вход по аккаунту


Effects of stroke on local cerebral metabolism and perfusion Mapping by emission computed tomography of 18FDG and 13NH3.

код для вставкиСкачать
Effects of Stroke on
Local Cerebral Metabolism and Perfhion:
Mapping by Emission Computed
Tomography of I8FDG and 13N&
David E. Kuhl, MD, Michael E. Phelps, PhD, Arthur P. Kowell, MD, PhD,
E. Jeffrey Metter, MD, Carl Selin, MS, and James Winter, MD, PhD
By means of emission computed tomography (ECT), we used 18F-fluorodeoxyglucose ("PDG) and I3N-ammonia
(13NH3)as indicators of abnormalities in local cerebral glucose utilization (LCMRgI,)and relative perfusion, respectively. The ECAT positron tomograph was used to scan normal control subjects and 10 stroke patients at
various times during recovery. In normal subjects, mean CMhl, was 5.28 & 0.76 m g per 100 gm tissue per minute
(mean f ,SD; N = 8). In patients with stroke, mean CM&Ic in the contralateral hemisphere was moderately decreased during the first week, profoundly depressed in irreversible coma, and normal after clinical recovery.
Quantification was restricted by incomplete understanding of tracer behavior in diseased brain, but relative local
distributions of 18FDGand 13NH3trapping qualitatively reflected the increases and decreases as well as coupling
and uncoupling expected for local alterations in glucose utilization and perfusion in stroke. Early after cerebrovascular occlusion there was a greater decrease in local trapping of l3NH, than 18FDG within the infarct, probably
because of increased anaerobic glycolysis. Otherwise, "FDG was a more sensitive indicator of cerebral dysfunction
than was 13NHs.Hypometabolism, due to deactivation or minimal damage, was demonstrated with the 18FDGscan
in deep structures and broad zones of cerebral cortex that appeared normal on x-ray computed tomography and
technetium 99m pertechnetate scans.
In its present state of development, the "FDG ECT method should aid in defining the location and extent of
altered brain in studies of disordered function after stroke. With improved knowledge of tracer behavior in diseased brain, the method has promise for mapping the response t o therapeutic intervention and increasing our
understanding of how the human brain responds to stroke.
Kuhl DE, Phelps ME, Kowell AP, et al: Effects of stroke o n local cerebral metabolism and perfusion: mapping by
emission computed tomography of "FDG and I3NH,. Ann Neurol 8:47-60, 1980
Details of local cerebral damage, hemorrhage, and
edema can be demonstrated in the stroke patient by
means of x-ray computed tomography (XCT) [2], but
local physiological correlates are more difficult to assess in human beings. Most measurements of cerebral metabolism and perfusion in humans lack
three-dimensional resolution and are restricted t o
whole brain [161 or large regions of brain [30, 39,
491, where local events can be obscured. In experimental animals, quantitative autoradiography has
been applied extensively for absolute measurements
of local cerebral blood flow (LCBF) [5,42] and local
cerebral glucose utilization (LCMK,,) [501 in small
cerebral structures. In humans, emission computed
tomography (ECT) [18, 311 can serve as an in vivo
analog of quantitative autoradiography. It is a nonin-
vasive scanning method which produces a similar but
less detailed cross-sectional picture of brain radioactivity in man following intravenous injection of a
labeled indicator. The potential of ECT for study of
human brain physiology is still relatively undeveloped, but the ECT method has already been applied
in humans to absolute measurements of local cerebral
blood volume [8, 9, 17, 23, 361 and LCMRgI, [ I l ,
20-22, 37,43, 441. The project reported here was a
first step in exploring the potential value of ECT in
t h e study of stroke patients, in this instance using
radioactive fluorine 18-labeled fluorodeoxyglucose
(18FDG) [ l l , 20-22, 37, 43, 441 and ammonia
labeled with nitrogen 13 (13NH3)[27, 32, 33, 351 as
indicators of LCMRgI, and relative perfusion, respectively.
From the UCLA School of Medicine and the Laboratory of Nuclear Medicine and Radiation Biology, L o s Angeles, CA.
Address reprint requests to D r Kuhi, Division of Nuclear
Medicine, UCLA School of Medicine, L o s Angeles, CA 90024.
Received Aug 13,1979, and in revised form Nov 16. Accepted for
publication Nov 24, 1979.
0364-5134/80/070047-14$01.25 @ 1979 by David E. Kuhl
The in vivo "FDG-ECT method [ l l , 2 1 , 2 2 , 3 7 ,
4 3 , 441 for determining LCMR,,, in individual brain
structhres is derived from the carbon 14-deoxyglucose autoradiographic method of Sokoloff et
a1 1501. "FDG enters the brain rapidly, is phosphorylated by brain hexokinase, and the metabolic
product, "FDG-6-P04, remains fixed, with little
further metabolism [ l l , 2 1 , 2 2 , 3 7 , 43,441. Calculations are based on a model of the biochemical behavior of deoxyglucose and glucose in brain [ l l , 37,
4 4 , 501. The time course of specific activity in cerebral capillary blood is estimated by measuring arterialized venous blood samples [ 3 71 obtained while
the blood is clearing of tracer. At times greater than
40 minutes, local cerebral 18F concentrations are
measured by ECT scan. With knowledge of predetermined rate coostants and the lumped constant
(LC), the operational equation (see Appendix) allows
calculation of the LCMR,,, corresponding to each
zone of the tomographic image. The method measures exogenous glucose utilization, the glycolytic
rate under the assumption that there is no n e t glycogen accumulation or glycogenolysis. The method
does not distinguish aerobic from anaerobic glycolysis.
Sokoloff et a1 [50] have found the LC to be stable
and uniform throughout normal brain regardless of
alterations in carbon dioxide tension, blood glucose
concentration, and state of anesthesia. Providing that
sufficient time is allowed for a near steady state condition to be approached, calculation of LCMR,,, is
quite insensitive to even several-fold inaccuracies in
the values of rate constants [ 11, 501. In diseased brain
the LC may be just as stable and uniform and the
change in rate constants may be unimportant, but this
possibility must be experimentally verified before it
can be accepted. In a severely ischemic zone, the
ratio of distribution spaces for "FDG and glucose (A)
and the hydrolysis of glucose-6-phosphate to free
glucose (4) may be so altered as to change the LC substantially, and large changes in rate constants may
greatly affect the calculation of LCMbl , [ l l , 3 7 , 501.
The magnitude of inaccuracy caused by assuming
norrhal values of the LC and rate constants in calculating LCMR,I,. for the severely ischemic zone remains to be determined by careful measurement.
Until then, absolute values of L C M h i , for diseased
tissue must be interpreted with caution.
We chose I3NH3[12, 27, 3 3 , 351 as an indicator of
relative cerebral perfusion because ammonia has a
cerebral uptake that varies with capillary perfusion, a
static cerebral distribution which is desirable for ECT
scanning, a short physicd half-life (10 minutes) which
permits its use before an "FDG scan without residual
interference, and simple chemical preparation. We
made no attempt t o quantify I3NH3distributions in
48 Annals of Neurology Vol 8 No 1 July 1980
absolute units of LCBF. Phelps et a1 [ 3 5 ] have reported that after a bolus intravenous injection, I3NH,
is rapidly extracted from blood into brain tissue
through the diffusible form of ammonia, is rapidly
incorporated into a glutamate-glutamine pool of large
size and slow turnover rate, and thus is effectively
trapped in a cerebral distribution that depends nonlinearly on local capillary perfusion. For example,
after local compression of dog brain, we found that
the relative distribution of trapped 13NH3underestimated both increases and decreases in relative distribution of LCBF as measured by the microsphere
method (unpublished results). Local 13NH, trapping
matched large LCBF decreases, was altered less than
half as much as LCBF when flow changes were between -20% and +50%, and rose only 37% for a
100% increase in LCBF.
The relationship between "NH, uptake and LCBF
depends on metabolic trapping and might be unpredictable in diseased tissue. Thcre is evidence [ 3 5 ]
that cerebral extraction of I3NH3is not greatly dependent upon blood p H in a range of 7.2 to 7.7,
bIood ammonia concentrations of 80 to 1,340 pgldl,
and reductions in glucose and oxygen metabolism
with and without elevated blood ammonia levels.
However, it is not well known how "NH, uptake in
diseased brain is influenced locally by alterations such
as depletion of t h e glutamine pool, decreases in
glutamine synthetase, alterations in the blood-brain
barrier, or marked changes in the blood-brain p H
gradient [ 2 5 , 26, 3 2 , 351. Caution in interpreting
quantitative measurements in diseased tissue is required for other potential CBF tracers as well as for
In this project we questioned: (1) d o absolute
values of mean CMR,,, determined by the "FDG
method in normal or nearly normal human brain
agree with known values?; ( 2 ) are there consistent
alterations in "FDG and I3NH3scans within abnormal brain zones identified in stroke patients by XCT
or technetium 99m (99mTc)pertechnetate scans?; and
( 3 ) will evaluation of cerebral function with ECT
provide better assessment of the degree and extent of
injury to the brain in stroke than is estimated by
X CT or 99mTcpertechnetate scans?
For preparation of ''FDG, lRF-labeled fluorine was produced in the UCLA medical cyclotron by the 20Ne(d,a)1RF
nuclear reaction. Synthesis was by the method of Ido et a1
[13] as modified by Robinson and associates 1451. Specific
activity was 10 to 15 mCi pet milligram. Radiochemical
purity, as assayed by high-pressure chromatography, was
greater than 95% I'FDG; the remainder was determined
by thin-layer chromatography to be deoxymannose. T h e
usual intravenous dose for patients was 5 to 10 mCi.
Fig 1 . Normal W H , and “FDG images in humans, made
parallel to the canthomeatal plane at three different le6’els;
levels 28 and 3A (not shown) are between levels 2A and 3 8 .
Level I A (basal): MOFC = medial orbital frontal cortex; SC
= sylviun cistern; TL = temporal lobe: PS = pons; PB =
petrous bone; C H = cerebellur hemisphere. Lei!el2A (middle):
FLC =frontal lobe cortex; AHLV = anterior horn of lateral
ventricle; I = insulu; TLC = temporal lobe cortex; ALV =
atrium of luteral ventricle; VC = visual cortex; OLC = occipital lobe cortex; GCC = genu of corpus callosum; HCN =
head of caudate nucleus: PT = putamen; I C = internul cup.rule; T M = thalamus; SCC = splenium of corpus cullosum.
Level 3 B (vertex): FLC =frontal lobe cortex: IHF = interhemi~phericfissure;PLC = parietal lobe cortex; CS = centrum
semiaide; OLC = occipital lobe cortex.
Nitrogen 13 was produced in the cyclotron by bombardment of water with protons, involving the l60(p, a)13N
nuclear reaction, followed by reduction of the “N compounds to ammonia with deVarda’s alloy [ 5 3 ] . The radiochemical purity of the product was greater than 99%
I3NH, and contained less than
M carrier ammonia
as determined by liquid and gas chromatography. The usual
intravenous dose for patients was 20 mCi.
Emission Computed Tomography
All “FDG and “NH, scanning was performed with the
ECAT positron tomograph [ 10, 341 (ORTEC Inc. Life Sciences, Oak Ridge, T N ) operated in the medium-resolution
mode. The full width at half maximum (fwhm) measure of
spatial resolution was 1.3 cm (intrinsic) and 1.7 cm (final)
within the image plane and 1.8 cm in the axial direction.
For l3NNH3 scans, started several minutes after injection,
the count rate was approximately 22,000 per minute per
millicurie injected; for ‘TDG scans, started 40 minutes
after injection, the count rate was approximately 25,000.
Scan duration was adjusted so that each image contained 1
to 1.5 million counts. The “NH, scan (13Nphysical halflife, 10 minutes) preceded the l8FDG scan (18F physical
half-life, two hours) by at least one hour, sufficient time to
allow radioactive decay of “N and avoid interference. Most
commonly, six levels were scanned sequentially parallel to
the canthomeatal plane from upper cerebellum to above
the cerebral ventricles (Fig 1).
For calculation of LCM&,, we used an operational
equation developed by Huang et a1 [ l l ] and validated
by Phelps et a1 [37] that was an extension of the autoradiographic model developed by Sokoloff et a1 [501 (see Appendix). This required measurement of local cerebral “F
concentration by ECT, determination of the time course
Kuhl et al: 18FDG ECT for Detecting Brain Damage in Stroke
of blood lsF activity and glucose concentration, and knowledge of certain rate constants and an LC [ll, 371 (see Appendix). The model was incorporated into the system
software of the tomograph, and a protocol was established
to allow operation by a nuclear medicine technologist [37].
When determination of LCMR,,, was required, the patient was positioned with one hand heated to 44°C in a hot
water glove-box, a procedure which made possible sampling of "arterialized" venous rather than arterial blood
[37]. '*FDG was injected intravenously in the opposite
arm, and blood was sampled only from a vein of the heated
hand. Venous blood samples were taken every 10 to 15
seconds for the first 1.5 to 2.0 minutes, with sample intervals progressively lengthened. Samples were immediately
put on ice and subsequently centrifuged to separate plasma
for the determination of plasma glucose and "FDG concentration. Plasma glucose concentrations were measured
in duplicate by enzymatic techniques. "FDG plasma concentrations were determined by counting weighed samples
in a sodium iodide well counter. A calibration of the tomograph relative to the well counter allowed determination of LCMR,,, in milligrams per 100 gm of tissue per
minute [371.
Because of the relatively slow trapping of IKF by the
brain, the first "FDG scan was begun 40 minutes after injection, when it was assumed that a near steady condition
had been established. Unless otherwise noted, all subjects
were awake with their'eyes open during the time between
injection and the first scan.
Regions of interest were selected on the tomograph display screen, and local values of I3N concentration, "F concentration, or LCMR,,, were then determined. I3N and IFF
concentrations for anatomical zones were averaged among
multiple levels and normalized to the maximum concentration (peak visual cortex value = 100) among the levels.
DN-,3,,and DMR
were defined as the percentage differences existing between selected zones and the corresponding contralateral zones for mean values of "N concentration, '*F concentration, and LCMR,,,, respectively.
Usually these were measured in lesions. The average
LCMRgI, for both gray and white matter was determined as
the simple average among all anatomical zones measured.
In calculating mean CMR,,, for whole brain in each subject,
zonal averages of gray and white LC-,,
were weighted
on the assumption that the brain is 50% gray and 50%
white matter.
Other Studies
All patients had at least one XCT study. Some patients
underwent scintillation camera or ECT studies after injection of ""T
pertechnetate. Pertechnetate flow studies
were performed to estimate cerebral circulation; a sequence of images was made at +second intervals immediately following intravenous injection of 20 mCi of "TC
pertechnetate with the scintillation camera placed against
the patient's forehead. Two hours later, alterations in the
blood-brain barrier were determined both by conventional
imaging using the scintillation camera and by ECT imaging
using the Mark IV scanner [19, 211.
50 Annals of Neurology Vol 8 No 1 July 1980
Seven normal volunteers underwent 13NNH3scanning. AIJ
were young males (median age, 24 years; range, 18 to 30
years) and all were right-handed. Ten normal volunteers
underwent "FDG scanning. All were young males (median age, 23 years; range, 21 to 35 years); 9 were righthanded and 1 was left-handed. Four of these subjects had
both '"H3 and lRFDGscans, 8 had determinations of
LCMR,,,., and 3 had a second determination of LCMR,,, on
another day.
W e studied 10 patients (8 men and 2 women) with clinical evidence of completed stroke. In 1 patient (aged 22
years) the cause of acute cerebral ischeqia was a knife
wound of the internal carotid artery; in the others (median
age, 6 1 years; range, 46 to 72 years) there were thrombotic
or embolic infarctions in the distribution of the middle cerebral (8 patients) or anterior cerebral ( 1 patient) arteries.
Sixteen paired '"H3 and 18FDGscans were performed in
this group; 5 patients were studied during the first week
after stroke and 5 recovered patients with residual aphasia
were studied after six weeks; 3 patients had studies at both
times. In addition, LCMR,,, measurements were made o n 2
patients during the first week and o n 5 recovered patients
later than six weeks after stroke.
Normal Distributions
In normal subjects the cerebral activity distributions
were the same for "FDG and 13NH3 scans (Fig 1,
Table), and tbere was no difference between left and
right hemispheres. The mean coefficient of variation
among subjects for normalized zonal concentrations
of 13N and "F was 6%; for zonal LCMh,, it was
15%. When l*FDG scans were repeated in the same
normal subjects, the average variation from the mean
in paired determinations of normalized zonal 18Fconcentration was 3%. The same normal distributions of
cerebral activity concentration and LCM%,, were
found in the hemispheres contralateral to chronic
infarcts in 5 patients who were scanned later than six
weeks after their stroke (see the Table).
Meun CMR,,,
The mean C M k , , was the same in the brains of normal subjects (median age, 23 years) and in the contralateral hemispheres of recovered stroke subjects
(median age, 61 years) who had residual aphasia: 5.28
i 0.76 mg/100 gmlmin (mean f SD; N = 8) and
5.23 ? 1.1 (N = 5), respectively. In duplicate '*FDG
scans of normal subjects, the average variation from
the mean in the paired determinations of mean
C M k , , was 12%. Two patients had measurements of
LCMR,,, during the first week after stroke. One was
alert at seven days, but the mean CMR,,, in his contralateral hemisphere was reduced 32% below the
normal value obtained three months later. The other
Cerebral Distributions of Metabolism and Perfusion from 13NH, and "FDG Scans"
Normal Subjects
(both hemispheres)
(mg/100 g m h i n )
(N = 8 )
80.0 f 3.8
81.8 i 3.2
77.6 f 2.3
76.8 & 3.1
80.7 f 5.9
6.80 f 0.92
6.84 f 0.45
6.43 f 0.83
6.41 0.92
6.71 f 0.56
f 3.3
75.6 f 6.5 6.64 t 1.2
77.3 f 3.2 6.36 -t 1.7
75.3 f 7.8 6.87 t 1.5
72.3 f 4.2 6.20 f 1.2
73.8 f 11.0 4.05 f 1.5
f 5.3
t 6.6
f 2.9
f 3.3
f 2.9
7.87 t 0.92
4.16 f 1.0
3.27 f 0.59
3.51 f 0.56
6.86 f 0.83
3.70 f 0.71
1.88 f 0.17
Brain Zone
Frontal gray
Temporal gray
Parietal gray
Occipital gray
Caudate nucleus,
Visual cortex
Frontal white
Parietal white
Occipital white
Mean gray
Mean white
Mean gray/
mean white
f 5.3
t 5.0
t 3.9
f 4.0
f 5.0
f 3.0
f 4.1
f 0.15
f 3.9
t 2.7
t 3.1
Recovered Stroke Patients
(contralateral hemisphere)
f 3.7
f 0.10
f 6.2
f 3.3
f 2.6
f 3.1
95.1 f 1.5
45.9 t 4.2
f 3.3
43.5 f 1.7
f 3.3
43.0 f 1.4
f 1.0
78.3 f 4.6
f 3.2
44.1 f 2.3
f 0.13
1.78 t 0.04
f 2.4
f 4.2
(mg/100 gm/min)
(N = 5 )
f 2.4
f 0.77
f 0.70
f 1.5
f 0.72
f 0.10
aRecovered stroke patients were scanned later than six weeks after cerebrovascular occlusion. I3N and 'SF values are relative cerebral
concentrationsnormalized to the maximum concentration in each subject (peak visual cortex value = 100). L C m , , calculations were based
on the operational equation and constants contained in the Appendix. Values for cerebellar structures were not obtained for all subjects and
are not included here. The values are mean standard deviation.
patient was in irreversible coma when scanned five
days after suffering extensive bilateral cerebral emboli, and mean CMRgI, was reduced 63% below
normal bilaterally.
( v e r y early)
- 100
- 50
+ 50
F i g 2. Relationships of DF-,X
and DN.13
and abnormal brain
zonex of .stroke patients. Data points falling on the dotted line
represent metabolism = perfusion deficits, assuming that the
relationship of relative 13N concentration to relative LCBF was
the same us we found in compressed dog brain (unpublished
observations). (A)In the first two days after a stroke, perfusion
is decreused more than glucose utilization within the infarct.
( B ) In the first two weeks, luxury petfusion appears at the infarct margins; local perfusion is increased but glucose utilization is not. (C) In oldpermanent infarcts that are obvious on
X C T scan, glucose utilization and petfusion are both markedhi reduced. (0)In some ipsilateral tiJsues thwe is a less
marked reduction i n glucose utilizution and perfusion but no
X C T evidence of structural damage. (MR = metabolic rate;
CRF = cerebral bloodjow.)
Local Relationships i n Abnormal Bruin
FOCAL LESIONS. In 5 patients with acute or chronic
infarcts we determined the relationship of DMRto
DF.18,measuring a total of fifteen lesions sites in data
obtained in nine 18FDG scan sessions. Scans were
begun an average of 64 2 22 minutes after injection;
no scan was performed earlier than 38 minutes after
injection. D,,, ranged from -12% to -71%. For
these conditions we found D,, = (1.17 -t- 0.05)DF-1x.
In abnormal zones of scans in stroke patients,
DF.18and D,-,, were in four different relationships,
grouped as A through D in Figure 2. Data points falling on the dotted line represent metabolism = perfusion deficits (coupled relationship in which local metabolism and local perfusion are equally decreased),
assuming that the relationship of relative I3N concentration to relative LCBF was the same as we found
in locally compressed dog brain (unpublished results). A comparison of data A through D with the
expected course of events after stroke and with results of X CT and 99mTcpertechnetate scans was consistent with the assumption that DF-IRand D,-,,
reflected relative metabolic changes and relative
perfusion changes, respectively.
Kuhl et al: IRFDGECT for Detecting Brain Damage in Stroke 51
Fig 3 . (Patient C. G.) Serial Jcans (lwel2A)of a 52-year-old
Group A represents data from focal infarcts in 2
man following a focal infarction within the distribution of the
patients (C. G. and M. R.) scanned within the first
ldt middle cerebral artery. Within the infarct (i),pe4usion
two days after stroke and one hour after injection of
was redmed more than glucose utilization ut one duy. but both
I*FDG (Figs 3, 4). There were metabolism > perfuwere decreased equally within the stable infarct at three
sion mismatches (uncoupled relationship in which
months. L u x u y perfusion waJ Jeen i n the infarct margins (m)
local metabolism exceeds local perfusion); perfusion
daring thejirst two weeks. The ipsilateral thalamus ( t ) was
hypometabolic even though it appeared normal on X C T . See
was depressed out of proportion to the decrease in
Figure 10 for quantitative relationships.
glucose utilization.
Group B represents data from hyperemic infarct
margins in 3 patients scanned during the second
week after their infarctions (Figs 3 , 5, 6). There were
distinct regions of decreased density measuring more
perfusion > metabolism mismatches, or luxury perthan 2 cm wide (Figs 3, 4 , 7, 8).These were marked
fusion [24] (uncoupled relationship in which local
metabolism = perfusion deficits; values of DF-18were
perfusion exceeds local metabolism); while perfusion
below -53% (LCMR,,, = -62%) in all these inwas increased, glucose utilization was decreased or
unchanged. These were the only instances of focal
Group D data represent local cerebral dysfunction
increase in 13N concentration found in this study. In
in the ipsilateral thalamus of 5 patients. These were
each patient, regional hyperemia was corroborated
metabolism = perfusion deficits even though the
by a !'!'"'Tc pertechnetate flow study using scintillation
thalamus appeared normal on XCT scan. Findings
camera imaging. In 2 patients (Figs 3, 6), the region
were similar in other ipsilateral structures such as the
of maximum I3N uptake was adjacent to but did not
head of the caudate nucleus and broad areas of cerecoincide with the region of maximum damage to the
bral cortex, including the visual cortex of 2 patients
blood-brain barrier as determined by Mark IV emission tomography after intravenous injection of Y S m T ~ with contralateral homonymous hemianopia (Figs 4 ,
9). In all these zones, DF.18 was never depressed more
than -48% (LCMR,,, = -56%), a clear separation
The scan data in Group C represent permanent
from the greater metabolic depressions found in
gross tissue destruction in old (> two months) infarct
nearly complete tissue destruction (Group C).
zones which were identified on XCT scans as stable,
Annals of Neurology
Vol 8 No 1 July 1980
Pig 4. (Patient M . R.) Serial scans (level2A) of a 22-year-old
m n after sudden transection of the le,4 internal carotid artery
by a knzfe wound. Within the ldt frontal lobe infarct (i), perfusion was decreased more than glucose utilization i n the earliest scan, but i n later scans this changed to a metabolism =
perfusion dejicit. Bloodjow increased i n the infarct margin
(m) during the second week. Distant from the local infarct.
ipsilateral cortex and thalamus (t) were hypometabolic. Ipsilateral primary visual cortex was hypometabolic and was associated with a contralateral homonymous hemianopia. All
these distant stractures appeared normal on X C T . See Figure
10 for quantitative relationships.
F i g 6 . Scans (le-uel3A) of a 79-year-old woman made ten days
after she experienced bilateral embolic occlusion of the anterior
cerebral arteries followed by reperfusion on the left and infarction on the right. Within the luxury perfusion zone on the
left, "FDG trapping W ~ normal.
13NH, trapping uas increased, however (solid arrow), and hyperemia was confirmed
by a 9mTc-pertechnetate$ow study. Within the infarct zone
on the right, both lsFDG and I3NH, trapping were decreased
(dotted arrow), hypoperfusion was seen i n the LJYmT~pertechnetate $ow study, and 99mTcECT by the Mark IV tomograph showed maximum blood-brain barrier damage.
l8FDG scans are not shown here.
Three patients were scanned early
after stroke and later after recovery. In 1 (see Fig 8),
ECT scans made five days after a focal cerebral infarction showed wider distribution of dysfunction
than was noted on XCT, and measurements of DF.18
and D,-,, were unchanged at six months. More detailed results from the other 2 patients are shown in
Figure 10. Both patients had acute focal infarction, 1
F i g 5 . Scans (level 3A) of a 46-year-old man made eleven days
after an infarction i n the distribution of the right middle cerebral artery. Luxury perfusion is seen in the margins of the
infarct (arrows). Localperjksion is increased but local glucose
utilization is decreased (DF.15= -2696, DN.,;$= +23 5%). A
pertechnetatej a w .study confirmed the local hyperemia.
Kuhl et al: "FDG ECT for Detecting Brain Damage in Stroke
F i g 7 . Scans (level 2 B ) of a 57-year-old man made five months
after a cerebral infarction in the distribution of the left middle
cerebral artery. There is marked depression of local glucose
utilization and perfusion (DF-18= -5996, D,.,,= -53 %)
within the large permanent infarct (i). The left thalamus (t)
appears dysfunctionalon ECT scans (DF.18
= -48%, D,.,, =
-30%) but normal on X C T scan. Findings from a visual
field examination were normal. Compare the normal-appearing
left primary visual cortex (PVG) with that of the patient
shown in Figure 9, who has a right homonymous hernianopia.
F i g 8. Scan.i ( b e 1 2 A j of a 61 -year-old woman made five days
after a right middle cerebral artery occlusion. Although there is
X C T evidence of a low-density infarct (arrow) limited to the
head of the caudate nucleus. anterior limb of the internal capsule, and anterior putamen, the ECT scans .rhow more exten.rive dysfunction involving the ipsilateral thalamus (t) and
overlying cortex (arrow). I n a repeat scan at six months, DF.lrc
and DN.13
were unchanged.
after chronic occlusion of both internal carotid arteries (Patient c. G . ) ,and the other after acute internal carotid artery occlusion (Patient M. R.).
Patient C. G. was a 52-year-old man who underwent a
left superficial temporal-middle cerebral artery anastomosis after several months of transient ischemic attacks
that had been characterized by right facial and upper extremity weakness and aphasia. Before surgery, arteriography o f t h e aortic arch had demonstrated total occlusion of
both internal carotid arteries; "FDG and l3NH, scans were
normal. Three days after surgery he experienced the onset
of right hemiparesis and severe nonfluent aphasia. Arteriography confirmed occlusion of branches of the left
middle cerebral artery. An XCT scan performed one day
after the stroke was normal, but a repeat study at three
months demonstrated a low-density infarct in the anterior
distribution of the left middle cerebral artery (see Fig 3).
One week after the stroke he was found to have a deficit in
sensation in the right hand. He improved steadily, and at
54 Annals of Neurology Vol 8 No 1 July 1980
F i g 9. Deactivation of intact left visual cortex due t o an anterior infarct. Scans of a 61-year-old man (level2A) with
right homonymous hemianopia and m infarct (i) restricted t o
the distribution of the left middle cerebral artery. ECT scans
show decreased metabolism and perfusion in the strucruralllly
intact left primary (PVC) and associative (AVC) visual cortex, demon.rtrating functional deactivation caused by the more
anterior destraction ofthe ldt optic tract (OT), Iuteral
geniculate body (LGB), or optic radiation (OR).In the leji
PVC, Dp.lti= -16% and DNU-,$
= -5%. I n the lefr AVC,
= -37% and DN.13
= -27%. Compure with Figure
three months there was residual weakness of the right arm
and right side of the face, mild dysphasia, and no sensory
abnormalities. His scan results are shown in Figures 3 and
10. Within the infarct, perfusion was reduced more than
glucose utilization at one day, but both were decreased
equally within the stable infarct at three months. Luxury
perfusion was seen in the infarct margins during the first
two weeks. The ipsilateral thalamus was hypometabolic
even though it appeared normal o n XCT.
Patient M. R., a 22-year-old man, was admitted in coma
after a knife wound had completely transected his left
internal carotid artery. After regaining consciousness, he
was found to have right hemiparesis, right hemisensory
deficit, right h yperreflexia, right homonymous hemianopia,
and global aphasia. After two months the patient had persistent mild right hemiparesis, dysarthria, and dysphonia.
His scan results are shown in Figures 4 and 10. Within the
left frontal lobe infarct, perfusion was decreased more than
glucose utilization in the earliest scan, but in later scans this
changed to a metabolism = perfusion deficit. Bkood flow
increased in the infarct margin during the second week.
resolution is much more limited. Because of this, the
gray to white matter ratio for LCMR,,, in normal
subjects (see the Table) was only 1.88, a value conlnfarcl
siderably lower than that reported for monkeys (3.0)
by Kennedy et a1 [ 151. Unless there is a priori knowledge of the size and shape of smaller structures, accurate quantification of concentration by ECT is limited
to tissue dimensions greater than the fwhm system
for spatial resolution [lo]. In this study, fwhm was
1.7 cm. Since the width of gray and white matter
structures is commonly smaller than this, zonal values
in the Table represent mixtures of the two in unprel p s i lateral
dictable proportions. Gray matter values are underestimated, white matter values overestimated. Con--40
centrations in larger structures, such as hemisphere
averages and infarcts greater than 2 cm in diameter
I p s 1 iaterai
(see Fig 2), are quantified more accurately.
Interpreting the quantitative relationships of DN-lS,
- 40
and D M R for focal lesions will remain probI Day
I Week
3 Months 2 Doys 2 Weeks 2 Months
lematic until there is more certainty of the trapF i g 10. Quantitative relationships of DF.laand D N .in~ serial
ping conditions for 13NH3and '*FDG in all states
scans of the 2 patients ilhstrated in Figures 3 and 4.
of diseased tissue. In addition, the basis for
contralateral zone-is assumed to
of both blood flow and glucose
Distant from the local infarct, the ipsilateral cortex and
function in the contralateral
thalamus were hypometabolic. Ipsilateral primary visual
hemisphere, such as diaschisis [6, 481, must be reccortex was hypometabolic and was associated with a conognized if side-to-side comparisons are to be intertralateral homonymous hemianopia. All these distant
preted correctly. D,.,, represents the percentage
structures appeared normal on XCT.
difference of activity concentration in contralateral
Figure 10 illustrates the time-dependent pattern of
regions of combined phosphorylated and nonphoslocal relative glucose utilization and relative perfuphorylated '*FDG. The zonal metabolic rate differsion in these 2 patients. Within the focal infarct, gluence, D M R , is calculated on the assumption that norcose utilization fell progressively with time. Initially,
mal rate constants and LC apply bilaterally. If this is
perfusion was decreased more, but it rose transiently
so, D M R is independent of the actual value chosen for
during the first two weeks only to fall again, and, after
LC; but if not, the magnitudes of D,R reported here
several months it matched the decrease in glucose
have the same uncertainty as discussed for LCMK,,
utilization within the stable infarct. The margins of
in diseased tissue. Under the conditions of this study
the infarct had a variable pattern. Initially, perfusion
we found a constant relationship between D M R and
was either very high (luxury perfusion), as in Patient
DF.18,but this would not necessarily apply to other
C. G., or maximally depressed (Patient M. R.); initial
glucose utilization was almost normal but decreased
Compatibility of ECT with Other Dutu
progressively with time. While local glucose utilizaABSOLUTE VALUES OF MEAN CM&,,.
We found
tion and perfusion were unchanged in the distant ipthe same mean CMR,,, values in whole brain of norsilateral cortex of Patient C. G., both were persismal young subjects and in the hemisphere contralattently reduced in Patient M. R. as a result of the
era1 to infarction in recovered older stroke patients:
global postischemic damage to the left hemisphere
5.28 f 0.76 and 5.23 -+ 1.1 mgper 100 gm tissue/min
that followed the sudden transection of his left ICA.
(mean t SD), respectively. Close correspondence
There was no evidence of this widespread damage on
between these results and the average of published
X C T scan. The ipsilateral thalamus was hypometavalues [ 3 , 7 , 29, 47, 521 of mean CMR,,, in humans
bolic in both patients even though it appeared normal
determined by the Kety-Schmidt method (5.38 &
on XCT scans; both patients had contralateral
0.77 mgl100 gm/min, mean k SD) established resensory deficits.
producibility but should not be considered further
confirmation of the numerical value of mean CMK,,
in normal human brain; these published data obRestrictions in ECT Quantz$cation
tained by the Kety-Schmidt method served as the
While ECT of '*FDG and l3NNH, is a process analobasis for estimating the LC value used in our metagous to quantitative autoradiography 142, 50J, spatial
Potient C.G
Chronic I C A Occlusion
Patient M R
Acute I C A Occlusion
Kuhl et al: "FDG ECT for Detecting Brain Damage in Stroke
bolic rate equation [ l l , 371. In normal conscious
macaque monkeys, Kennedy et a1 [151 estimated
mean CMRgI, to be 4.76 mg per 100 gm per minute
by the 14C-deoxyglucose autoradiographic method.
Two patients had CMR,], measurements during
the first week after stroke. In Patient C. G. (see Fig
3 ) , mean CMR,,, in the contralateral hemisphere at
seven days was reduced 32% below the normal value
found three months later. The patient was fully
awake during each study. The contralateral depression of glucose utilization in the early study may have
reflected metabolic changes in diaschisis [6, 491, i.e.,
the process of inhibition in neural and vascular tissues remote from the injury site. The second patient
was comatose when scanned five days after suffering
bilateral cerebral embolism and did not survive.
Mean CMR,,, in this patient was reduced 63% below
normal, the same magnitude of LCMk,, depression
which we found in permanent focal infarcts (see Fig
is generally accepted that changes in regional cerebral
activity are accompanied by local variations in metabolism and blood flow functions which are coupled
in healthy brain [46, 501 and may not be in diseased
brain [22, 24, 39, 401. Our ECT scans in stroke patients gave results compatible with expected directions of change, coupling, and uncoupling of local
blood flow and metabolism and suggest good qualitative correspondence of relative local 13NH3 and
'*FDG trapping with relative regional perfusion and
glucose utilization, respectively.
Although relative local cerebral distributions of
13N and LCMk,, were approximately the same in
normal humans (see the Table), animal studies led us
to expect that local I3N trapping would vary nonlinearly with LCBF change ([35] and unpublished results), and there were concerns about the relationship
in diseased tissues [25,26, 32, 351. What we found in
stroke patients was good qualitative correspondence
between 13NH3uptake in diseased zones and other
corroborative information. In acute infarcts, "NH,
uptake was very low during the first two days, when
LCBF should be minimum. Uptake was increased in
infarct margins only during the first two weeks, when
local reactive hyperemia is expected, and in all 3
cases this hyperemia was confirmed by pertechnetate
flow studies. There were no instances of elevation of
local I3N concentration without corroborative evidence of increased LCBF. Uptake was decreased, not
increased, in zones of maximum alteration of the
blood-brain barrier. Uptake was markedly decreased
in all old infarcts that measured larger than 2 cm wide
on XCT. Intermediate decreases in l3NH, uptake
were accompanied by intermediate reductions in
56 Annals of Neurology
Vol 8
No 1 July 1980
LCMR,,, in deactivated or minimally damaged tissues
that appeared structurally intact on XCT scan.
The maximum depression of local glucose utilization occurred many weeks after cerebrovascular occlusion. All zones where "FDG trapping was decreased below DF.lX
= -53% (LCMR,,, = -62%)
were permanent infarcts detectable by XCT. Although decreased metabolism and perfusion were
coupled in late stable infarcts, they were uncoupled
within the infarct during the first two days after
stroke. There was less impairment of glucose utilization than of perfusion during this early period, probably due to transient enhancement of anaerobic
glycolysis 16, 38, 511. Good evidence has been obtained in animal models that cerebral glucose utilization is increased immediately after the onset of
hypoxia L3, 281, but little is known concerning infarct
LCMR,,, during the first few days after stroke in humans. Our earliest scans were made one or two days
after stroke, and these showed no evidence of the
marked zonal increases in LCMR,,, reported by investigators who have used the ''C-deoxyglucose autoradiographic method in animal brains immediately
after an ischemic insult. For example, Ginsberg et a1
[6] found infarcts rimmed by markedly enhanced
glucose utilization 90 minutes after infarcts were induced in cat brain, and Pulsinelli and Duffy [38]
found greatly increased glucose utilization in subcortical white matter 10 minutes following a hypoxic
ischemic insult to rat brain. Scans made earlier than
one day after stroke will be required to test if these
phenomena occur in human beings.
DyJfunction Detected Only by ECT
IuFDG and 13NH, scans consistently detected hypofunction in broad zones of remote cerebral cortex, striatum, and thalamus that appeared structurally
intact on XCT and normal on Y g m T
scans. IsFDG was a more sensitive indicator of this
dysfunction than 13NH3,as expected from the nonlinear response of 13NH3 trapping to changes in
LCBF (1351 and unpublished results). The 18FDGand
l3NNH3studies showed that more brain had been affected by stroke than was suggested by the other
radiological studies. Soh et a1 [49]and Orgogozo et a1
[30] have also found broad zones of dysfunctional but
intact cerebral cortex in cerebral blood flow activation studies using the xenon 133 method in patients
with stroke, and have suggested that undercutting of
cortical fiber tracts is responsible for this deactivation.
Deactivation of the left visual cortex was detected
in 2 patients who had right homonymous hemianopia caused by infarcts restricted to the distribution of
the left middle cerebral artery (see Figs 4, 9). No
ischemia was apparent in the territory of the pos-
left ratio of LCMR,,, in the primary visual cortex for
two adjacent scan IeveIs was 1.01 2 0.04, mean 2
SI>.) Approximately the same reduction in metabolic
rate was found in the visual cortex of normal subjects
with eyes closed, as compared to opened in ambient
room light, both by us (Fig 11) and by Reivich et a1
[411. The scans in these stroke patients showed the
metabolic and perfusion response of the left visual
cortex to destruction of the left optic tract, left lateral
geniculate body, or optic radiations of the left cerebral hemisphere. Therefore, these examples (see Figs
4 , 9, 11) are clinical correlates of the experimental
lesions demonstrated by Kennedy et a1 [ 141 in
metabolic mapping of the visual system of monkey brain
by means of '4C-deoxyglucose autoradiography.
Ipsilateral thalamic dysfunction was found in 5 patients with infarcts restricted to the middle cerebral
artery distribution, even though the primary thalamic
blood supply comes from the posterior cerebral circulation (see Fig 2, Group D; Figs 3, 4 , 7, 8). Four
of these patients were aphasic, 3 had contralateral
sensory loss, and none experienced the thalamic pain
syndrome. These scan findings may have resulted
from ischemic necrosis due to compression of
thalamic circulation by postinfarction edema of
perithalamic tissue, wallerian degeneration of
thalamic neurons efferent to the region of cortical
infarction, decreased excitation of thalamic neurons
secondary to a loss of afferent neurons from the region of cortical infarction, or combinations of these
F i g 11. Typical deartioation of primary visual cortex (PVC)
with eye closure in a normal subject (leuel2B). I n the first
study, the subject bad eyes open t o ambient room light for 30
minutes following 'RFDGinjection; i n the second .stndy the
eyes remained closed. Zonal metubalic rates were normalized to
a tmalue o f 100 for the initial PVC measi~rement.Eye closure
reduced PVC metabolic rate by 14$6 bat made little chunge in
other zones. Note the difficulty i n vimalzzing sucb a small
metubolic change without benefit of side-to-side asymmetry, as
.reen in Figures 4 and 9. (FLC = frontul lobe cortex; PLC =
parietal lobe cortex; OLC = occipital lobe cortex; FJlr =frontal
white matter: OW = occipital white matter.) For anatomical
detuils see Figure 16 in DeArmord e t a l [ 4 ] .
terior cerebral artery, which supplies the visual cortex. In each patient the ECT scan demonstrated de-
creased glucose utilization and perfusion in the left
primary (area 17) and associative (areas 18, 19)visual
cortex, which appeared structurally normal on X CT
scan. Compared to the opposite side, LCMR,],. in the
ipsilateral primary visual cortex was decreased by
about 15%. (In 6 normal subjects the mean right to
These studies of the brain in normal subjects and of
the contralateral hemisphere in recovered stroke patients show that absolute values of mean CMR,,, determined by the lsFDG technique agree well with
published values determined in normal brain by
the Kety-Schmidt method. In abnormal brain,
quantification was restricted by our limited understanding of tracer behavior in diseased tissues; but
qualitatively, relative local distributions of "FDG
and I3NH3trapping reflected the increases and decreases as well as coupling and uncoupling expected
for local alterations of glucose utilization and perfusion in stroke. The ECT scans of 18FDG and 13NH3
showed that stroke caused widespread physiological
alterations in brain tissues that appeared to be structurally intact. This suggests that the ECT method,
even in its early state of development, may be quite
useful in defining the location and extent of altered
brain in the study of disordered function, such as
aphasia, after stroke. Also, these scans might aid in
early identification of threatened but potentially recoverable cortex and, later, in estimation of surviving
cortex. For example, in Patient M. R. (see Fig 4),
Kuhl et al: IXFDG ECT for Detecting Brain Damage in Stroke
early XCT scans showed widespread decreased density throughout the ischemic left hemisphere that
could have represented either necrosis or edema, but
the "FDG scans showed functioning, and therefore
potentially recoverable, left cerebral cortex. In the
same patient, 18FDGscans at two months showed the
overall extent and magnitude of permanent residual
damage to the entire left cerebral hemisphere,
whereas XCT scan showed only the more limited lesion where focal structural damage was maximum.
The future usefulness of ECT in estimating the
prognosis for local cerebral tissues after stroke depends in large part on improvements in radiopharmaceuticals and increased knowledge of their behavior in diseased tissue. A tracer is needed that will
respond consistently and linearly to LCBF in normal
and abnormal brain. Better knowledge of the kinetic
behavior of 18FDG in an early infarct is required, as
is the ability to distinguish aerobic and anaerobic
glycolysis, perhaps by ECT of radioactive oxygen 15
[ 11 proximate with 18FDG. For example, a local metabolic disruption causing heightened "FDG trapping
could be associated with decreased or increased
oxidative metabolism, yet the local prognosis might
be different in each instance. With such advances, the
ECT method could permit us to study in human beings those measurements of local cerebral metabolic
and circulatory physiology previously restricted to
animal models of stroke [54]. This capability would
prove useful in improving our understanding of how
the human brain responds to stroke and for mapping
subsequent progressive alterations with and without
therapeutic intervention.
Supported by Department of Energy Contract DE-AM03-76SF00012 and by US Public Health Service Research Grants
7R01-GM-24839 and 1P01-NS-15654.
Presented in part at the Eighth International CBF Symposium in
Copenhagen, Denmark, June 1977 [22].
We thank E. J. Hoffman, PhD, S. C. Huang, DSc, N. S. MacDonald, PhD, G. D. Robinson, Jr, PhD, J. Miller, NMRT, F.
Aguilar, NMRT, and A. Ricci, AB, for assistance in the research
effort reported here.
Operational equation for calculation of LCMR,,, (from
Phelps et al L371 and Huang et a1 [ 111):
Cp (CF(T) LCMR,,, =
ffz -
0.130; white, 0.109) capillary membrane transport,
respectively (min-')
are first-order rate constants for phosphorylation
of lRFDG(gray, 0.062; white, 0.045) and dephosphorylation of lRFDG-6-P04(gray, 0.0068; white,
0.005€9, respectively (min-')
is a lumped constant (0.420)-the
ratio of arteriovenous extraction fraction of 18FDG to that of
glucose under steady-state conditions and when :
is small
is cerebral tissue concentration of "FDG plus
IRFDG-6-PO, in region i at a single time (T), determined by ECT scan
is capillary plasma glucose concentration (steady
by average value from peripheral artery o r vein
is capillary plasma "FDG concentration as a function of time-approximated
by values from peripheral artery or vein
are rate constants for the model response to an impulse change in lsFDG capillary plasma concentration-a,: gray, 0.0046; white, 0.0041; az:gray,
0.194; white, 0.156
metabolism = perfusion
metabolism > perfusion
perfusion > metabolism
[(k;- a,)e-"lt + (az- k:)e-a2t]
percentage differences existing
between selected zones and corresponding contralateral zones
for mean values of ''N concentration, "F concentration, and
LCMR,,,, respectively
emission computed tomography
full width at half maximum, a
measure of spatial resolution
lumped constant
local cerebral blood flow
Local cerebral metabolic rate for
coupled relationship in which
local metabolism and local perfusion are equally decreased
uncoupled relationship in which
local metabolism exceeds local
uncoupled relationship in which
local perfusion exceeds local
primary visual cortex
x-ray computed tomography
@ Cb(t))
a 2
0 cg(t)
denotes operation of convolution
k:, k: are first-order rate constants for lRFDGforward
(gray, 0.102; white, 0.054) and reverse (gray,
58 Annals of Neurology Vol 8 No 1 July 1980
1. Ackerman RH, Alpert NkI, Correia JA, et al: Correlations of
positron emission scans with TCT scans and clinical course.
Acta Neurol Scand 6O:suppl 72:230-231, 1979
2. Alcala H, Gad0 M, Torack R: The effect of size, histologic
elements, and water content on the visualization of cerebral
infarcts. A computerized cranial tomographic study. Arch
Neurol 35:l-7, 1978
3. Cohen PJ, Alexander SC, Smith TC, et al: Effects of hypoxia
and normocarbia on cerebral blood flow and metabolism in
conscious man. J Appl Physiol 23:183-189, 1967
4. DeArmord SJ, Fusco MM, Dewey MM: Ftructure of the
Human Brain, a Photographic Atlas. Second edition. New
York, Oxford University Press, 1976, p 30
5. Freygang W H Jr, Sokoloff L Quantitative measurement of
regional circulation in the central nervous system by the use
of radioactive inert gases. Adv Biol Med Phys 6:263-279,
6. Ginsbetg MD, Reivich M, Giandomenico A, et al: Local glucose utilization in acute focal cerebral ischemia: local dysmetabolism and diaschisis. Neurology (Minneap) 27: 10421048, 1977
7. Gottstein E, Bernsmeier A, Sedlmeyer 1: Der Kohlenhydratstoffwechsel des meschlichen Gehirns. Klin Wochenschr
41~943-448, 1963
8. Greenberg JH, Alavi A, Reivich M, et al: Local cerebral blood
volume 'response to carbon dioxide in man. Circ Res
43~324-331, 1978
9. Grubb RL Jr, Raichle ME, Higgins CS, et al: Measurement of
regional cerebral blood volume by emission tomography. Ann
Neurol 4:322-328, 1978
10. Hoffman EJ, Huang SC, Phelps ME:Quantitation in positron
emission computed tomography: 1. Effect on object size. J
Comput Assist Tomogr 3:299-308, 1979
11. Huang SC, Phelps ME, Hoffman EJ, et al: Non-invasive determination of local cerebral metabolic rate of glucose in normal man with (F-18)2-fluoro-2-deoxyglucoseand emission
computed tomography: theory and results. Am J Physiol238:
E69-E82, 1980
12. Hunter W, Monahan WG: 13N-Ammonia: a new physiologic
radiotracer for molecular medicine. J Nucl Med 12:368,
197 1
13. Ido T, Wan CN, Casella V, et al: Labeled 2-deoxy-D-glucose
2-deoxyanalogs. 18-F-labeled-2-deoxy-2-fluoro-~-glucose,
2-fluoro-p-mannose and 14-C-2-deoxy-2-fluoro-D-glucose.
Label Comp Radiopharm 14:175-183, 1978
14. Kennedy C, DesRosiers MH, Sakurada 0, et al: Metabolic
mapping of the primary visual system of the monkey by means
of the autoradiographic [14C] deoxyglucose technique. Proc
Natl Acad Sci USA 734230-4234, 1976
15. Kennedy C , Sakurada 0, Shinohara M, et al: Local cerebral
glucose utilization in the normal conscious macaque monkey.
Ann Neurol 4:293-301, 1978
16. Kety SS, Schmidt CG: The determination of cerebral blood
flow in man by the use of nitrous oxide in low concentrations.
Am J Physiol 143:53-66, 1945
17. Kuhl DE, Alavi A, Hoffman EJ, et al: Local cerebral blood
volume in head-injured patients: determination by emission
computed tomography of Tc-99m red cells. J Neurosurg 52:
309-322, 1980
18. Kuhl DE, Edwards RQ: Image separation radioisotope scanning. Radiology 80:653-662, 1963
19. Kuhl DE, Edwards RQ, Ricci AR, et al: The Mark IV system
for radionuclide computed tomography of the brain. Radiology 121:405-413, 1976
20. Kuhl D , Engel J, Phelps M, et al: Patterns of local cerebral
metabolism and perfusion in partial epilepsy by emission
computed tomography of I*F-fluorodeoxyglucose and 13Nammonia. Acta Neurol Scand 6O:suppl 72:538-539, 1979
2 1. Kuhl DE, Hoffman EJ, Phelps ME, et al: Design and application of the Mark IV scanning system for radionuclide tomog-
3 1.
raphy of the brain, in Medical Radionuclide Imaging. International Atomic Energy Agency Symposium on Medical
Radionuclide Imaging, Los Angeles, 1976. Vienna, International Atomic Energy Association, 1977, vol l, pp 309-320
Kuhl DE, Phelps ME, Hoffman EJ, et al: Initial clinical experience with L8F-2-deo~y-D-gl~~ose
for determination of local
cerebral glucose utilization by emission computed tomography. Acta Neurol Scand 56:suppl 64:192-193, 1977
Kuhl DE, Reivich M, Alavi A, eta]: Local cerebral blood volume determined by three-dimensional recorptruction of
radionuclide scan data. Circ Res 36:610-619, 1975
Lassen NA: The luxury-perfusion syndrome and its possible
relation to acute metabolic acidosis localized within the brain.
Lancet 2:1113-1115, 1966
Lockwood AH: Cerebral ammonia metabolism (letter).
Stroke 9:521, 1978
Lockwood AH, Finn RD, Campbell JA, et al: Factors affecting
uptake of 13N-ammoniaby the brain. J Comput Assist Tomagr
2~654-655, 1978
Monahan WG, Tilbury RS, Laughlin JS: Uptake of I3Nlabelled amm9nia. J Nucl Med 13:274-277, 1972
Norberg K, Siesjo BK: Cerebral metabolism in hypoxic
hypoxia. I. Pattern of activation of glycolysis: a reevaluation.
Brain Res 8:3!-44, 1975
Novak P, Goluboff B, Bortin A , et al: Studies of the cerebral
circulation and metabolism in congestive heart failure. Circulation 7:724-731, 1953
Orgogozo JM, Larsen B, Skyhoj T, et al: Evidence of cortical
disconnection in deep hemispheric strokes as revealed by
rCBF. Acta Neurol Scand 6O:suppl 72:258-259, 1979
Phelps ME: Emission computed tomography. Semin Nucl
Med 7~337-365, 1977
Phelps ME: Cerebral ammonia metabolism (letter). Stroke
9:521-522, 1978
Phelps ME, Hoffman EJ, Coleman RE: Tomographic images
of blood pool and perfusion in brain and heart. J Nucl Med
17:603-612, 1976
Phelps ME, Hoffman EJ, Huang SC, e t al: E C A T a pew computerized tomographic imaging system for positron-emitting
radiopharmaceuticals. J Nucl Med 19:635-647, 1978
Phelps ME, Hoffman EJ, Rayband C: Factors which affect cerebral uptake and retention of ''NH,. Stroke 8:694-702,
Phelps ME, Huang SC, Hoffman EJ, et al: Tomographic measurement of cerebral blood volume with C-11 labeled carboxyhemoglobin. J Nucl Med 20328-334, 1979
Phelps ME, Huang SC, Hoffman EJ, et al: Tomographic measurement of local cerebral glucose metabolic rate in humans
validation of method.
with (F-l8)2-fluoro-2-deoxy-D-glucose:
Ann Neurol 6:371-388, 1979
Pulsinelli WA, Duffy TE: Local cerebral glucose metabolism
during controlled hypoxemia in rats. Science 204:626-629,
Raichle ME, Grubb
Gado MH, et al: Correlation between
regional cerebral blood flow and oxidative metabolism. In
vivo stydies in man. Arch Neurol 33:523-526, 1976
Reivich M: Blood flow metabolism couple in brain, in Plum F
(ed): Brain Dysfunction in Metabolic Disorders. New York,
Raven, 1974, pp 125-139
Reivich M, Greenberg J, Alavi A, et al: The use of the I'Ffluorodeoxyglucose technique for mapping of functional
neural pathways in man. Acta Neurol Scand 6O:suppl 72:
198-199, 1979
Reivich M, Jehle J, Sokoloff L, et al: Measurement of regional
cerebral blood flow with antipyrine-14C in awake cats by an
autoradiographic technique. J Appl Physiol 27296-300,
Kuhl et al: lSFDG ECT for Detecting Brain Damage in Stroke
43. Reivich M, Kuhl D, Wolf A, et al: Measurement of local cerebral glucose metabolism in man with 18-F-2-fluoro-2-deoxy~-gIucose.Acta Neurol Scand 56:suppl 64:170-171, 1977
44. Reivich M, Kuhl D, Wolf A, et al: The ["F] fluorodeoxyglucose method for the measurement of local cerebral glucose
utilization in man. Circ Res 44:127-137, 1979
45. Robinson GD Jr, MacDonald NS, Easton MP, et al: I'Ffluorodeoxyglucose:rrmote, semiautomated production using
a compact cyclotron. J Nucl Med 19:701-702, 1978
46. Roy CS, Sherrington MB: On the regulation of the blood
supply of the brain. J Physiol (Lond) 11:85-108, 1890
4?. Scheinberg P, Stead EA: The cerebral blood flow in male
subjects as measured hy the nitrous oxide technique. Normal
values for blood flow, oxygen utilization, glucose utilization
and peripheral resistance, with observations on the effect of
tilting and anxiety. J Clin Invest 28:1163-1171, 1949
48. Slater R, Reivich M, Goldberg H, et al: Diaschisis with cerebral infarction. Stroke 8684-690, 1977
60 Annals of Neurology Vol 8 No 1 July 1980
49. Soh K, Larsen B, Skinhof E, et al: Regional cerebral blood
flow in aphasia. Arch Neurol 35:625-632, 1978
50. Sokoloff L, Reivich M, Kennedy C, et al: The ( I T ) deoxyglucose method for the measurement of local cerebral glucose
utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J Neurochem 28:877-916,
51. Spatz M, Mrsulja BB, Micic D, et al: Ischemic and postischemic effects on 2-de0xy-D-[~H] glucose uptake in cerebral
capillaries. Brain Res 120:141-145, 1777
52. Takeshita H, Okuda V, Sari A: The effects of ketamine on
cerebral circulation and metabolism in man. Anesthesiology
36:69-75, 1972
53. Vaalburg W, Kamphius JA, Beerling-van der Molen HB, et al:
An improved method for the cyclotron production of IJNlabelled ammonia. Int J Appl Radiat Isot 26:316-318, I975
54. Waltz AG: Clinical relevance of models of cerebral ischemia.
Stroke 10:211-213, 1979
Без категории
Размер файла
1 438 Кб
local, stroki, effect, metabolico, mapping, 18fdg, tomography, emissions, 13nh3, perfusion, computer, cerebral
Пожаловаться на содержимое документа