817
MR Angiography in Children with Cerebral Neurovascular Diseases: Findings
in 31 Cases
H
.
American Journal of Roentgenology 1992.159:817-823.
Thomas J. VogI1 JOrn 0. Balzer1 Joachim Stemmler1 Clifford Bergma&
E. Egger Josef
Lissner1
OBJECTIVE. We evaluated the suitability of MR angiography for routine use in children with suspected intracranial vascular disease. SUBJECTS AND METHODS. Thirty-one children, 6 months to 14 years old, with intracranial lesions or clinically suspected vascular malformations were studied prospectively with conventional MR imaging and time-of-flight MR angiography. In nine cases, MR angiographic findings were verified with digital subtraction angiography or conventional angiography All MR studies were performed on a 1.5-T MR system using a circularly polarized head coil. RESULTS. Arterial MR angiography, performed in 24 cases, revealed congenital abnormalities of the arterial vessels in 20 cases. Vessel stenosis was observed in nine patients, and displacement of intracranial arteries due to tumors could be seen in 10 patients. Seven children had no abnormal findings. Venous MR angiography was performed in seven children, with depiction of sinus thrombosis in six cases. The comparative analysis of MR angiography and digital subtraction angiography showed equivalent results in nine patients; in one patient the degree of stenosis was overestimated with MR angiography. CONCLUSION. MR angiography, when combined with MR imaging, reveals information about soft-tissue and vascular structures in a single setting. At this point, MR angiography can replace invasive conventional angiography or digital subtraction angiography only in selected cases because of software and hardware limitations. Arterial or venous MR angiography can be helpful as an additional scan in MR examinations of children with suspected cerebral neurovascular diseases, and its noninvasive nature makes it well suited for routine use in children. AJR
159:817-823,
Although
October
conventional
1992
X-ray
angiography
and
digital
subtraction
angiography
(DSA) are the gold standards for the assessment of intracranial vessels [1 -3], the invasiveness of these procedures makes them difficult to perform, particularly in children. Thus, they are not routinely available for use in children in many centers. Among the noninvasive methods of imaging blood flow, MR angiography is an attractive alternative. It can be used to evaluate vessel patency, flow magnitude,
Received March 12, 1992; accepted s/on April 30, 1992.
after revi-
1
Departmentof
Radiology,
University
1 , 8000 Munchen
of Munich,
2, Germany. Ad-
dress reprint requests to T. J. yogi. 2
Pediatric
Munchen
Hospital,
0361 -803x/92/1 CAmerican
University
2, Germany.
594-0817
Roentgen
Ray Society
As an adjunct
raphy can provide tumors that would
information otherwise
zation.
This study was supported by a grant (no. 842501) of the Wilhelm Sander Foundation. ZiemssenstraBe
and flow direction.
of Munich,
8000
MR angiography
to conventional
MR imaging
studies,
about blood supply and vascular be available only through angiography
exploits
the intrinsic
properties
of flowing
MR angiog-
topography of via catheteri-
blood so that
catheterization and the need for contrast material are eliminated. Because of its safety, it is likely that MR angiography will play an increasingly significant role in the workup of children with suspected neurovascular disease. Nevertheless, MR angiography is still subject to a number of technical limitations, because of motioninduced artifacts and artifacts caused by alterations in flow behavior through
abnormal vessels. Furthermore, it is not yet possible to consistently visualize vessels smaller than 1 mm in diameter [4, 5]. We present the MR angiographic findings in 31 children, correlate the findings
81 8
VOGL
ET AL.
AJR:159, October 1992
with those of conventional angiography or digital subtraction angiography, when available, and briefly summarize the ex-
amination
strategies
and parameters
we found most useful. C U:
Subjects
and
Methods
6
Thirty-one children 6 months to 14 years old with suspected intracranial lesions or vascular malformations seen on conventional spin-echo MR images were studied prospectively with MR
.
angiography. MR imaging examinations were carried out with the patient under general anesthesia in 27 cases. All MR studies were performed on a 1 .5-T MR system (Magnetom SP 63, Siemens AG, Erlangen, Germany) with a circularly polarized head coil 30 cm in diameter. Conventional
MR
imaging
was
performed
according
to the
.
$
thickness
3-1 0 mm, 256 x 256 matrix,
images (600/i 5, i 3 slices, slice thickness
x 256 matrix,
acquisitions)
two
one acquisition)
c .
following
were obtained.
and then Ti
3
.
American Journal of Roentgenology 1992.159:817-823.
For all axial images,
a saturation
plane
below the imaging slices to eliminate motion artifacts.
c
coronal
. .
was placed
The total width
of the field of view (FOV) was 20 cm in all cases. In i 1 children with intracranial tumors, Ti-weighted images were obtained after IV bolus injection of gadopentetate dimeglumine (dosage: 0.1 mmol per kilogram body weight). MR angiography was performed by using three-dimensional (3D) Fourier transformed, rephased gradient-recalled echo sequences. Three-dimensional
fast
imaging
with
steady
ing was used to depict vessels with dimensional (2D) or 3D fast low-angle
precession
fast shot
(FISP)
> -E (‘
I-.
w .
imag-
blood flow, and (FLASH) imaging
‘-
-
>
we had previously found gave the best results in visualizing fast-
.
..
flowing blood with 3D FISP imaging. A TRITE of 36/iO was best for
,
E
placed arterial venous The tracing
adjacent or venous or arterial 3D or 2D algorithm
(Table
i
).
Presaturation
slices
intensity projection (MIP) images [6, 7]. Rotatable projections of the MR angiograms were usually calculated with a range increment of i 50 The computing time required for image reconstruction was 3-
i 2 mm, depending For each
on range increment
scale
of 0 to 3: grade
were not depicted
data
0, vessels
on MR images;
-
c#{176}
0
0
I
.
5
U: .
.
9)
‘
5
.-
C’)
‘
‘ (,
9
‘
.
c
c
L
*1)
.
and resolution.
set of MR angiographic
and corresponding
MIP
images, the quality and extent of vessel visualization within the slab were ranked by a concensus of three independent observers on an ordinal
o
o o I
were
to the imaging volume in order to suppress either blood flow as appropriate to produce more purely MR angiograms, respectively. data set was processed by using a computer rayon an integrated workstation to provide maximum
o
(#{176}
.
the dural sinus system
‘-
#{176}-
twowas
used for vessels with slow flow. For arterial MR angiography, we used a TE of7 msec, a TR of4O msec, and a flip angle of iS#{176}, which
visualizing
X
X
-
or sagittal images were acquired as needed to show any lesions seen on axial images.
3
X X
3-i 0 mm, 256
Ti-weighted
.
-
scheme: After a preliminary sagittal image was obtained, axial T2weighted spin-echo images (3000/22,60,120 [TRITE], 17 slices, slice weighted
,
‘
that lay within
grade i
,
vessels
!
‘
?
i
< 2.
the FOV but
j
that could be
detected only over a short distance; grade 2, vessels that could be visualized over a greater distance but had interruptions in flow signal; grade 3, vessels that were imaged over their entire length within the
C’)
C’)
u
FOV and without signal interruption. The following arteries were evaluated: internal carotid, anterior cerebral, middle cerebral, anterior communicating,
ophthalmic,
ebral, superior cerebellar,
posterior
communicating,
anterior inferior cerebellar,
posterior
posterior
cer-
inferior
cerebellar, basilar, and vertebral (Table 2). On venous MR angiography, depiction of the following segments of the sinus system was evaluated: superior sagittal sinus, inferior sagittal sinus, straight sinus, confluens of sinuses, transverse sinus,
sigmoid sinus, cavernous sinuses,
the sphenoparietal
vein (Table
3).
sinus and its draining sinus,
and
the superior
superior bulb
and inferior of the jugular
cs
.
,
c
.
U-
Z
MRA
1992
OF
AJR:159,
October
TABLE Imaging
2: Delineation of Intracranial Arteries Quality on Arterial MR Angiography
Could Not Be
Vessel8
CEREBRAL
NEUROVASCULAR
and Evaluation
of
TABLE 3: Delineation Evaluation of Imaging
Grade 1 Grade 2 Grade 3
(%)
(%)
(%)
0 0
i 5 20
25 35
60 45
0 5
10 20
45 5
45 70
70 45
20 25
5 25
0 5
90 95 I 00
0 5 0
10 0 0
0 0 0
0 0
20 0
25 25
0 70
65
25
5
0
50 40
20 iO
0 15
0 0
0
20
i0
40
Evaluated
DISEASES
of Intracranial Sinus System and Quality on Venous MR Angiography
V esse I
(%) Internal Anterior
carotid artery cerebral artery
Middle cerebral artery Posterior cerebral artery
ACA PCA Ophthalmic artery Thalamostriate artery Choroid artery
Vertebral artery Basilar artery Superior
cerebellar
artery
AICA PICA Abnormal
vessels
819
Could Not Be Evaluated
Grade
1 Grade
2 Grade
(%)
(%)
(%)
0 0 50 66 0 0 0 17 0 0 17 0 0
33 0 17 17 17 0 0 17 17 17 53 33 17
50 100 17 17 83 100 83 0 0 0 17 67 50
3
(%) Superior sagittal sinus Straight sinus Inferior sagittal sinus Vein of Galen Confluens of sinuses Transverse
Sigmoid Cavernous
sinus
sinus sinus
Superior petrosal sinus Inferior petrosal sinus Sphenoparietal sinus Superior bulb of jugular vein Abnormal vessels
17 0 17 0 0 0 17 67 83 83 13 0 0
Note-Seven patients were examined; 14% had no abnormalitses. Grade 1 poor visualization, grade 2 = good visualization with some flow-signal interruption, grade 3 = optimal visualization.
American Journal of Roentgenology 1992.159:817-823.
=
Note.-24 patients were examined; 30% had no abnormalities. Grade 1 = poor visualization, grade 2 = good visualization with some flow-signal interruption, grade 3 = optimal visualization. a ACA = anterior communicating artery. PCA = posterior communicating artery, AICA = anterior inferior cerebellar artery, PICA = posterior inferior cerebellar artery.
Results Arterial
MR Angiography
Arterial
MR angiography
was
the nine cases in which results
performed
in 24 children.
of angiography
In
via catheteri-
zation were available, depiction of the vascular structures on MR angiograms compared favorably with depiction on DSA
images, although vessels smaller than 1 mm in diameter could not be visualized consistently on MR angiograms (Table 2). The anterior cerebral artery could be seen as far as the high cortical segment in nine patients, the middle cerebral artery could be followed to the angular gyrus in eight patients.
Excellent for
the
imaging upper
quality
segments
could of the
be achieved internal
in 1 2 patients
carotid
artery,
in
1 4 patients for the entire posterior cerebral and basilar artery, and in one patient for the posterior communicating artery. Smaller arteries such as the anterior communicating
Fig.
1.-3-year-old
with
venous malformation encephalocele. A, Coronal
intracranial
arterio-
and a frontoethmolds
Ti-weighted
(600/15) MR Image of frontal lobe. FISP (40/7, 15#{176})
shows #{149}ncephalocele (arrows) B, Axial,
three-dimensional
arterial MR anglogram shows a network of abnormal vessels (arrowhads). Ophthalmic art#{149}ries (0) can be seen because they are dilated.
artery, the ophthalmic artery (Fig. 1), the superior cerebellar artery, the anterior inferior cerebellar artery (AICA), and the posterior inferior cerebellar artery (PICA) could be visualized in only a few cases. On arterial MR angiograms, vessel stenoses were detected in eight cases involving the middle cerebral artery (n = 4; Figs. 2-5), the anterior cerebral artery (n = 2; Figs. 2 and 3), the posterior cerebral artery (n = 1 ), and the internal carotid artery (n = 1 , Fig. 5). Correlation with DSA findings was available in four patients and confirmed the diagnosis based on MR angiographic findings in three cases (Figs. 2B and 3B). In one case the grade of stenosis was overestimated on the basis of MR angiographic findings (Figs. 4B and 4C). In a fifth patient with hemiparesis, stenosis of the middle cerebral artery, consistent with clinical findings, was seen on MR angiograms; in the remaining three patients, infarcts evident on the spin-echo images corresponded to the areas of flowsignal loss as seen on the MR angiograms. Displacements of the anterior cerebral artery by an astrocytoma in one case and of the middle cerebral artery by a large hydrocephalus in
820
VOGL
ET AL.
AJR:1 59, October
1992
Fig. 2.-2-year-old with an astrocytoma in frontal lobe and stenoses of both anterior cerebral arteries and right middle cerebral artery. A, Axial, three-dimensional FISP (40/7, 15#{176}) arterial MR anglogram, rotated to a more sagittal view, artery
reveals stenosis of right middle cerebral (arrowhead) and stenosis of anterior cerebral arteries (arrows). m = middle cerebral ar-
tery, I
=
internal carotid
artery.
B, Conventional angiogram (lateral view) of rightintemal carotid artery verifies the diagnosis of stenosis of right middle cerebral artery made by MR anglography (arrowhead). Comparison of
American Journal of Roentgenology 1992.159:817-823.
stenosis on MR angiography and conventional anglography shows that MR angiography overestimates grade of stenosis.
Fig. 3.-2-year-old with right-sided hemiparesis and local vascular lesion (infarct) in left internal capsule. A, Axial, three-dimensional FISP (40/7, 15#{176}) arterial MR angiogram. Axial > coronal -30#{176} rotated maximum intensity projection angiogram reveals stenosis of both anterior cerebral arteries (arrowheads) and of left middle cerebral ar tery (arrows). B, Digital subtraction angiogram of left Internal carotid artery verifies resufts of MR angiography in this patient stenosis of anterior and middle cerebral arteries (arrowheads). a = anterior cerebral artery, m = middle cerebral artery, i = internal carotid artery.
C FIg. 4.-6-year-old with idiopathic hemorrhage in region of right internal capsule. A, Coronal TI-weighted (600/15) MR Image shows hemorrhagic lesion in region of right internal capsule (arrowheads) B, Axial, three-dimensional ASP (40/7, 15#{176}) arterial MR anglogram. Coronal rotated maximum intensity projection angiogram of right middle cerebral artery (arrows). C, Digital subtraction angiogram of right Internal carotid artery confirms stenosis of right middle cerebral artery (arrowheads).
shows
suspected
stenosis
MRA
October 1992
AJR:159,
OF CEREBRAL
and middle
American Journal of Roentgenology 1992.159:817-823.
intensity
cerebral
arteries. Ax15#{176}) arterial 30#{176} rotated angiogram shows
projection
821
bacterial
hemisphere (arrowheads). B, Axial, three-dimensional
FISP (40/7, 15#{176}) arterial MR angiogram. Coronal > sagittal -30#{176} rotated maximum intensity projection anglogram shows displacement of right middle cerebral artery due to occlusive hydrocephalus (arrowheads). Additionally, a common anterior cerebral artery is detected (arrows). a = anterior cerebral artery, m = middle cerebral artery.
ial, three-dimensional FISP (40/7, MR angiogram. Coronal > sagittal maximum
DISEASES
Fig. 6.-12-year-old with enlargement of right ventricle due to occlusion of foramen of Monro after meningitis, resulting in displacement of vessels. A, Coronal Ti-weighted (600/15) enhanced MR image shows an occlusive hydrocephalus In right
Fig. 5.-i4-year-old with giant-cell arteritis (Takayasu’s syndrome) and stenoses of upper segments of left internal carotid artery and origin of left anterior
NEUROVASCULAR
stenoses
of upper segments of left internal carotid artery (arrows) and at origin of left anterior and middle cerebral artery (arrowhead). a = anterior cerebral artery, m = middle cerebral artery, i = internal carotid artery, b = basilar artery.
another case were clearly seen on the MR angiograms. In one case, an occlusion of both middle cerebral arteries was seen. MR angiograms showed anatomic variations in five patients.
These
included
primitive
trigeminal
arteries
(three),
a
common anterior cerebral the internal carotid artery
artery (one, Fig. 6) and kinking of (one, Fig. 7). In the patient with the
occlusive
and a common
artery,
hydrocephalus a displacement
was suggested, additional
with
but this was excluded
projections
anterior
superimposition
and rotated
cerebral
of both
arteries
after we examined
the angiograms
(Fig. 6). In
one patient with a frontoethmoidal encephalocele, MR angiograms showed an arterial network in the area of the anterior cerebral artery. Dilatation of both ophthalmic arteries resulted in particularly erwise findings
good
visualization
of these
vessels,
which
are not usually seen on MR angiograms (Fig. confirmed the presence of a primitive trigeminal
oth-
1 ). DSA artery
in one case and kinking of the internal carotid artery in another case (Fig. 7). In one case, the mass effect of an astrocytoma resulted in displacement of the anterior cerebral artery and occlusion
superior
of
and confirmed
Venous
the
high
curvature;
cortical
this was
segments
observed
of
with
the
ascending
MR angiography
with DSA (Fig. 2).
superior sagittal, straight, transverse, and sigmoid and the bulb of the internal jugular vein were visualin all patients (Table 3). The cavernous sinus and its
ized draining superior and inferior petrosal sinus were in most cases ranked with grade 0 or grade 1 , largely because these sinuses lay at the edge of the imaging slices. Detection of the sagittal
sinus
abnormal
with kinking
of both internal
carotid
arteries
and no
findings.
Axial, three-dimensional FISP (40/7, 15#{176}) arterial MR angiogram. Sagittal > coronal 30#{176} rotated maxium intensity projection angiogram reveals kinking of left internal carotid artery (arrows).
Venous MR angiography was performed in seven patients with suspected sinus thrombosis and showed sinus thrombosis in six cases; collateral drainage could be seen in one. In one patient, venous MR angiography revealed no pathologic findings. Sinus thrombosis involved the transverse sinus in four of the seven cases. One patient with extensive throm-
bosis of the transverse and sigmoid sinuses was examined before and after treatment with heparin; venous flow improved
MR Angiography
The sinuses
inferior
Fig. 7.-2-year-old other
and the vein of Galen
was inconsistent.
markedly after administration of heparin (Fig. 8). Complete and partial sinus thrombosis could be differentiated by including the original MR angiographic data set in the review.
Use of gadopentetate dimeglumine in 1 1 children did not improve the overall imaging quality of angiography, although the relationship depicted in five
of tumor to adjoining cases. Large tumors
vessels was better or tumors with high
822
VOGL
ET AL.
AJR:159, October 1992
Fig. 8.-14-year-old girl under chemotherapy for acute myelogenous leukemia with sinus thrombosis before (A and B) and 1 month after (C and D) application of hepann. A, Coronal, two-dimensional (2D) FLASH (36/ 10, 60#{176}) venous MR angiogram. Coronal view shows occlusion of right transverse sinus (t), right sigmoid sinus (5), right jugular bulb (jb),
and jugular vein B, Coronal,
(fl.
2D FLASH
(36/10,
60#{176})venous MR
anglogram. Sagittal view shows thrombosis of superior sagittal sinus (arrowheads) as well. st = straight sinus, g = vein of Galen, t = transverse sinus, s = sigmoid sinus, j = jugular vein.
American Journal of Roentgenology 1992.159:817-823.
C, Coronal, 2D FLASH (36/10, 60#{176}) venous MR angiogram after heparin application shows reperfusion of right sigmoid sinus (s) and jugular bulb (jb). D, Coronal, 2D FLASH (36/10, 60#{176}) venous MR angiogram. Sagittal view after heparin applicahon shows imperfect reperfusion of superior sagittal sinus (arrowheads). is = inferior sagittal sinus, St = straight sinus, g = vein of Galen.
enhancement
reduced
MIPs in six cases, neighboring
proved giography,
angiographic
lesions
Use
of contrast
vessels.
the visualization especially
the
as these
imaging material
of small vessels in three
infants
quality
were superimposed slightly
in venous
where
3D
in
on im-
MR an-
FLASH
im-
aging was used. We achieved the best results by using the sequence parameters for 3D FISP and 2D FLASH imaging listed in Table 1. Visualization of smaller intracranial veins, such as the great vein of Galen and the inferior sagittal sinus, was better on 3D
FLASH than on 2D FLASH images. Discussion
Time-of-flight MR angiography is a useful technique for imaging intracranial vessels, although it does have several limitations. Because it depends on the refocusing of inflowing unsaturated spins, it may suffer from increasing saturation within a large imaging volume and incomplete refocusing [2, 5, 8]. Alterations in flow dynamics can cause partial signal
loss, because from from
these spins are dephased.
Dephasing
the random, tumbling motion in areas of turbulence the increased duration of stay in the measuring
Because of signal loss induced by poststenotic vessel stenosis cannot be graded appropriately
results and slice.
turbulence, on the basis
of MR angiographic findings [5, 9]. Nevertheless, the grade of stenosis or occlusion as seen on MR angiograms could
more accurately be assessed by including conventional MR images and the original 3D FISP data set in the review. Another source of error stems from the lack of signal from partially thrombosed aneurysms, which may lead to underestimation of aneurysm size [1 0, 1 1 ]. However, spin-echo images obtained during the same examination usually provide more information on true size and composition of the aneurysm; indeed, MR angiography together with conventional MR imaging may yield more information than DSA alone. Progressive saturation within an imaging volume occurs sooner with slow-flowing blood (e.g., venous blood) but also depends on the vessel’s course (perpendicular, oblique, or parallel to the imaging volume). Slice thickness and the RF deposition, which are operator-controlled parameters, also
cause saturation effects [8, 12]. As scanning parameters depend on the Ti relaxation of blood, a variation of these parameters
is possible
within
a narrow
Before this study, we determined
range
optimal
only [6, 8].
scanning
param-
eters in healthy children. We found that a TR of 40 msec and a flip angle of 1 5-20#{176}produced the maximum contrast-tonoise and signal-to-noise ratios. Slice thickness must be held to a minimum to decrease saturation effects; however, this also reduces the length of the vessel segment that can be imaged during a given examination [5, 8, 1 2-1 4]. A difficulty arises in positioning small axial planes in order to achieve maximum
be kept
information
as small
about
as possible
vessels,
as slice
in order
thickness
to increase
must
spatial
MRA
October1992
AJR:i59,
resolution. To increase slice thickness ration effects, two 3D volumes can
partitions
and 25% overlapping
A more
severe
problem
vessels
in the process
Imaging
of small
NEUROVASCULAR
while minimizing satube used with half the
[101.
is the loss of visualization
of acquisition
vessels
OF CEREBRAL
or reduced
of small
and MIP reconstruction. venous
flow
in partially
thrombosed sinuses requires an increase of in-plane spatial resolution, which can be achieved if the FOV is kept as small as possible. The best achievable in-plane resolution in our study was 1 mm, with a loss of signal from smaller vessels. A further problem arises in image reconstruction. When a vessel’s size is small compared with pixel size, small vessels may be lost completely on MIP reconstruction because of partial volume averaging and statistical fluctuations in background noise [1 0, 15]. This problem can be overcome by including the original MR angiographic data set in the review, as small vessels or restricted venous flow in partial sinus
American Journal of Roentgenology 1992.159:817-823.
thrombosis
is documented
on these
images
be
effect (e.g., TE = 7 or 1 1 msec), combined with the smallest possible slice thickness and field of view. In 89% of our healthy children and five of our patients, we found
partial
signal
loss from
the confluens
of sinuses
that
seemed not to correlate with pathologic changes. in stenosis or sinus thrombosis, the extent of the be overestimated, because the atheromatous thrombus creates a poststenotic turbulent flow loss adjacent to vessel boundaries, simulating a occlusion. It has been reported that fresh thrombus a high signal intensity, mimicking that of flowing
Particularly lesion can plaque or with signal high-grade can have blood, re-
suIting
[1 3]. How-
in missed
diagnosis
of sinus
thrombosis
ever, we found that in evaluating the original sections of the MR angiographic sequences, we had no difficulty in differentiating
between
Despite graphic
thrombus
the reasonable findings
with
DSA
and the laminar
anatomic findings,
flow of blood.
correlation MR
of MR angio-
angiography
is not
likely to be sufficient for the preoperative staging of aneurysms. At the current state of the art, small vessels arising near an aneurysm, knowledge of which is crucially important in planning neurosurgical strategies, are likely to be missed [10, 14]. For the evaluation of MR angiograms, rotatable projection images in two planes are recommended, and the original data set as well as conventional MR images should be included. This reduces the risk of overestimating the degree of a stenosis or underestimating the size of an aneurysm. The cine technique facilitates detection perimposed vessels can be easily
of vessels, distinguished
because from
sequences
jection for depiction not. Use of contrast
suone
another and their identification verified by rotating the images. Use ofcontrast material contributed little to the visualization of intracranial vessels in MR angiography. Whereas 3D
823
required
gadopentetate
dimeglumine
in-
of vessels, 3D FISP and 2D FLASH did material improves the depiction of vessels
only in patients with an intact blood-brain barrier, that is, when contrast material remains intravascular [1 6]. In patients with tumor, contrast-enhanced MR angiography can be an excellent tool for visualizing the lesion and adjacent vessels, but is not applicable for tumors with strong contrast enhancement
or tumors lying near the nasal cavity, as the high signal intensity of the lesion or the mucosal tissue can interfere with the vascular
signal.
As a noninvasive technique, MR angiography can be added to routine MR imaging, providing additional information about intracranial vessels and soft tissues in a single setting. It may be useful as a screening method when vascular disease is suspected, as a mapping technique before surgery or catheterization of abnormal vessels, or as a method for diagnosis and follow-up
but is eliminated
by the MIP algorithm on projection angiograms. High-resolution 3D FISP or FLASH sequences can be used to depict small intracranial vessels such as the anterior inferior cerebellar artery (AICA), posterior inferior cerebellar artery (PICA), anterior cerebral artery, and ophthalmic artery or small intracranial veins. A disadvantage of the 3D FLASH sequence is the simultaneous visualization of subcutaneous fatty tissue, which is superimposed on vessels in MIPs. This can diminished by using echo times that induce a fat-suppressive
FLASH
DISEASES
of sinus
thrombi
[1 0, 1 7]. Additional
clinical
experience and improvements in flow-compensation technique will expand the role of MR imaging in the diagnosis of intracranial vascular abnormalities. REFERENCES 1 . Brown DG, Riederer SJ, Jack CR, Farzaneh F, Ehman RL. MR angiography with oblique gradient-recalled echo technique. Radiology 1990;176:
461-466 2. Edelman RR, Mattle HP, Atkinson
DJ, Hoogewoud
HM. Magnetic
reso-
nance angiography. In: Cardiovascular imaging: ARRS categorical course syilabus. Reston, VA: American Roentgen Ray Society, 1990:51-60 3. KrayenbUhl H, Yasargil MG. Zerebrale Angiographie f#{252}r Klinik und Praxis.
Stuttgart:
Thieme Verlag, 1979:71-211
4. Edelman RR, Wentz KU, Mattle HP, et al. Intracerebral arteriovenous malformations: evaluation with selective MR angiography and venography. Radiology i989;173:831-837 5. Edelman RR, Hesselink JR. Clinical magnetic resonance imaging. Philadelphia: Saunders, 1990: 1 1 0-i 82 6. Angiography Numaris Il/Version A 2.1 . In: Magnetom SP user guide, 5th ed. Erlangen, Germany: Siemens AG, 1990 7. Ehricke H-H, Laub G. Integrated 3D display of brain anatomy and intracranial vasculature in MR imaging. J Comput Assist Tomogr 1990;14: 846-852 8. Lissner J, Solderer M. Klinische Kernspintomographie. Stuttgart: Ferdinand
Enke Verlag, 1990:59-83,
570-607
9. Masaryk TJ, Medic MT, Ruggieri PM, et al. Three-dimensional (volume) gradient-echo imaging of the carotid bifurcation: preliminary clinical experience. Radiology i989;1 71 :801-806 10. Sevick RJ, Tsuruda JS, Schmalbrock P. Three-dimensional time-of-flight MR angiography in the evaluation of cerebral aneurysms. J Comput Assist Tomogr 1990;14:874-881 1 1 . Marchal G, Bosmans H, Van fraeyenhoven L, et al. Intracranial vascular lesions: optimization and clinical evaluation of three dimensional time-offlight MR angiography. Radiology i990;1 75:443-448 12. Bongartz G, Vestnng TH, Fahrendorf G, Peters PE. Einsatz schneller Sequenzen beE der kraniozerebralen MR-Diagnostik. Fortschr Rontgenstr 1990;1 53:669-677 13. Nadel L, Braun IF, Kraft KA, Fatouros PP. Lame FJ. Intracranial vascular abnormalities: value of MR phase imaging to distinguish thrombus from flowing blood. AJNR 1990;1 1:1133-1140 14. Peters PE, Bongartz G, Drews C. Magnetresonanzangiographie der himversorgenden Artenen. Fortschr ROntgenstr i990;1 52: 528-533 15. Anderson C, Saloner D, Tsuruda J, Shapeero LG, Lee RE. Artifacts in maximum-intensity-projection display of MR angiograms. AJR 1990; 154: 623-629 16. Creasy JL, Price RR, Presbrey TH, Goins D, Partain L, Kessler RM. Gadolinium-enhanced MR angiography. Radiology i990;1 75:280-283 17. Schuierer G, Laub G, Huk WJ. MR angiography of the primitive trigeminal artery: report on two cases. AJNR 1990;1 1 :1131-1132