- Email: [email protected]
Age-related changes in cerebral MR angiography
J O U OFTHE R N A L
NEUROLOGICAL SCIENCES Journal of tfre Neurological Sciences 145 (1997) 195-203
Age-relatedchanges in cerebral MR angiography Yoriko Kajiya *, YoshikiKajiya,MasayukiNakajo Deparrrrrentof Radiology, Faculryof Medicine, Kagoshima University, 8-35-I, Sakuragaoka,Kagoshima 890, Japan Received 3 January 1996; revised 21 May 1996; accepted 23 May 1996
Abstract Two hundred forty four neurologically normal subjects (118 males and 126 females, aged 7 to 82 yrs.) were evaluated to assess the correlation of the degree of visualization of cerebral arteries on magnetic resonance angiography (MRA) with aging, sex, literality or lacunar infarctions on magnetic resonance imaging (MM). A single volume three-dimensional transformation time-of-flight pulse sequence with slab thickness of”115 mm was used to obtain axial, saggital and coronal projection images. A total of 6 arteries including bilateral anterior cerebral arteries (ACAS), middlecerebralarteries(MCAS)andposteriorcerebralarteries(PCAS)foreverysubjectwere
eachgradedinto8 scoresby thedegreeof visualization of thearteries.Resultsshowthatan age-relateddeclineof visualization of the cerebralarteriesin ACA(r = –0.603, p < 0.001), MCA(r = – 0.452,p < 0.001) andPCA(r= – 0.537,p < 0.001). The arterieswere bettervisualizedin femalesthan malesduringthe fifth decade(p < 0.01). No substantial relationship was observed between the visualization and sex in the other decades, literality or lacunar infarctions. Thus, the cerebral MRA showed an age-related decline of visualization of the arteries and a sex-difference of the visualization in the fifth decade, probably due to the decline of the blood flow velocity with aging and a difference in it between females and males in the fifth decade. Published by Elsevier Science B.V. Keywords; Magnetic resonance angiography (MRA); Cerebral artery; Age; Sex; Lacunar infarction
1. Introduction
Many studieshave documentedthe clinicalefficacyof a three-dimensionaltime-of-flightMRA in the evaluationof intracranialvascular pathology(Korogi et al., 1994; Yamada et al., 1995;Heisennanet al., 1992).To our knowledge, however, there have been no reports evaluating age-relatedchanges in cerebral MRA in many neurologically normal subjects.In this article, we explore age-related changes in cerebral MRA. We also study about the effects of sex, literality and lacunar infarctionson the visualizationof the cerebralarterieson MRA.
2. Materials and methods 2.1. Patients
Two hundred forty four patients(126 females and 118 males)aged 7 to 82 yearswere includedin this study,who
‘ Correspondingauthor. Tel.: +81 (992) 75-5417. Fax: +81 (992) 75-5420. O022-510X/97/$17.00 Published by Elsevier Science B.V. PII S0022-5 10X(96)O0 199-2
received MR examinationsbetween December 1990 and June 1991. Two-thirdsof them underwentMR examinations because of headache,dizziness,paresthesiaor other symptomes,but who showedno specificneurologicaldisorders and one third of them underwent for a thorough cerebralcheck up. They had no tumor,hemorrhage,infarction except lacunar infarction and no vascular pathology such as intracranialaneurysms. 2.2. Methods of MRA and MRI The MR examinationswere performedon a 1.5-Timaging system(Signa; GE Medical System,Milwaukee).The single-volumethree dimensional Fourier transformation (3DlW) time-of-flight(TOF) (Wehrli, 1990) MR angiographic images were obtainedwith a gradientecho pulse sequencewith ‘spoiling’of residualtransversemagnetization known as spoiledgradient-recalledacquisitionin the steady state (Prorok, 1989).The imagingparameterswere 30/1 1/1 (TR/TE/excitations) using a 20° flip angle.No presaturation pulse was used. Flow compensationwas performed in the read and phase encoding directions.A 256 X 192matrix was used, with a field of view of 22 cm and a partitionthicknessof 0.9 mm, resultingin a 0.9 X
196
Y.Kajiya et al. /Jourrral of the Neurological Sciences 145 (1997) 195–203
1.1x 0.9 mm voxel. One hundred twenty eight partitions with a section thicknessof 115 mm to cover almost the whole brain were acquired in an imagingtime of 12 min 18 s. The 128 axial source images obtained in this way were postprocessedby means of a maximum-intensity pixel projectionalgorithmto produce multiple projection rotated about the section axis. We used 3 images; axial, sagittaland coronalprojectionimages. T1- and T2-weightedspin-echoimages were obtained to get axial and sagittalMR images.Repetitiontime, echo time and excitation were 400 msec, 11 msec and two (400/1 1/2) for T1-weightedimagesand were 2200/90/1 for T2-weightedimages.An imagematrix was 256X 192.
10.
U* 5 A : ACA Y= -O. 11X+13(.=2U) (7=-0 .s03. P
1
2
345a
789
131
I
H
G
!5-
not {8) !61 us
041 U*
I I 1
678
aO
[58
8: MCA V= –0.LWOX+13(.=244)
E
(r=–0.452,
3
01
1
2
p< O.@ll)
3456789
2: A B
?
A : ACA C : PCA Y= -0. ffi4x+ II (n=244) (r= -0.537, P< O.001)
-1....u
3/--’).
$“--’& ...
f’-&
o!
4
5 [h ,.,,
34
56
7a9
range
Fig. 2. Age-related changes in visualization of the cerebral arteries on ACAS (A), MCAS (B) and PCAS (C). Solid circles with bars show mean scores with SD.S. The number in parenthesis below the bar is number of patients. X: Actual age, Y: Score of visualization of arteries. Y declined with age rapidly in the ACAS and gradually in MCAS and PCAS.
--)2 [~ ~ , :,
2
Age
B : MCA
!,-:
1
67
the the the the the
{4 ,., ,!, :
8
c:PCA
Fig. 1. The methods of estimation of the degree of cerebral arterial visualization. Scoring of the arterial visualization was made for the ACA on a 3D MRA sagittal projection image as illustrated with the scheme (A) artd for the MCA and pCA on a 3D MRA axial projection image as illustrated with the schemes (B) and (C). The solid and broken lines represent the linear and beaded arteries, respectively. See text about the details of scoring criteria.
2.3. Scoring of the degree of visualization of the cerebral arteries on MRA
Gradingof the cerebral arteriesis shown schematically in Fig. 1. Each vessel has a uniquescoring systemwith a maximumscore of 8 per vessel. The ACA was estimated on a sagittalprojectionimage. A coronalprojectionimage
Y. Kajiya et aL/Journal of the Neurological Sciences 145 (1997) 195–203
was used to decidewhetherthe arterieslocatedin the right or left hemisphere.The ACA was evaluatedat 8 points as illustrated in Fig. 1A. The point A was set where the internal cerebral artery went upward after the carotid siphon. The upper and lower points of the pericallosal artery in front of the lamina terminals were set as C and
197
pointcorrespondingto the genu of corpuscallosum was set as E. The D point was set on the positionwhere the horizontaldistancebetweenC and D was equal to that between A and B. The point where the extension of BC intersectedthe pericallosalartery was set as G. The points F, H and I were set to make the horizontaldistancesequal
B. The
(A)
Fig. 3. Age-related changes in cerebral MRA of neurologically normal patients a: sagittal pro~ction images and b: axial projection images. All six arteries are excellently demonstrated and the mean scores of bilateral ACAS, MCAS and PCAS are 14, 16 and 14 in a 7-year-old girl (A). The cerebral arteries are moderately visualized and the scores of bilateral ACAS, MCAS and PCAS are 10, 11 and 10 in a 43-year-old woman (B). All the arteries, especially ACAS, w PoorlY visualized and tie scores of bilateral ACAS, MCAS and pCAS are o, 1’ and 6 in a 81-year-old woman (C). These images demonstrate that visualization of the cerebral arteries declines with aging.
198
Y.Kajiya et al. /JournaI of the Neurological Sciences 145 (1997) 195–203
betweenEF, FG, GH and HI. The score was made as 1, 2, ..., 7 when the tip of the visualizedvessellay betweenBC, CD, ..., HI, respectively. The score was 8 when the pericallosal artery was visualized clearly beyond the I point. The MCA was estimatedon the axial projectionimage. We divided the MCA into four segments, the sphenoid (Ml) segment, the insular (M2) segment, the opercukir
(M3) segmentand the terminal(M4) segment.Each segment was rather long and had a differencein appearance; the smoothlyand clearlyvisualizedvesselsand beadedand faintly visualized vessels. Therefore in scoring of the MCA, we used two parameters including the terminal point and clearness of the visualized vessel. When the terminalbranch lay on Ml segment and was beaded and faint, the scorewas 1. If it was smoothand clear, the score
(B)
Fig. 3 (continued).
Y. Kajiya et al. /Journal of the Neurological Sciences 145 (1997) 195-203
was 2. The scores were 3 and 4 on M2 segment,5 and 6 on M3 segment,and 7 and 8 on M4 segmentin the same way. The PCA was also estimated on the axial projection image and divided into 8 scores by the place and the scoringsystemwas the same used in the MCA. When the terminal branch was found before the posterior temporal
199
artery branchsoff, the degree of visualizationwas graded as 1 or 2, and graded as 3 or 4 before the parietooccipital artery and the calcarinearterybranchoff, graded 5 or 6 on these arteriesand 7 or 8 on their branches. The ACA, MCA and PCA were separately scored in both hemispheresand the scores were summed. A total score of the abovesix arterieswas set as [T]. Theoretically
(c)
Fig. 3 (continued).
Y. Kajiya et al. /Journal of the Neurological Sciences 145 (1997) 195-203
fifth decade (Dozono et al., 1991). So we estimated the lacunarinfarctionsin the patientsof sixthto ninethdecades (n= 164).Lacunarinfarctionshave been definedto be old small deep infarctions,0.5 to 15 mm in diameter(Fisher, 1965).However, dilated perivascularspace cannot be excluded. Therefore we estimated the lesions larger than 5 mm in order to excludethe perivascularspace, which was usuallyunderthan 5 mm in diameter(Jungreiset al., 1988; Braffmartet al., 1988).The high intensitylesions on T2 weighted images were scored as O–2 based on their size (O= none, 1 = 5-10 mm in diameter, 2 = 11-15 mm in diameter)and the gradeswere summed.Degree of lacunar infarctions were evaluated in the following three parts; basal ganglia and thalamus, cerebral white matter, and pens and cerebellum.The sum was made for each part in each patient. The maximumgrades of Iacurtarinfarction were 7, 12 and 2 in basal ganglia and thalamus,cerebral white matter, and pens and cerebellum,respectively.The number of the patients without lacunar infarctions was 121, 109 and 152,respectively.
(2)
0
(3)
6
(;)
(4)
1
1I
(7)&9 (4)
(13) (3)
IJ t I
(9)
@l)
or)
=o~
(2)
=
I
2
4
3
5
6
7
8
9
+ Age
range
Fig. 4. Sex differences in visualization of the cerebral arteries in neuroIogically normal males (0) and females (0). [T] is tie total score of rhe six arteries. Circles with bars show the mean ratios with S.D.S. In the fifth decade, the arteries were visualized significantly better in females than in males.
[T] can be scored from 6 to 48. However,it was actually scored from 6 to 43. Two radiologists(Y.K. and Y.K.) evaluated independentlyMRA and MR images of each patient, using the above mentionedscoring systems.The difference of scores was always one score. When the assessmentdiffered, a consensus interpretationwas obtained.
2.5. Stratijcation of age The ages of patientswere divided into 9 ranges based on the decade: 1 (O–9-year-old;n = 3), 2 (10–19-year-old; n = 10), 3 (2f)-29-year-o]d; n = 8), 4 (3t)-39-year-old; n = 16), 5 (4.@49_yem.o]d;n = 43), 6 (50–59-year-old; n = 56), 7 (6&69-yew-old; n = 73), 8 (70–79-year-old; n = 30) and 9 (80–89-year-old;n = 5).
2.4. Grading of lacunar infarctions Lacunarinfarctionswere foundin the patientsabovethe age of 40 years, however were not commonlyseen in the
PCA
MCA
ACA
6 T
5
I I
4
3
2
1
c
6767
6767
6767 Age
range
Fig. 5. Literality of visualization of the cerebral arteries between the right and left hemispheres. ●, Right hemisphere and O, left hemisphere of the male (n= 27 in the sixth decadeand 35 in the seventh decade), A; the right hemisphere and A; the left hemisphere of the female (n = 29 in the sixth decade and 38 in the seventh decade). Circles or triangles with bars show the mean ratios with S.D.S. There were no statistically significant differences in the visualization between the right and left hemispheres.
201
Y. Kajiya et aL/Journal of the Neurological Sciences 145 (1997) 195–203
2.6. Analysis of data Visualizationof cerebral arteries on MRA was compared to the age. First, the linear correlationcoefficient was obtainedbetweenthe actual age (X) and the sum (Y) of the scoresof rightand left arteriesof the ACA, MCA or PCA and its significancewas calculated. Secondly, the mean score+ S.D. of the sumof right and left arterieswas calculated in each age range. Maximum and minimum scoresof Y were 16 and 2. A sex differencewas estimatedwith the mean score+ S.D. of [T] using unpaired t-test in each age range. The right-leftdifferencein the ACA, MCA or PCA was estimatedwith the mean score+ S.D. of each artery(maximum and minimumscores were 8 and 1) using paired t test. Multiple regression analysis with visualizationof the cerebralarterieson MRA as a dependentvariablewas used to investigatethe dependencyon age and lacunar infarctions (independentvariables). The statisticalsignificantlevel was assumedto be less than 0.05. 3. Results The relation of visualizationof the cerebral arteries to age is shown in Fig. 2. A statisticallysignificantnegative linear correlationexisted between the actual age (X) and visualizationof the cerebral peripheral arteries (Y) (r= –0.603, p <0.001 for age vs the ACAS, r = –0.452, p <0.001 for age vs the MCAS, r = –0.537, p <0.001 for age vs the PCAS).The age-relateddeclinewas rapid in the ACAS and gradual in the MCAS and PCAS. Three instancesof the first, fifth and ninth decadesare shownin Fig. 3. Projectedthree-dimensionalTOF MR images of a 7-year-oldgirl demonstratedexcellentlyboth central and peripheralarteries. However, the peripheralarteries were moderately detected in a 43-year-old woman and they were poorlydetectedin a 8l-year-old woman. The relation of visualizationof the cerebral arteries to sex is shownin Fig. 4. For the total population,there was no statisticallysignificantdifferencebetweenmales([T]= 24.4 + 6.3, n = 118) and females ([T]= 24.9+ 6.3 n = 126). However, in the fifth decade, the visualizationwas higher (p < 0.01) in females ([T]= 28.3 ~ 4.8, n = 22) than in males([T]= 24.3+ 4.6, n = 21). No statisticallysignificantright-left differencesin the ACAS,MCASand PCASwas found in any decadeswhich had a sufficientnumberof patientsfor statisticalanalysis. Fig. 5 shows the results of literality in the sixth and seventhdecades. A multipleregressionanalysisof the data in the group of 50 or more years old patientsis shown in Table 1. [T] was set as dependentand age and lacunarinfarctionswere set as independent.Therewas a statisticalsignificantcorre-
ctable 1 Multiregression analysis with [T] as dependent using 4 factors of age (Xl), lacunar infarctions in basat ganglia and thafamus (X2), cerebral white matter (X3), and pens and cerebellum (X4) as independent Factors
xl x2 x3 x4
[T] Simple regression p-Value Coefficient -0.323
-0.220 -0.080
–0.047
p p p p
<0.001 <0.01 >0.05 >0.05
Multiple regression p-Value Coefficient –0.285 –0.148 0.024
0.019
p <0.001 p >0.05 p >0.05 p >0.05
lationbetweenage and [T]( p < 0.001).It revealedthatthe increaseof lacunar infarctionsof basal gangliaand thalamus was concomitantof age-dependentdeclineof [T], and there was no statisticallysignificantrelationbetweenlacunar infarctionsand visualizationof the cerebralarteries. 4. Discussion Various approachesto MRA have recently been proposed and have already been shown to be useful in the clinicalevaluationof cerebralvasculardiseases(Korogiet al., 1994; Yamada et al., 1995; Heisermanet al., 1992). However, most of them are investigationsabout given diseases.Very few studieswere reportedaboutthe cerebral arteries themselves using MRA (Portman et al., 1992; Mattleet al., 1991).Marks et al. describedan age-related declinein mean velocityof the internalcarotid artery and basilar artery (Marks et al., 1992).An age-relateddecline in mean velocityof the internalcarotid artery and basilar artery is so far available(Marks et al., 1992).However, there have been no reports dealing with the relation between aging and visualizationof the cerebral peripheral arterieson MRA in many neurologicallynormal subjects. The presentstudyrevealedthat the degreeof visualization of the cerebralperipheralarteriescorrelatednegatively with age. The declinewas rapid in the ACASand gradual in the MCAS and PCAS. The flow-relatedenhancement obtainedwith the singlevolume3DlWTOF is veryhigh at the point that a vessel enters the imaging volume. The inflowingspinsundergosaturationas the bloodtransverses the region of excitation,resulting in progressivelydiminishing contrast (Lewin and Laub, 1991). The ACA runs relativelyparallel to the long axis of body and intersects the region of excitation.Therefore, the slow flow in the ACA tendsto decreasethe signalof inflowingbloodmore obviouslythan those in the MCA and PCA. Studies using transcranial pulsed Doppler technique showed that the measured blood velocity of the middle cerebralartery decreasedsignificantlywith increasingage (Vriens et al., 1989; Ackerstaff et al., 1990). With this single 3DIT volume method, the slow flow cannot span the entire volume before becoming sufficientlysaturated,
202
Y.Kajiya et aL/Journal of the Neurological Sciences 145 (1997) 195-203
and results in signalloss (Lewin and Laub, 1991).Standing on the above,the age-relateddecreasein visualization of the cerebral arteries seems to be caused by age-related decreaseof the flow velocity.The singlevolumetechnique is less sensitiveto visualizea slowflow than the sequential 2-D slicemethodand multiplethin volumemethod(Lewin and Laub, 1991).Usually42 to 72 mm thicknessof slab is used (Yamadaet al., 1995;Heisermanet al., 1992;Davis et al., 1994)to avoid signal loss caused by saturationof the flow in the slab. Howeverin this studywe used a very thick single volume of 115 mm to cover almost all the intracranialvessels,which decreasedsensitivityto vessels which containedslowlymovingblood,resultingin making the more prominentdifferencein visualizationof the cerebral arteries between the patients. Using the usual slab thickness,we would not have such differenceobservedin this study. Recentlywe can use the multipleoverlappingthin slab acquisitiontechnique(MOTSA)which has decreasedsensitivity to the effects of flow saturation and allows improvedvisualizationof the intracranialvascularabnormalities (Davis et al., 1994).Using this technique,we will get better images even in elder people and will not have so distinctdifferencecompareto youngerpeople. Arteriosclerosismay be another factor that influences visualizationof the cerebral arterieson MRA. The severe stenosiswith atherosclerosiswas reportedto cause a segmental decreasein visualizationof the carotid arterieson MRA (Laster et al., 1993). Atherosclerosismay also decrease the flow velocity. We examinedthe relation betweenvisualizationof the cerebral arteries on MRA and lacunar infarctionswhich have art associationwith atherosclerosis(Dozono et al., 1991; Braffman et al., 1988). There was no statistically significantrelation between visualizationof the arteries and lacunar infarctions.It has been shown that MRI receives clear effects from flow in large vessels, but it is difficultto measuretissueperfusionby MR studies(Axe], 1984). It may be the reason why there was no relation betweenvisualizationof the cerebralarterieson MRA and the number of Iacunar infarctions which are ischemic change of the very small arteries. Even without atherosclerosis,vesselsbecomemoretortuouswith aging(Adams and Victor, 1993).It may also causethe TOF effect down. Femalesfrom the third to fifth decadesshowedgreater scoresof the visualizationthan males,and the fifth decade femaleshad statisticallysignificantbettervisualizationthan males.The sex-relateddifferencesin humancerebralblood flow and blood flow velocity have been described by several investigatorsusing Xenon inhalationtechniqueor transcranial Doppler examination (Vriens et al., 1989; Ackerstaffet al., 1990;Shawet al., 1984).They foundthat premenopausalfemales had significantlyhigher cerebral blood flow valuesthan their male counterparts.It has been shown that sex differencein frequencyof cerebralatherosclerosis appears from the fourth decade and the largest
sex differenceis found in the fifth decade, indicatingthat the males have considerablyhigher degrees of cerebral atherosclerosisthan the femalesin these age groups(Flora et al., 1968).These reportswill supportour result. There were also no statisticallysignificantright–left differencesin visualizationof the cerebral arteries. In the study using transcranialDoppler examination,it has been noted that the mean blood velocityof the MCA shows no right–leftdifferences(Ackerstaffet al., 1990). We use MRA in order to get informationabout the morphologyof the vessels. Conventionalarteriographyis superiorto MRA in visualizingthe vessels.However, the techniqueof MRA can be used in additionto routineMRI and is noninvasive.In addition, in this study MRA was considered to reflect the flow velocity like transcranial Doppler and Xenon inhalationmethods.MRA may show the natural state of vesselsbecause it is not necessaryfor MRA to inject a contrastmediumby force.
References Ackerstaff, R.G.A., R.W.M. Keunen, W. van Pelt, A.D.M. van Swijndregt and T. Stijnen (1990) Influence of biological factors on changes in mean cerebral blood flow velocity in normal ageing: a trarrscranial Doppler study. Neurol. Res., 12: 187-191. Axel, L. (1984) Blood flow effects in magnetic resonance imaging. A.J.R., 143: 1157-1166. Braffman, B.H., R.A. Zimmerman, J.Q. Trojarrowski; N.K. Gonatas, W.F. Hickey and W.W. Schlaepfer (1988) Brain MR: pathologic correlation with gross and histopathology. 1. Lacunar infarcton and Virchow-Robirr spaces. A.J.N.R., 9: 621–628. Davis, W.L., D.D. Blatter, H.R. Hamsberger and D.L. Parker (1994) Intracranial MR arrgiography: comparison of single-volume three-dimensional time-of-flight and multiple overlapping thin slab acquisition techniques. A.J.R., 163: 915–920. Dozono, K., N. Ishii, Y. Nishihara and A. Hone (1991) An autopsy study of the incidence of Iacunes in relation to age, hypertension, and atherosclerosis. Stroke, 22: 993–996. Fisher, C.M. (1965) Lacunes: small, deep cerebral infarcts. Neurology, 15: 774–784. Flora, G.C., A.B. Baker, R.B. Loewenson and A.C. Klassen (1968) A comparative study of cerebral atherosclerosis in males and females. Circulation, 38: 859–869. Heiserman, J.E., B.P. Drayer, P.J. Keller and E.K. Fram (1992) hrtracranial vascular stenosis and occlusion: evaluation with three-dimensional time-of-flight MR arrgiography. Radiology, 185: 667-673. Jungreis, C.A., E. Kanal, W.L. Hirsch, A.J. Martinez and J. Moossy (1988) Normal perivascular spaces mimicking Iacunar infarction: MR imaging. Radiology, 169: 101– 104. Korogi, Y., M. Takahashi, N. Mabuchi, H. Miki, S. Fujiwara, Y. Horikawa, T. Nakagawa, T. O’Uchi, T. Watabe, H. Shiga and M. Furuse (1994) Intracranial aneurysms: diagnostic accuracy of three-dimensional, fourier transform, time-of-flight MR angiography. RadiolWY7193: 181–186. Laster, R.E., J.D. Acker, H.H. Halford III and T.C. Nauert (1993) Assessment of MR angiography versus arteriography for evaluation of cervical carotid bifurcation disease. A.J.N.R., 14: 681–688. Lewin, J.S. and G. Laub (1991) Intraerrmial MR angiography: a direct comparison of three time-of-flight techniques. A.J.N.R., 12: 1133– 1139. Marks, M.P., N.J. Pelc, M.R. Ross and D.R. Enzmarm (1992) Deternrina-
— Y. Kajiya et al./Journal of the Neurological Sciences 145 (1997) 195–203
tion of cerebral blood flow with a phase-contrastcine MR imaging technique: evaluation of normal subjects and patients with arteriovenous malformations. Radiology, 182: 467-476. Mattle, H., R.R. Edelmatt, K.U. Wentz, M.A. Reis, D.J. Atkinson and T. Ellert (1991) Middle cerebmt artery: determination of flow velocities with MR angiography. Radiology, 181: 527–530. Portman, M.A., T.G. Cooper and E.J. Potchen (1992) Physiologic alterations in cranial blood flow demonstrated by magnetic resonance angiography. Invest. RadioI., 27: 240-244. Prorok, R.J. (1989) Signa Applications Guide. Volume 2. 2nd cd. GE Medical Systems. Milwaukee, pp. 12-15. Shaw, T.G., K.F. Mortel, J.S. Meyer, R.L. Rogers, J. Hardenberg and
203
M.M. Cutaia (1984) Cerebral blood flow changes in benign aging and cerebrovascular disease. Neurology, 34: 855–862. Vnens, E.M., V. Kraaier, M. Musbach, G.H. Wieneke and A.C. van Huffelen (1989) Transcranial pulsed doppler measurements of blood velocity in the middle cerebral artery: reference values at rest and during hyperventilation in healthy volunteers in relation to age and sex. Ultrasound Med. BioL, 15: 1–8. Wehdi, F.W. (1990) Time-of-flight effects in MR imaging of flow. Magn. Reson. Med., 14: 187-193. Yamada, I., S. Suzuki and Y. Matsushita (1995) Moyamoya disease: comparison of assessment with MR angiography and MR imaging versus conventional angiography. Radiology, 196: 21 1–218.
NEUROLOGICAL SCIENCES Journal of tfre Neurological Sciences 145 (1997) 195-203
Age-relatedchanges in cerebral MR angiography Yoriko Kajiya *, YoshikiKajiya,MasayukiNakajo Deparrrrrentof Radiology, Faculryof Medicine, Kagoshima University, 8-35-I, Sakuragaoka,Kagoshima 890, Japan Received 3 January 1996; revised 21 May 1996; accepted 23 May 1996
Abstract Two hundred forty four neurologically normal subjects (118 males and 126 females, aged 7 to 82 yrs.) were evaluated to assess the correlation of the degree of visualization of cerebral arteries on magnetic resonance angiography (MRA) with aging, sex, literality or lacunar infarctions on magnetic resonance imaging (MM). A single volume three-dimensional transformation time-of-flight pulse sequence with slab thickness of”115 mm was used to obtain axial, saggital and coronal projection images. A total of 6 arteries including bilateral anterior cerebral arteries (ACAS), middlecerebralarteries(MCAS)andposteriorcerebralarteries(PCAS)foreverysubjectwere
eachgradedinto8 scoresby thedegreeof visualization of thearteries.Resultsshowthatan age-relateddeclineof visualization of the cerebralarteriesin ACA(r = –0.603, p < 0.001), MCA(r = – 0.452,p < 0.001) andPCA(r= – 0.537,p < 0.001). The arterieswere bettervisualizedin femalesthan malesduringthe fifth decade(p < 0.01). No substantial relationship was observed between the visualization and sex in the other decades, literality or lacunar infarctions. Thus, the cerebral MRA showed an age-related decline of visualization of the arteries and a sex-difference of the visualization in the fifth decade, probably due to the decline of the blood flow velocity with aging and a difference in it between females and males in the fifth decade. Published by Elsevier Science B.V. Keywords; Magnetic resonance angiography (MRA); Cerebral artery; Age; Sex; Lacunar infarction
1. Introduction
Many studieshave documentedthe clinicalefficacyof a three-dimensionaltime-of-flightMRA in the evaluationof intracranialvascular pathology(Korogi et al., 1994; Yamada et al., 1995;Heisennanet al., 1992).To our knowledge, however, there have been no reports evaluating age-relatedchanges in cerebral MRA in many neurologically normal subjects.In this article, we explore age-related changes in cerebral MRA. We also study about the effects of sex, literality and lacunar infarctionson the visualizationof the cerebralarterieson MRA.
2. Materials and methods 2.1. Patients
Two hundred forty four patients(126 females and 118 males)aged 7 to 82 yearswere includedin this study,who
‘ Correspondingauthor. Tel.: +81 (992) 75-5417. Fax: +81 (992) 75-5420. O022-510X/97/$17.00 Published by Elsevier Science B.V. PII S0022-5 10X(96)O0 199-2
received MR examinationsbetween December 1990 and June 1991. Two-thirdsof them underwentMR examinations because of headache,dizziness,paresthesiaor other symptomes,but who showedno specificneurologicaldisorders and one third of them underwent for a thorough cerebralcheck up. They had no tumor,hemorrhage,infarction except lacunar infarction and no vascular pathology such as intracranialaneurysms. 2.2. Methods of MRA and MRI The MR examinationswere performedon a 1.5-Timaging system(Signa; GE Medical System,Milwaukee).The single-volumethree dimensional Fourier transformation (3DlW) time-of-flight(TOF) (Wehrli, 1990) MR angiographic images were obtainedwith a gradientecho pulse sequencewith ‘spoiling’of residualtransversemagnetization known as spoiledgradient-recalledacquisitionin the steady state (Prorok, 1989).The imagingparameterswere 30/1 1/1 (TR/TE/excitations) using a 20° flip angle.No presaturation pulse was used. Flow compensationwas performed in the read and phase encoding directions.A 256 X 192matrix was used, with a field of view of 22 cm and a partitionthicknessof 0.9 mm, resultingin a 0.9 X
196
Y.Kajiya et al. /Jourrral of the Neurological Sciences 145 (1997) 195–203
1.1x 0.9 mm voxel. One hundred twenty eight partitions with a section thicknessof 115 mm to cover almost the whole brain were acquired in an imagingtime of 12 min 18 s. The 128 axial source images obtained in this way were postprocessedby means of a maximum-intensity pixel projectionalgorithmto produce multiple projection rotated about the section axis. We used 3 images; axial, sagittaland coronalprojectionimages. T1- and T2-weightedspin-echoimages were obtained to get axial and sagittalMR images.Repetitiontime, echo time and excitation were 400 msec, 11 msec and two (400/1 1/2) for T1-weightedimagesand were 2200/90/1 for T2-weightedimages.An imagematrix was 256X 192.
10.
U* 5 A : ACA Y= -O. 11X+13(.=2U) (7=-0 .s03. P
1
2
345a
789
131
I
H
G
!5-
not {8) !61 us
041 U*
I I 1
678
aO
[58
8: MCA V= –0.LWOX+13(.=244)
E
(r=–0.452,
3
01
1
2
p< O.@ll)
3456789
2: A B
?
A : ACA C : PCA Y= -0. ffi4x+ II (n=244) (r= -0.537, P< O.001)
-1....u
3/--’).
$“--’& ...
f’-&
o!
4
5 [h ,.,,
34
56
7a9
range
Fig. 2. Age-related changes in visualization of the cerebral arteries on ACAS (A), MCAS (B) and PCAS (C). Solid circles with bars show mean scores with SD.S. The number in parenthesis below the bar is number of patients. X: Actual age, Y: Score of visualization of arteries. Y declined with age rapidly in the ACAS and gradually in MCAS and PCAS.
--)2 [~ ~ , :,
2
Age
B : MCA
!,-:
1
67
the the the the the
{4 ,., ,!, :
8
c:PCA
Fig. 1. The methods of estimation of the degree of cerebral arterial visualization. Scoring of the arterial visualization was made for the ACA on a 3D MRA sagittal projection image as illustrated with the scheme (A) artd for the MCA and pCA on a 3D MRA axial projection image as illustrated with the schemes (B) and (C). The solid and broken lines represent the linear and beaded arteries, respectively. See text about the details of scoring criteria.
2.3. Scoring of the degree of visualization of the cerebral arteries on MRA
Gradingof the cerebral arteriesis shown schematically in Fig. 1. Each vessel has a uniquescoring systemwith a maximumscore of 8 per vessel. The ACA was estimated on a sagittalprojectionimage. A coronalprojectionimage
Y. Kajiya et aL/Journal of the Neurological Sciences 145 (1997) 195–203
was used to decidewhetherthe arterieslocatedin the right or left hemisphere.The ACA was evaluatedat 8 points as illustrated in Fig. 1A. The point A was set where the internal cerebral artery went upward after the carotid siphon. The upper and lower points of the pericallosal artery in front of the lamina terminals were set as C and
197
pointcorrespondingto the genu of corpuscallosum was set as E. The D point was set on the positionwhere the horizontaldistancebetweenC and D was equal to that between A and B. The point where the extension of BC intersectedthe pericallosalartery was set as G. The points F, H and I were set to make the horizontaldistancesequal
B. The
(A)
Fig. 3. Age-related changes in cerebral MRA of neurologically normal patients a: sagittal pro~ction images and b: axial projection images. All six arteries are excellently demonstrated and the mean scores of bilateral ACAS, MCAS and PCAS are 14, 16 and 14 in a 7-year-old girl (A). The cerebral arteries are moderately visualized and the scores of bilateral ACAS, MCAS and PCAS are 10, 11 and 10 in a 43-year-old woman (B). All the arteries, especially ACAS, w PoorlY visualized and tie scores of bilateral ACAS, MCAS and pCAS are o, 1’ and 6 in a 81-year-old woman (C). These images demonstrate that visualization of the cerebral arteries declines with aging.
198
Y.Kajiya et al. /JournaI of the Neurological Sciences 145 (1997) 195–203
betweenEF, FG, GH and HI. The score was made as 1, 2, ..., 7 when the tip of the visualizedvessellay betweenBC, CD, ..., HI, respectively. The score was 8 when the pericallosal artery was visualized clearly beyond the I point. The MCA was estimatedon the axial projectionimage. We divided the MCA into four segments, the sphenoid (Ml) segment, the insular (M2) segment, the opercukir
(M3) segmentand the terminal(M4) segment.Each segment was rather long and had a differencein appearance; the smoothlyand clearlyvisualizedvesselsand beadedand faintly visualized vessels. Therefore in scoring of the MCA, we used two parameters including the terminal point and clearness of the visualized vessel. When the terminalbranch lay on Ml segment and was beaded and faint, the scorewas 1. If it was smoothand clear, the score
(B)
Fig. 3 (continued).
Y. Kajiya et al. /Journal of the Neurological Sciences 145 (1997) 195-203
was 2. The scores were 3 and 4 on M2 segment,5 and 6 on M3 segment,and 7 and 8 on M4 segmentin the same way. The PCA was also estimated on the axial projection image and divided into 8 scores by the place and the scoringsystemwas the same used in the MCA. When the terminal branch was found before the posterior temporal
199
artery branchsoff, the degree of visualizationwas graded as 1 or 2, and graded as 3 or 4 before the parietooccipital artery and the calcarinearterybranchoff, graded 5 or 6 on these arteriesand 7 or 8 on their branches. The ACA, MCA and PCA were separately scored in both hemispheresand the scores were summed. A total score of the abovesix arterieswas set as [T]. Theoretically
(c)
Fig. 3 (continued).
Y. Kajiya et al. /Journal of the Neurological Sciences 145 (1997) 195-203
fifth decade (Dozono et al., 1991). So we estimated the lacunarinfarctionsin the patientsof sixthto ninethdecades (n= 164).Lacunarinfarctionshave been definedto be old small deep infarctions,0.5 to 15 mm in diameter(Fisher, 1965).However, dilated perivascularspace cannot be excluded. Therefore we estimated the lesions larger than 5 mm in order to excludethe perivascularspace, which was usuallyunderthan 5 mm in diameter(Jungreiset al., 1988; Braffmartet al., 1988).The high intensitylesions on T2 weighted images were scored as O–2 based on their size (O= none, 1 = 5-10 mm in diameter, 2 = 11-15 mm in diameter)and the gradeswere summed.Degree of lacunar infarctions were evaluated in the following three parts; basal ganglia and thalamus, cerebral white matter, and pens and cerebellum.The sum was made for each part in each patient. The maximumgrades of Iacurtarinfarction were 7, 12 and 2 in basal ganglia and thalamus,cerebral white matter, and pens and cerebellum,respectively.The number of the patients without lacunar infarctions was 121, 109 and 152,respectively.
(2)
0
(3)
6
(;)
(4)
1
1I
(7)&9 (4)
(13) (3)
IJ t I
(9)
@l)
or)
=o~
(2)
=
I
2
4
3
5
6
7
8
9
+ Age
range
Fig. 4. Sex differences in visualization of the cerebral arteries in neuroIogically normal males (0) and females (0). [T] is tie total score of rhe six arteries. Circles with bars show the mean ratios with S.D.S. In the fifth decade, the arteries were visualized significantly better in females than in males.
[T] can be scored from 6 to 48. However,it was actually scored from 6 to 43. Two radiologists(Y.K. and Y.K.) evaluated independentlyMRA and MR images of each patient, using the above mentionedscoring systems.The difference of scores was always one score. When the assessmentdiffered, a consensus interpretationwas obtained.
2.5. Stratijcation of age The ages of patientswere divided into 9 ranges based on the decade: 1 (O–9-year-old;n = 3), 2 (10–19-year-old; n = 10), 3 (2f)-29-year-o]d; n = 8), 4 (3t)-39-year-old; n = 16), 5 (4.@49_yem.o]d;n = 43), 6 (50–59-year-old; n = 56), 7 (6&69-yew-old; n = 73), 8 (70–79-year-old; n = 30) and 9 (80–89-year-old;n = 5).
2.4. Grading of lacunar infarctions Lacunarinfarctionswere foundin the patientsabovethe age of 40 years, however were not commonlyseen in the
PCA
MCA
ACA
6 T
5
I I
4
3
2
1
c
6767
6767
6767 Age
range
Fig. 5. Literality of visualization of the cerebral arteries between the right and left hemispheres. ●, Right hemisphere and O, left hemisphere of the male (n= 27 in the sixth decadeand 35 in the seventh decade), A; the right hemisphere and A; the left hemisphere of the female (n = 29 in the sixth decade and 38 in the seventh decade). Circles or triangles with bars show the mean ratios with S.D.S. There were no statistically significant differences in the visualization between the right and left hemispheres.
201
Y. Kajiya et aL/Journal of the Neurological Sciences 145 (1997) 195–203
2.6. Analysis of data Visualizationof cerebral arteries on MRA was compared to the age. First, the linear correlationcoefficient was obtainedbetweenthe actual age (X) and the sum (Y) of the scoresof rightand left arteriesof the ACA, MCA or PCA and its significancewas calculated. Secondly, the mean score+ S.D. of the sumof right and left arterieswas calculated in each age range. Maximum and minimum scoresof Y were 16 and 2. A sex differencewas estimatedwith the mean score+ S.D. of [T] using unpaired t-test in each age range. The right-leftdifferencein the ACA, MCA or PCA was estimatedwith the mean score+ S.D. of each artery(maximum and minimumscores were 8 and 1) using paired t test. Multiple regression analysis with visualizationof the cerebralarterieson MRA as a dependentvariablewas used to investigatethe dependencyon age and lacunar infarctions (independentvariables). The statisticalsignificantlevel was assumedto be less than 0.05. 3. Results The relation of visualizationof the cerebral arteries to age is shown in Fig. 2. A statisticallysignificantnegative linear correlationexisted between the actual age (X) and visualizationof the cerebral peripheral arteries (Y) (r= –0.603, p <0.001 for age vs the ACAS, r = –0.452, p <0.001 for age vs the MCAS, r = –0.537, p <0.001 for age vs the PCAS).The age-relateddeclinewas rapid in the ACAS and gradual in the MCAS and PCAS. Three instancesof the first, fifth and ninth decadesare shownin Fig. 3. Projectedthree-dimensionalTOF MR images of a 7-year-oldgirl demonstratedexcellentlyboth central and peripheralarteries. However, the peripheralarteries were moderately detected in a 43-year-old woman and they were poorlydetectedin a 8l-year-old woman. The relation of visualizationof the cerebral arteries to sex is shownin Fig. 4. For the total population,there was no statisticallysignificantdifferencebetweenmales([T]= 24.4 + 6.3, n = 118) and females ([T]= 24.9+ 6.3 n = 126). However, in the fifth decade, the visualizationwas higher (p < 0.01) in females ([T]= 28.3 ~ 4.8, n = 22) than in males([T]= 24.3+ 4.6, n = 21). No statisticallysignificantright-left differencesin the ACAS,MCASand PCASwas found in any decadeswhich had a sufficientnumberof patientsfor statisticalanalysis. Fig. 5 shows the results of literality in the sixth and seventhdecades. A multipleregressionanalysisof the data in the group of 50 or more years old patientsis shown in Table 1. [T] was set as dependentand age and lacunarinfarctionswere set as independent.Therewas a statisticalsignificantcorre-
ctable 1 Multiregression analysis with [T] as dependent using 4 factors of age (Xl), lacunar infarctions in basat ganglia and thafamus (X2), cerebral white matter (X3), and pens and cerebellum (X4) as independent Factors
xl x2 x3 x4
[T] Simple regression p-Value Coefficient -0.323
-0.220 -0.080
–0.047
p p p p
<0.001 <0.01 >0.05 >0.05
Multiple regression p-Value Coefficient –0.285 –0.148 0.024
0.019
p <0.001 p >0.05 p >0.05 p >0.05
lationbetweenage and [T]( p < 0.001).It revealedthatthe increaseof lacunar infarctionsof basal gangliaand thalamus was concomitantof age-dependentdeclineof [T], and there was no statisticallysignificantrelationbetweenlacunar infarctionsand visualizationof the cerebralarteries. 4. Discussion Various approachesto MRA have recently been proposed and have already been shown to be useful in the clinicalevaluationof cerebralvasculardiseases(Korogiet al., 1994; Yamada et al., 1995; Heisermanet al., 1992). However, most of them are investigationsabout given diseases.Very few studieswere reportedaboutthe cerebral arteries themselves using MRA (Portman et al., 1992; Mattleet al., 1991).Marks et al. describedan age-related declinein mean velocityof the internalcarotid artery and basilar artery (Marks et al., 1992).An age-relateddecline in mean velocityof the internalcarotid artery and basilar artery is so far available(Marks et al., 1992).However, there have been no reports dealing with the relation between aging and visualizationof the cerebral peripheral arterieson MRA in many neurologicallynormal subjects. The presentstudyrevealedthat the degreeof visualization of the cerebralperipheralarteriescorrelatednegatively with age. The declinewas rapid in the ACASand gradual in the MCAS and PCAS. The flow-relatedenhancement obtainedwith the singlevolume3DlWTOF is veryhigh at the point that a vessel enters the imaging volume. The inflowingspinsundergosaturationas the bloodtransverses the region of excitation,resulting in progressivelydiminishing contrast (Lewin and Laub, 1991). The ACA runs relativelyparallel to the long axis of body and intersects the region of excitation.Therefore, the slow flow in the ACA tendsto decreasethe signalof inflowingbloodmore obviouslythan those in the MCA and PCA. Studies using transcranial pulsed Doppler technique showed that the measured blood velocity of the middle cerebralartery decreasedsignificantlywith increasingage (Vriens et al., 1989; Ackerstaff et al., 1990). With this single 3DIT volume method, the slow flow cannot span the entire volume before becoming sufficientlysaturated,
202
Y.Kajiya et aL/Journal of the Neurological Sciences 145 (1997) 195-203
and results in signalloss (Lewin and Laub, 1991).Standing on the above,the age-relateddecreasein visualization of the cerebral arteries seems to be caused by age-related decreaseof the flow velocity.The singlevolumetechnique is less sensitiveto visualizea slowflow than the sequential 2-D slicemethodand multiplethin volumemethod(Lewin and Laub, 1991).Usually42 to 72 mm thicknessof slab is used (Yamadaet al., 1995;Heisermanet al., 1992;Davis et al., 1994)to avoid signal loss caused by saturationof the flow in the slab. Howeverin this studywe used a very thick single volume of 115 mm to cover almost all the intracranialvessels,which decreasedsensitivityto vessels which containedslowlymovingblood,resultingin making the more prominentdifferencein visualizationof the cerebral arteries between the patients. Using the usual slab thickness,we would not have such differenceobservedin this study. Recentlywe can use the multipleoverlappingthin slab acquisitiontechnique(MOTSA)which has decreasedsensitivity to the effects of flow saturation and allows improvedvisualizationof the intracranialvascularabnormalities (Davis et al., 1994).Using this technique,we will get better images even in elder people and will not have so distinctdifferencecompareto youngerpeople. Arteriosclerosismay be another factor that influences visualizationof the cerebral arterieson MRA. The severe stenosiswith atherosclerosiswas reportedto cause a segmental decreasein visualizationof the carotid arterieson MRA (Laster et al., 1993). Atherosclerosismay also decrease the flow velocity. We examinedthe relation betweenvisualizationof the cerebral arteries on MRA and lacunar infarctionswhich have art associationwith atherosclerosis(Dozono et al., 1991; Braffman et al., 1988). There was no statistically significantrelation between visualizationof the arteries and lacunar infarctions.It has been shown that MRI receives clear effects from flow in large vessels, but it is difficultto measuretissueperfusionby MR studies(Axe], 1984). It may be the reason why there was no relation betweenvisualizationof the cerebralarterieson MRA and the number of Iacunar infarctions which are ischemic change of the very small arteries. Even without atherosclerosis,vesselsbecomemoretortuouswith aging(Adams and Victor, 1993).It may also causethe TOF effect down. Femalesfrom the third to fifth decadesshowedgreater scoresof the visualizationthan males,and the fifth decade femaleshad statisticallysignificantbettervisualizationthan males.The sex-relateddifferencesin humancerebralblood flow and blood flow velocity have been described by several investigatorsusing Xenon inhalationtechniqueor transcranial Doppler examination (Vriens et al., 1989; Ackerstaffet al., 1990;Shawet al., 1984).They foundthat premenopausalfemales had significantlyhigher cerebral blood flow valuesthan their male counterparts.It has been shown that sex differencein frequencyof cerebralatherosclerosis appears from the fourth decade and the largest
sex differenceis found in the fifth decade, indicatingthat the males have considerablyhigher degrees of cerebral atherosclerosisthan the femalesin these age groups(Flora et al., 1968).These reportswill supportour result. There were also no statisticallysignificantright–left differencesin visualizationof the cerebral arteries. In the study using transcranialDoppler examination,it has been noted that the mean blood velocityof the MCA shows no right–leftdifferences(Ackerstaffet al., 1990). We use MRA in order to get informationabout the morphologyof the vessels. Conventionalarteriographyis superiorto MRA in visualizingthe vessels.However, the techniqueof MRA can be used in additionto routineMRI and is noninvasive.In addition, in this study MRA was considered to reflect the flow velocity like transcranial Doppler and Xenon inhalationmethods.MRA may show the natural state of vesselsbecause it is not necessaryfor MRA to inject a contrastmediumby force.
References Ackerstaff, R.G.A., R.W.M. Keunen, W. van Pelt, A.D.M. van Swijndregt and T. Stijnen (1990) Influence of biological factors on changes in mean cerebral blood flow velocity in normal ageing: a trarrscranial Doppler study. Neurol. Res., 12: 187-191. Axel, L. (1984) Blood flow effects in magnetic resonance imaging. A.J.R., 143: 1157-1166. Braffman, B.H., R.A. Zimmerman, J.Q. Trojarrowski; N.K. Gonatas, W.F. Hickey and W.W. Schlaepfer (1988) Brain MR: pathologic correlation with gross and histopathology. 1. Lacunar infarcton and Virchow-Robirr spaces. A.J.N.R., 9: 621–628. Davis, W.L., D.D. Blatter, H.R. Hamsberger and D.L. Parker (1994) Intracranial MR arrgiography: comparison of single-volume three-dimensional time-of-flight and multiple overlapping thin slab acquisition techniques. A.J.R., 163: 915–920. Dozono, K., N. Ishii, Y. Nishihara and A. Hone (1991) An autopsy study of the incidence of Iacunes in relation to age, hypertension, and atherosclerosis. Stroke, 22: 993–996. Fisher, C.M. (1965) Lacunes: small, deep cerebral infarcts. Neurology, 15: 774–784. Flora, G.C., A.B. Baker, R.B. Loewenson and A.C. Klassen (1968) A comparative study of cerebral atherosclerosis in males and females. Circulation, 38: 859–869. Heiserman, J.E., B.P. Drayer, P.J. Keller and E.K. Fram (1992) hrtracranial vascular stenosis and occlusion: evaluation with three-dimensional time-of-flight MR arrgiography. Radiology, 185: 667-673. Jungreis, C.A., E. Kanal, W.L. Hirsch, A.J. Martinez and J. Moossy (1988) Normal perivascular spaces mimicking Iacunar infarction: MR imaging. Radiology, 169: 101– 104. Korogi, Y., M. Takahashi, N. Mabuchi, H. Miki, S. Fujiwara, Y. Horikawa, T. Nakagawa, T. O’Uchi, T. Watabe, H. Shiga and M. Furuse (1994) Intracranial aneurysms: diagnostic accuracy of three-dimensional, fourier transform, time-of-flight MR angiography. RadiolWY7193: 181–186. Laster, R.E., J.D. Acker, H.H. Halford III and T.C. Nauert (1993) Assessment of MR angiography versus arteriography for evaluation of cervical carotid bifurcation disease. A.J.N.R., 14: 681–688. Lewin, J.S. and G. Laub (1991) Intraerrmial MR angiography: a direct comparison of three time-of-flight techniques. A.J.N.R., 12: 1133– 1139. Marks, M.P., N.J. Pelc, M.R. Ross and D.R. Enzmarm (1992) Deternrina-
— Y. Kajiya et al./Journal of the Neurological Sciences 145 (1997) 195–203
tion of cerebral blood flow with a phase-contrastcine MR imaging technique: evaluation of normal subjects and patients with arteriovenous malformations. Radiology, 182: 467-476. Mattle, H., R.R. Edelmatt, K.U. Wentz, M.A. Reis, D.J. Atkinson and T. Ellert (1991) Middle cerebmt artery: determination of flow velocities with MR angiography. Radiology, 181: 527–530. Portman, M.A., T.G. Cooper and E.J. Potchen (1992) Physiologic alterations in cranial blood flow demonstrated by magnetic resonance angiography. Invest. RadioI., 27: 240-244. Prorok, R.J. (1989) Signa Applications Guide. Volume 2. 2nd cd. GE Medical Systems. Milwaukee, pp. 12-15. Shaw, T.G., K.F. Mortel, J.S. Meyer, R.L. Rogers, J. Hardenberg and
203
M.M. Cutaia (1984) Cerebral blood flow changes in benign aging and cerebrovascular disease. Neurology, 34: 855–862. Vnens, E.M., V. Kraaier, M. Musbach, G.H. Wieneke and A.C. van Huffelen (1989) Transcranial pulsed doppler measurements of blood velocity in the middle cerebral artery: reference values at rest and during hyperventilation in healthy volunteers in relation to age and sex. Ultrasound Med. BioL, 15: 1–8. Wehdi, F.W. (1990) Time-of-flight effects in MR imaging of flow. Magn. Reson. Med., 14: 187-193. Yamada, I., S. Suzuki and Y. Matsushita (1995) Moyamoya disease: comparison of assessment with MR angiography and MR imaging versus conventional angiography. Radiology, 196: 21 1–218.