol 19: 571—574, 1964

28. OLDENDORFWH, KITANOM, SHIMIzu 5, et al: Hematocritof the human cranial blood pool. Circ Res 17: 532—539,1965

29. SKLARFH, BURKEEF, JR, LANGFITrTW: Cerebral bloodvolume: values obtained with “Cr-labeled red blood cellsand RISA. J App! Physiol 24: 78—82,1968

30. STUDERR, POTCHENEJ: The radioisotopic assessmentofregional microvascular permiability of macromolecules. Microvasc Res 3: 35—48,1971

31. EICHLINGJO, RAICHLEME, GRUBBRL, JR, et al: In vivodetermination of cerebral blood volume with radioactiveoxygen-IS in the monkey. Circ Res 37: 707-714, 1975

32. ROUGHTON RJW, RooT WS: The fate of CO in the bodyduring recovery from mild carbon monoxide poisoning inman. Am J Physiol 145: 239—252, 1945

33. WENNESLANDR, N0MOF N, BROWNE, et al: Distributionof CO and radiochromium in blood and tissue of rabbit anddog. I. Carbon monoxide. Proc Soc Exper Biol Med 96: 655—657, 1957

34. INOVARND, CR0NQvI5T5, EKBERGR, et al: Normal valuesof regional cerebral blood flow in man, including flow andweight estimates of grey and white matter. In Regional Cerebral Blood Flow. Ingvar DH, Lassen NA, eds. Copenhagen, Munksgaard, 1965,pp 72—78

35. LIERSE W, HORSTMANN E: Quantitative anatomy of thecerebral vascular bed with special emphasis on homogeneityand inhomogeneity in small parts of the grey and whitematter. In Regional Cerebral Blood Flow. Ingvar DH, Lassen NA, eds. Copenhagen, Munksgaard, 1965, pp 15-19

April 27-28, 1979 Glenedon, Oregon

The Pacific Northwest Chapter of the Society of Nuclear Medicine will hold its Annual Spring Meeting on April27-28, 1979,at SalishanLodge,Glenedon,Oregon.

The program will consistof 1)a tutorial on approachesto nuclear cardiology: pathophysiologytechniques, interpretation, and clinical utility of current procedures—Chairman: Dr. Glen Hamilton; 2) a panel discussionwith representatives of the major nuclear medicine computer companies—Chairman:Dr. Hamilton; and 3) aclinical refresher course on pediatric nuclear medicine: handling of the pediatric patient, sedation, restraint,special problems and interesting cases,conducted by Drs. Gary Gates and Thomas Rudd.

For further information contact:

Justine ParkerP.O. Box 40279San Francisco, Ca 94140(415) 647-0722

334 THE JOURNAL OF NUCLEkR MEDICINE

PACIFIC NORTHWEST CHAPTERANNUAL SPRING MEETING

Salishan Lodge

by on April 5, 2019. For personal use only. jnm.snmjournals.org Downloaded from

1979;20:328-334.J Nucl Med.   M. E. Phelps, S. C. Huang, E. J. Hoffman and D. E. Kuhl  CarboxyhemoglobinValidation of Tomographic Measurement of Cerebral Blood Volume with C-11-Labeled

http://jnm.snmjournals.org/content/20/4/328This article and updated information are available at:

  http://jnm.snmjournals.org/site/subscriptions/online.xhtml

Information about subscriptions to JNM can be found at:  

http://jnm.snmjournals.org/site/misc/permission.xhtmlInformation about reproducing figures, tables, or other portions of this article can be found online at:

(Print ISSN: 0161-5505, Online ISSN: 2159-662X)1850 Samuel Morse Drive, Reston, VA 20190.SNMMI | Society of Nuclear Medicine and Molecular Imaging

is published monthly.The Journal of Nuclear Medicine

© Copyright 1979 SNMMI; all rights reserved.

by on April 5, 2019. For personal use only. jnm.snmjournals.org Downloaded from