Get documents about "MRA visualization of cerebral aneurysms
W
Shared by: sleepover76
Categories
Tags
cerebral aneurysms, intracranial aneurysms, cerebral aneurysm, mr angiography, the brain, blood flow, unruptured aneurysms, intracranial aneurysm, subarachnoid hemorrhage, endovascular treatment, blood vessels, magnetic resonance angiography, brain aneurysm, neurosurgical residents, communicating artery
-
Stats
- views:
- 194
- posted:
- 11/5/2009
- language:
- English
- pages:
- 8
Document Sample
W. M. Adams1 , R. D. Laitt1, J. Thorne 2 and A. Jackson 3
MRA visualization of cerebral aneurysms
MR angiography With the development of newer techniques in Techniques
plays an increas- the endovascular treatment of intracerebral
ingly important aneurysms, magnetic resonance angiography The most frequently used methods of acquisition
role in endovas- (MRA) is playing an increasingly important role are the ‘bright blood’ sequences TOF (time-of-
cular treatment in the diagnosis, assessment and management of flight) and PCA (phase contrast angiography),
of cerebral this condition. Diagnosis has traditionally relied which rely on suppression of background station-
aneurysms. on conventional catheter angiography, but this is ary tissue and increased vessel-to-soft-tissue con-
invasive and potentially dangerous. Image reso- trast from the flow of moving spins. Flowing spins
Black blood lution and signal-to-noise ratios have significantly in T2-weighted ‘black blood’ MRA create a signal
techniques avoid improved with the introduction of improved void when compared to the surrounding static
artifacts arising gradient systems and new background suppres- tissue. This avoids some of the artifactual prob-
from haematoma sion techniques. This allows the interrogation of lems resulting from haematoma and turbulence
and turbulence. aneurysm morphology; the determination of associated with the bright blood techniques.
parent and branch vessel patency, and the assess-
ment of the ratio of aneurysm neck to fundus. Time-of-flight magnetic resonance
The latter is an important predictor of an angiography
aneurysm’s ability to retain a platinum coil.
In time-of-flight MRA, the longitudinal magne-
Although the treatment of choice in posterior tization of stationary tissue within a three-
circulation aneurysms, the endovascular tech- dimensional imaging volume becomes saturated
nique is still in its infancy and meticulous follow- into a steady state by multiple RF pulses. Flow-
up is required. MRA in the presence of platinum related enhancement occurs in those relaxed
coils is non-hazardous. One of its future roles may
be in the long term follow-up of these patients Fig. 1 b. A 3D reconstruction of the same aneurysm
to obviate repeated conventional angiography. using the black blood technique (TR 1000, TE 90,
NSA 1) more accurately portrays the aneurysm,
although there is still some artifact degradation due to
Fig. 1. A 62 year old female patient with subarachnoid haematoma. Because of signal void from the bony sphe-
haemorrhage due to rupture of an anterior communi- noid wing this is also included in the isosurface recon-
cating artery aneurysm. struction. Because of reduced inherent contrast between
1Department of Fig. 1 a. The TOF MRA 3D isosurface reconstruction vessel and soft tissue in the T2 weighted black blood
Neuroradiology, (TR 45, TE 7, NSA 1) is non-diagnostic. The aneurysm technique the 3D isosurface reconstruction does not
Central Manchester is obscured by high signal sub-acute thrombus due to work as well as the TOF sequence and shows some cut
Healthcare Trust, conversion of oxyhaemoglobin to methaemoglobin. off in the A2 segments of the anterior cerebral arteries.
Manchester,
England.
2Department of
Neurosurgery,
Central Manchester
Healthcare Trust,
Manchester,
England.
3Department of
Diagnostic
Radiology,
University of
Manchester,
Manchester,
England.
Volume 43 Issue 1
2 medicamundi March 1999
spins possessing full longitudinal magnetization TONE (Tilted Optimized Nonsaturating Excita-
which enter the 3D slab, providing vessel/soft- tion) is a variable flip-angle excitation technique
tissue contrast. The use of a 3D technique and which decreases spin saturation as the vessel
larger imaging matrices ensures small voxel size traverses an imaging volume. The flip angle is set
and reduced phase dispersion. Short TE values to a lower value at the entry side, and is gradu-
also minimize signal loss due to phase dispersion ally increased as it approaches the exit side,
in areas of complex flow or in vessel systole [1]. increasing signal intensity in the distal arteries.
However, signal is lost when the flow is in-plane. MOTSA (multiple overlapping thin slab acquisi-
Enhancement is at its maximum when the vessel tion) combines the features of a thin slice 2D
is perpendicular to the imaging plane. The further acquisition, and therefore increased sensitivity to
a vessel penetrates into the imaging volume, the slow flow, with the 3D features of smaller voxel
more likely it is to lose its longitudinal magnet- size (less phase dispersion) and the potential
ization. The maximum slice thickness is therefore for 3D reconstruction. The use of narrow slabs
limited because of saturation effects. The optimum minimizes saturation effects.
flip angle is a compromise between adequate
background tissue suppression and preservation MRA in aneurysmal subarachnoid haemorrhage
of signal in moving spins penetrating distally
into the imaging volume. There are two reasons for performing a prompt
MR examination in aneurysmal subarachnoid Improved
Improved background tissue suppression tech- haemorrhage. Firstly, TOF is relatively insensi- background
niques include magnetization transfer saturation tive to slow flow. MRA is therefore less sensitive suppression en-
(MTS), which uses an off-resonance radio- than conventional angiography to the effects of hances contrast
frequency pulse to saturate protons in the bound spasm, which commonly occurs 4–10 days after between brain
water of the brain parenchyma, and so enhance the initial ictus. This can be a cause of poor vessel parenchyma and
contrast between the parenchyma and the arteries. depiction and failure to visualize an aneurysm arteries.
on conventional angiography. Secondly, signal
Fig. 2 a. The lateral projection intra-arterial digital changes due to paramagnetic blood breakdown
subtraction angiogram of the left common carotid artery products affect aneurysm visibility, especially
demonstrates the left posterior communicating artery when there is extension into the brain paren-
aneurysm.
Fig. 2.
This 70 year old
female patient
presented with a
painful IIIrd nerve
palsy due to pressure
effect from a
left posterior
communicating
artery aneurysm.
Fig. 2 b. An axial projection of the MRA TOF MIP (TR 45, TE 7, NSA 1)
shows gradual loss of signal intensity along the medial border of the aneurysm due
to phase dispersion secondary to turbulence and slow flow. The posterior commu-
nicating artery is just visible but appears to be displaced medially when compared
to the contralateral side.
Fig. 2 c. Reference to the TOF source data shows that the aneurysm is much
larger than either the IA DSA or the TOF reconstructions suggest. The lumen of
the aneurysm is high signal but there is a large intra-mural hypointense thrombus
which is displacing the ipsilateral posterior communicating artery. The presence
of intramural thrombus is a relative contra-indication to treatment by endovas-
cular coil. With time the coil mesh may become impacted into the soft thrombus,
leading to recurrence of the aneurysm. This patient was treated surgically.
Volume 43 Issue 1
March 1999 medicamundi 3
chyma. Because of the higher oxygen tension Phase Contrast MR Angiography (PCA)
within CSF when compared with brain, conver-
sion of oxyhaemoglobin to methaemoglobin Spins moving within a magnetic gradient field
within the subarachnoid space tends to be accumulate phase, which is proportional to
Imaging should delayed. Methaemoglobin results in shortening velocity [5]. Stationary tissue yields no phase
be performed as of the T1 relaxation time, with a consequent high shift after the application of a bipolar magnetic
soon as possible artifactual signal which can obscure aneurysm gradient, allowing complete suppression of back-
to avoid artifacts visibility. Imaging should therefore be performed ground tissue, including those with short T1
from blood as soon as possible after the patient’s presenta- relaxation times such as fat or thrombus. Only
breakdown tion, in order to avoid high artifact signals from regions of flow appear as a bright signal. This
products. blood breakdown products [2] (Fig. 1). technique is therefore particularly appropriate in
When there is thrombus containing methaemo- those patients with peri-aneurysmal haematoma.
globin within the lumen of an aneurysm, the Using a TOF acquisition, vascular detail would
TOF technique may overestimate its size because be lost.
of its shortened T1 relaxation. If thrombus within
Imaging is an aneurysm is hypointense, then aneurysm size By manipulation of the amplitude and duration
performed on a will be underestimated (Fig. 2). These limitations of the bipolar magnetic gradient, the examina-
Philips Gyroscan also apply to conventional catheter angiography. tion can be tailored to particular flow velocities.
ACS NT 1.5 T It is important to look at either the source im- Giant aneurysms, subject to slow and turbulent
system with ages or conventional spin echo images in order flow, are particularly suitable for this technique.
quadrature to obtain an accurate assessment of an aneurysm’s It is important to ensure that the correct velocity
head coil. size. The sensitivity of prospective aneurysm encoding (Venc, measured in cm/s), is chosen.
detection also increases when reference is made If the setting is too high, signal will be lost from
to source images [4]. Imaging in our institution slower-flowing vessels. If the Venc chosen is too
is performed on a 1.5 T Philips Gyroscan small, velocity aliasing may occur with faster
ACS NT scanner using a quadrature head coil. flows being under-represented. Additional
A sagittal 2D phase contrast MR angiogram is information about the flow dynamics within an
used as a localizer. aneurysm can be obtained. The Venc can be
selected to target the inflow jet of an aneurysm
Parameters for TOF acquisition
or the slow vortex flow (Fig. 3). We normally
Repetition Time 45, Echo Time 7, NSA 1, select a Venc of 50 cm/s for intracranial arterial
Flip angle 20, Matrix 512, studies, and reduce the Venc to 20 or 30 cm/s
Slice thickness 0.7 mm, No. of slices 100. when imaging giant aneurysms.
Fig. 3 a. A right internal carotid injection oblique
angiogram during the early arterial phase with image Fig. 3 b. A 3D phase contrast MR angiogram (TR 14,
reversal shows high density contrast representing more TE 6, NSA 1) with a velocity encoding profile of 8 cm/s
rapid velocity flow lining a giant cavernous aneurysm. shows similar appearances to the angiogram. Signal
There is as yet no contrast opacification of the central intensity representing more rapid flow is seen around
portion of the aneurysm. the edges of the aneurysm.
Fig. 3.
A 73 year old
female patient who
presented with
progressive
headaches and
right-sided third
nerve palsy.
Volume 43 Issue 1
4 medicamundi March 1999
Parameters for PCA acquisition Like TOF, the black blood technique is relatively The black blood
insensitive to slow flow. Its niche within the technique has a
Repetition Time 14, Echo Time 6, NSA 1, head probably lies in the demonstration of those niche in demon-
Flip angle 15, Matrix 256, No. of slices 50, vessels surrounded by soft tissue such as the strating vessels
Slice thickness 1–2 mm, Field of view 230 mm. middle or anterior cerebral arteries. Resolution surrounded by
can be improved with the use of a surface coil. soft tissue.
Black blood MRA
Parameters for Black Blood acquisition
Bright blood techniques such as time-of-flight Repetition Time 850 –1000, Echo Time 70 –90, In the black
and phase contrast MRA, which make flowing NSA 1, Flip angle 90, Matrix 256, blood technique,
protons hyperintense when compared with sur- Slice thickness 0.8 mm, No. of slices 50. signal loss due to
rounding tissue, are prone to artifacts when flow turbulence
is slow. This occurs in complex aneurysms or in Display contributes to
turbulent vessels distal to stenoses. An alternative the desired
approach is to depict the vessel lumen itself by Although MR acquisition techniques are becom- signal void.
enhancing the signal void created by flowing ing increasingly sophisticated, the conventional
protons. This renders the vessel black in contrast method of post-processing has remained the
to the surrounding stationary tissue. This tech- maximum-intensity projection (MIP). This
nique is called ‘black blood’ or ‘dark blood’ provides excellent contrast as long as the vessel
angiography [6]. We use a turbo spin echo intensity is about 2 SD above the mean back-
sequence combined with flow presaturation and ground noise [7]. If there is a reduction in signal
thin sections to increase sensitivity to slow flow. intensity when compared with background tissue
then, as more sections are included, the overall
Inflowing blood, having acquired a 90° excitation fluctuations in noise may exceed the vascular
pulse, will exit the imaging slice before the appli- signal response and result in poor visualization.
cation of the 180° rephasing pulse, and therefore Apparent vessel width is reduced because of the
return no signal. Its advantage over bright blood lower signal intensity in marginal pixels due to
techniques is that signal loss due to turbulence, laminar flow and partial volume effects. Turbu-
which would otherwise cause image degradation, lence, slow flow and small vessels in relation
contributes to the desired signal void. to pixel size also contribute to poor vessel or
aneurysm depiction. The MIP display also makes
Instead of the maximum intensity projection it difficult to visualize depth information.
(MIP) method used in bright blood techniques,
black blood requires a minimum intensity pro- Alternative postprocessing algorithms have been Alternative post-
jection (MINIP) which is designed to depict only introduced which improve the visualization processing algo-
those pixels at least 2 standard deviations (SD) and characterization of known aneurysms [8]. rithms improve
below background intensity. However, 3D Surface modelling techniques place a surface on visualization.
reconstructions are problematical because of the boundaries within a three-dimensional data vol-
low inherent vessel/soft-tissue contrast. Viewing ume using a connectivity algorithm. A geometric
the data as multiplanar reformats has proved surface is generated for those data values that lie
more useful in our institution (Fig. 4). within a range specified by the user. Some form
of segmentation is required, usually in the form
Black blood images have found an application in of pixel intensity thresholding. Once a solid struc-
the extracranial carotid vessels, but are problem- ture has been created, other computer graphic
atic in the intracranial circulation because of the algorithms can be introduced. These include
intimate relationship of the internal carotid illumination rendering which produces a 3D
artery to the skull base. Bone produces a signal shaded image with depth features, a smoothing
void and is therefore difficult to differentiate algorithm which reduces the effect of noise and
from the signal void produced from flowing removes isolated pixels, and geometric transfor-
blood. In an attempt to address this problem we mations which allow real time movement and
angle the imaging slab in alignment with the rotation of the selected data, so that the vascular
clivus, in order to include the Circle of Willis structures can be viewed from any angle. This is
vessels, but avoid contact with bone. of particular importance in selecting the optimum
Volume 43 Issue 1
March 1999 medicamundi 5
Fig. 4.
This 55 year old
female patient
presented with IIIrd
and VIth nerve
palsies due to a
giant right sided
fusiform intracav-
ernous aneurysm.
Fig. 4 a. Lateral projection MRA TOF MIP reconstruc- Fig. 4 b. A lateral projection 3D TOF isosurface recon-
tion (TR 45, TE 7, NSA 1) showing the fusiform intra- struction using the same source data demonstrates the
cavernous aneurysm. Because of turbulence and slow fusiform aneurysm. No neck could be defined on any
flow within the aneurysm there is non-homogeneous sig- projection making this unsuitable for endovascular coil
nal loss, particularly in the fundus, due to phase inco- insertion. The surface reconstruction technique produces
herence. Streamlining and signal loss is evident within a smooth endoluminal margin avoiding the signal loss
the carotid siphon, partly due to turbulence, and also seen on the MIP.
due to susceptibility artifact from the adjacent sphenoid
sinus. This results in some artifactual reduction in
apparent vessel calibre.
Fig. 4 c. The coronal single-slice black blood MRA Fig. 4 d. The lateral IA study taken during the late
image (TR 1000, TE 90, NSA 1) shows the complex venous phase shows the profound stasis of contrast
internal structure of the aneurysm. Medially there is within the aneurysm extending into the terminal ICA.
hypointense flowing blood; centrally hyperintense acute
thrombus or slow flowing blood; and laterally a crescent
of hypointense thrombus. Some ghosting is seen in the
phase-encoding direction, due to motion artifact.
projection angle when considering treatment tor dependent, and requires manual input for
via the endovascular route, potentially reducing thresholding vascular structures and for selecting
procedure duration and radiation exposure. volumes of interest. The whole procedure takes
approximately 10 –15 minutes, but a thorough
In our institution both standard MIP and 3D technique is required. More than one reconstruc-
isosurface reconstructions are performed using a tion may be needed. In creating a surface around
Philips EasyVision Release 2.12 Workstation. a manually detected threshold, the signal void
The 3D isosurface segmentation editor is opera- that one sees in MIP reconstructions of giant
Volume 43 Issue 1
6 medicamundi March 1999
aneurysms is avoided (Fig. 4). Dextrous thresh- prior to endovascular therapy. The depth and
olding may also minimize the artifact due to shading algorithms allow a sense of spatial per-
peri-aneurysmal haematoma. Because the thresh- spective which give a more ‘tangible’ image and
old algorithm is to a certain extent subjective, allows a more precise appreciation of compli-
setting it too low may result in the elimination cated structures, particularly when the image is
of small parent or branch vessels. Analysis of the actively manipulated on the workstation rather
source data and assessment of the target aneurysm than viewed as a hard copy (Fig. 5). Conven-
from the diagnostic angiogram are important tional angiograms can be difficult to interpret
pre-reconstruction steps. Setting the threshold because of the presence of overlying vessels.
level too high can enhance artifacts and create
artificial contours or holes [9]. This is also true when applied to MIP images,
particularly around the carotid siphon. This is
We feel the 3D isosurface reconstruction offers less problematic with the isosurface techniques, Isosurface
several advantages when compared to the MIP and is partly overcome by selective targetting reconstruction
display. It provides a more faithful representation of the MRA image. In some cases aneurysm offers a realistic
of the angiographic appearances in terms of visualization is easier on the MRA than on view of aneurysm
aneurysm morphology, the ratio of fundus to the diagnostic angiogram (Fig. 6). Aneurysms morphology.
neck and the patency of parent or branch vessels. visualized on MRA but not seen at angiography
It is of particular value in assessing a patient have been reported in the literature [10].
Fig. 5 b. Right internal carotid artery injection per-
Fig. 5 a. At the time of the diagnostic angiogram there formed prior to attempted endovascular coil insertion.
is no evidence of vessel spasm. The aneurysm is easily There is now severe vessel spasm of the parent A1 and
seen, but the relationship of the neck to the A2 segments daughter A2 vessels. The relationship of the A2 vessels
of the anterior cerebral artery is not clear. to the aneurysm neck is still not clear.
Fig. 5.
This 65 year old
female patient
presented with
subarachnoid
haemorrhage due
to rupture of
an anterior
communicating
artery aneurysm.
Fig. 5 c. The aneurysm has been catheterized with a
microcatheter and a cautious intra-aneurysmal injec-
tion performed. The A2 segments are opacified confirm-
ing their origin from the neck of the aneurysm. This is Fig. 5 d. A TOF MRA 3D isosurface reconstruction
a relative contra-indication to coil treatment because (TR 45, TE 7, NSA 1) of the same aneurysm viewed
occlusion of the aneurysm would also sacrifice the A2 from behind and slightly to the right. The right A2
vessels. vessel is clearly seen arising from the aneurysm itself.
Volume 43 Issue 1
March 1999 medicamundi 7
Fig. 6.
This 29 year old
male patient
presented with
a subarachnoid
haemorrhage
diagnosed on CT.
The location of the
blood suggested
an anterior
communicating
artery aneurysm.
Fig. 6 a.
The IA DSA is Fig. 6 b. A TOF MRA 3D isosurface reconstruction
inconclusive. On (TR 45, TE 7, NSA 1) unequivocally demonstrates a
this perorbital teardrop saccular aneurysm originating from the junc-
projection the tion between the right A1 segment of the anterior cere-
A1/A2 junction bral artery and the anterior communicating artery.
is partly obscured The ratio of neck to fundus suggests that this aneurysm
by the overlying would be suitable for endovascular treatment.
ophthalmic artery.
Other projections
were also Future uses of MRA to exclude the aneurysm from the circulation by
non-diagnostic.
filling it with platinum microcoils (Fig. 7). The
Guglielmi Detachable Coils (GDC) were first long term occlusion rates of completely packed
used clinically for the treatment of intracranial aneurysms, or the potential for partly packed
aneurysm in 1990. The aim of the technique is aneurysms to re-bleed, have not yet been estab-
Fig. 7 a. A lateral IA DSA 6 month check angiogram
shows complete occlusion of the posterior communicat- Fig. 7 b. An oblique projection confirms complete
ing artery aneurysm with no compaction of the coil occlusion of the aneurysm. The double density is
ball mesh. A double density (arrow) overlies the supra- revealed as a separate ophthalmic segment aneurysm.
clinoid portion of the internal carotid artery.
Fig. 7.
This 41 year old
male patient
presented with
aneurysmal
subarachnoid
haemorrhage due
to rupture of a
lobulated posterior
communicating
artery aneurysm.
The patient was
treated by the
endovascular route Fig. 7 d. Lateral MRA TOF MIP reconstruction of the
with insertion of same data confirms the absence of recurrence of the
four GDC coils, treated aneurysm. There is no attenuation of the ICA on
with complete Fig. 7 c. 3D isosurface TOF reconstruction (TR 45, the reconstruction due to the presence of a coil ball mesh.
occlusion. TE 7, NSA 1) with a similar oblique projection to
the angiogram demonstrates the ophthalmic segment
aneurysm. The venetian blind effect is caused by over-
lap of imaging volumes.
Volume 43 Issue 1
8 medicamundi March 1999
lished. In some patients the aneurysm may recur Fig. 8.
This 55 year old
because of compaction, or because of growth of female patient with
a residual aneurysmal neck. The future role of a short history of
MRA may be in the long term follow-up of these headaches showed
angiographic evi-
treated patients [11].
dence of rupture of
a posterior commu-
Subjecting patients with coils in situ to an exter- nicating artery
nal magnetic field is non-hazardous, and the aneurysm during
a diagnostic
degree of artifact produced on gradient echo angiogram, and
images is minimal. Hartman et al. [12] placed suffered a fit after
coils within a magnetic field of 1.5 T and showed the procedure. She
was treated acutely
no magnetic deflection or rotational force on by the endovascular
suspended coils, or any significant thermal effect. Fig. 8 a. A 6 month check angiogram shows a small route.
Some artifact was seen in vitro on all imaging neck remnant due to coil ball compaction.
sequences, namely a thin rim of high signal in
MRA may have
the frequency encoding direction both posterior
a future role in
and anterior to the coil mesh. These changes
the long-term
were mirrored in vivo with, in addition, obscura-
follow-up of
tion of about 2–3 mm of surrounding tissue.
treated patients.
Despite this, MRA still provides useful diagnos-
tic information. High signal within an aneurysm
is suggestive of residual flow due to coil com-
Fig. 8 b. A lateral projection
paction, recurrence, or flow within the coil TOF MRA 3D isosurface
interstices (Fig. 8). However, haemorrhage and reconstruction (TR 45, TE 7,
artifacts can mimic residual aneurysm flow. NSA 1) demonstrates the neck
remnant and preserved patency
MRA is also able to assess parent or branch vessel of the posterior communicating
occlusion. artery.
References
[1] Graves MJ. Magnetic Resonance Angiography. [8] Atlas SW, Sheppard L, Goldberg HI, Hurst RW,
Review Article. The British Journal of Radiology Listerud J, Flamm E. Intracranial Aneurysms:
1997; 70: 6 –28. Detection and Characterization with MR Angio-
[2] Wilman AH, Huston J, Riederer SJ. Three- graphy with use of an Advanced Postprocessing
Dimensional Magnetization – Prepared Time-of Technique in a Blinded Reader Study. Radiology
Flight MR Angiography of the Carotid and 1997; 203: 807– 814.
Vertebral Arteries. MRM 1997; 37: 252 –259.
[9] Bontozoglou NP, Spanos H, Lasjaunias P, Zarifis
[3] Ida M, Kurisu Y, Yamashita M. MR Angiography G. Intracranial Aneurysms: Endovascular Evalua-
of Ruptured Aneurysms in Acute Subarachnoid tion with Three-Dimensional-Display MR Angio-
Hemorrhage. AJNR 1997; 18: 1025 –1032. graphy. Radiology 1995; 197: 876 – 879.
[4] Huston J III, Nichols DA, Luetmer PH et al.
[10] Curnes JT, Shogry ME, Clark DC, Elsner HJ.
Blinded Prospective Evaluation of Sensitivity of
MR Angiographic Demonstration of an Intra-
MR Angiography to Known Intracranial
cranial Aneurysm not seen on Conventional
Aneurysms: Importance of Aneurysm Size. AJNR
Angiography. AJNR 1993; 14: 971–973.
1994; 15: 1607–1614.
[5] Korosec FR, Turski PA. Velocity and Volume [11] Derdeyn CP, Graves VB, Turski PA, Masaryk
Flow Rate Measurements Using Phase Contrast AM, Strother CM. MR Angiography of Saccular
Magnetic Resonance Imaging. International Aneurysms after Treatment with Guglielmi
Journal of Neuroradiology 1997; 3: 293 – 318. Detachable Coils: Preliminary Experience. AJNR
[6] Ekelman RR, Mattle HP, Wallner B et al. 1997; 18: 279 –286.
Extracranial Carotid Arteries: Evaluation with
[12] Hartman J, Nguyen T, Larsen D, Teitelbaum
‘Black Blood’ MR Angiography. Radiology 1990;
GP. MR Artifacts, Heat Production and Ferro-
177: 45 –50.
magnetism of Guglielmi Detachable Coils. AJNR
[7] Anderson CM, Saloner D, Tsuruda JS, Shapeero 1996; 18: 497–501.
LG, Lee RE. Artifacts in Maximum-Intensity-
Projection Display of MR Angiograms. AJR 1990; [13] Schievink WI. Genetics of Intracranial
154: 623 – 629. Aneurysms. Neurosurgery 1997; 40: 651– 663.
Volume 43 Issue 1
March 1999 medicamundi 9
Related docs
