AJR 2003; 181:242-244
© American Roentgen Ray Society
Three-Dimensional Time-of-Flight Subtraction Angiography of Subacute Cerebral Hemorrhage
Jimmy H. M. Chan1,
Wilfred C. G. Peh2,
Edmund Y. K. Tsui1,
Kenneth P. C. Wong1 and
M. K. Yuen1
1 Department of Diagnostic Radiology, Tuen Mun Hospital, Tsing Chung Koon Rd.,
Hong Kong, China.
2 Program Office, Singapore Health Services. 02-09, 7 Hospital Dr., Singapore
169611.
Received October 24, 2001;
accepted after revision January 2, 2003.
Address correspondence to J. H. M. Chan
Introduction
Three-dimensional (3D) time-of-flight (TOF) MR angiography has been adopted
as a technique for routine assessment of stenosis and occlusion of
intracranial blood vessels. Difficulties in making a diagnosis arise when
patients present with subacute hemorrhage in the region of interest
[1,
2] because methemoglobin is
hyperintense on 3D TOF MR angiograms. Incorporation of this signal
hyperintensity to the maximum-intensity-projection reconstructed images may
mask the arteries of interest or mimic vascular abnormalities. This study
aimed to design an MR digital subtraction technique capable of removing
hyperintense hematoma from 3D TOF MR angiograms.
Subjects and Methods
All MR imaging was performed using a 1.5-T whole-body MR scanner (Signa
Horizon, Echo-speed, software version 5.8, General Electric Medical Systems,
Milwaukee, WI) equipped with high-speed gradients. Maximal gradient strength
was 23 mT/m and slew rate was 120 T/m/ per second. A circularly polarized head
coil was used.
Five adult patients (three men, two women; mean age, 53.5 years) with
findings of subacute hemorrhage around the circle of Willis on conventional MR
imaging were selected for the study. After completion of conventional MR
imaging using the standard brain protocol, 3D TOF MR angiography was performed
with a thick axial slab centered on the circle of Willis. The imaging
parameters for the 3D TOF MR angiogram pulse sequence were TR/TE, 36/6.9; flip
angle, 20°; number of partitions, 64; partition thickness, 1.2 mm; matrix
size, 512 x 224; field of view, 22 x 16.5 cm; receiver bandwidth,
± 15.6 kHz; and number of excitation, 1. Flow compensation and
magnetization transfer saturation techniques were used. A second set of MR
angiograms was then acquired by repeating the data acquisition using the same
pulse sequence, imaging parameters, and identical slice locations except that
superior and inferior spatial saturation (i.e., presaturation) slabs were
added. The purpose of applying presaturation slabs was to create flow voids of
all intracranial blood vessels in the second set of MR angiograms.
After data acquisition, all the angiographic base images were transferred
to the Advantage Windows workstation (software version 2.0, General Electric
Medical Systems) for postprocessing. The first set of MR angiographic base
images was subtracted from the second set of images on a pixel-by-pixel basis,
thus producing a third set of subtracted base images. Multiplanar angiograms
were obtained by maximum-intensity-projection reconstructions of the first set
of base images and the third set of subtracted base images. The two resulting
sets of MR angiograms were qualitatively assessed and compared.
Results
All patients had typical MR features of subacute hematoma, which appeared
hyperintense on both T1- and T2-weighted MR images. In two patients, the site
of hemorrhage was remote from the circle of Willis. Hence, the blood vessels
were not masked by hyperintense hematoma on MR angiography. In the other three
patients, the site of hemorrhage was close to the circle of Willis. Therefore,
part of the area of the blood vessels on the MR angiograms was obscured by the
hyperintense hematoma. In all patients, the hyperintense hematoma was
completely removed from the MR angiograms in the third set of subtracted base
images.
As an illustrative case, 3D TOF MR angiograms of a 44-year-old man with a
history of hypertension who presented with sudden onset of right-sided
weakness showed a large hematoma obscuring the M2 and M3 segments of the left
middle cerebral artery (Fig.
1A). A left vertebral artery aneurysm was displayed. The hematoma
was removed on subtracted 3D TOF MR angiograms, and the M2 and M3 segments
were clearly visible. No evidence was seen of an aneurysm (Figs.
1B and
1C). The left middle cerebral
artery was slightly attenuated. The left vertebral artery aneurysm was well
shown and was not affected by the subtraction technique.
View larger version (92K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A. —44-year-old hypertensive man with sudden onset of right-sided
weakness. Coronal maximum intensity projection of three-dimensional (3D)
time-of-flight (TOF) MR angiogram of circle of Willis shows large subacute
hematoma obscuring M2 and M3 segments of left middle cerebral artery. Note
left vertebral artery aneurysm (arrowhead).
|
|
View larger version (92K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B. —44-year-old hypertensive man with sudden onset of right-sided
weakness. Coronal maximum intensity projection of 3D TOF MR angiogram of
circle of Willis acquired with spatial saturation slabs placed immediately
above and below acquisition volume shows blood vessels as dark. Note that
subacute hematoma still appears bright.
|
|
View larger version (112K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C. —44-year-old hypertensive man with sudden onset of right-sided
weakness. Subtracted coronal maximum intensity projection of 3D TOF MR
angiogram clearly shows M2 and M3 segments after removal of hematoma. Note
lack of evidence of circle of Willis aneurysm. Left vertebral artery aneurysm
(arrowhead) remains easy to see.
|
|
Another patient with no clinical history of hypertension was suspected of
having an arteriovenous malformation. The 3D TOF MR angiograms showed a
hematoma (Figs. 2A and
2B) obscuring the M2 and M3
segments of the right middle cerebral artery. After the high signal of the
hematoma had been removed, no sign of arteriovenous malformation or aneurysm
was seen (Fig. 2C) except for
some attenuation in the distal right middle cerebral artery branches.
View larger version (168K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A. —35-year-old woman with no clinical history of hypertension
who presented after frequent episodic unilateral headaches. Axial
three-dimensional (3D) time-of-flight (TOF) MR angiogram shows subacute
hematoma (arrows) near distal end of right middle cerebral artery.
Such findings cannot be excluded on postprocessing.
|
|
View larger version (142K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B. —35-year-old woman with no clinical history of hypertension
who presented after frequent episodic unilateral headaches. Coronal maximum
intensity projection of 3D TOF MR angiogram of circle of Willis shows subacute
hematoma (arrows) obscuring M2 and M3 segments of right middle
cerebral artery.
|
|
View larger version (134K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2C. —35-year-old woman with no clinical history of hypertension
who presented after frequent episodic unilateral headaches. Subtracted coronal
maximum intensity projection of 3D TOF MR angiogram shows M2 and M3 segments
after removal of hematoma. Note lack of arteriovenous malformation or
aneurysm, although distal middle cerebral artery branches are attenuated.
|
|
A 41-year-old woman with no known history of hypertension presented with an
acute onset of mental confusion and disorientation. On MR imaging, a large
subacute hematoma was found at the circle of Willis. A 3D TOF MR axial source
angiogram showed a large hematoma and two adjacent small suspicious lesions
(Fig. 3A). These lesions were
present on multiple source images and could represent parts of the hematoma or
abnormal vessels such as a vascular malformation. On subtracted axial
maximum-intensity-projection (collapsed) images, the two suspicious lesions
disappeared, indicating that they were indeed part of the hematoma
(Fig. 3B).
View larger version (139K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A. —41-year-old woman with no clinical history of hypertension
who presented with sudden onset of mental confusion and disorientation. Axial
three-dimensional (3D) time-of-flight (TOF) MR angiogram shows large subacute
hematoma and two small lesions (arrowheads) that are suspected to be
parts of either hematoma or aneurysm.
|
|
View larger version (117K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B. —41-year-old woman with no clinical history of hypertension
who presented with sudden onset of mental confusion and disorientation.
Subtracted axial maximum intensity projection of 3D TOF angiogram shows that
suspicious lesions have disappeared, indicating that they were parts of
hematoma.
|
|
Discussion
Three-dimensional TOF MR angiography makes use of a TR that is much shorter
than the T1 relaxation time of stationary tissue. As a result, there is signal
saturation of the stationary tissue. In contrast, the fresh, unexcited moving
blood entering the excitation volume between pulse sequence repetitions is not
saturated and thus has high signal intensity compared with that of the
surrounding stationary tissue.
The main blood product of a subacute hematoma is either intracellular or
extracellular methemoglobin. Methemoglobin is paramagnetic, resulting in very
short T1 relaxation time. Thus, the longitudinal magnetization of the subacute
hematoma can substantially recover, even though the TR used with the 3D TOF
pulse sequence is short. This accounts for the signal hyperintensity of a
subacute hematoma on the 3D TOF MR angiograms. When superior and inferior
spatial saturation slabs are placed immediately above and below the
acquisition volume to saturate the fresh unexcited blood spins entering the
acquisition volume, all blood vessels are made to appear dark (i.e., flow
voids) on MR angiograms. The subacute hematoma, however, still appears bright.
Digital subtraction between the 3D TOF MR angiograms acquired without spatial
saturation slabs and those acquired with superior and inferior spatial
saturation slabs gives rise to a third set of MR angiograms with bright blood
vessels but with the subacute hematoma removed
[3]. When digital subtraction
is used between the 3D TOF MR angiograms and the avascular 3D TOF MR
angiograms, patient movement should be kept to the minimum needed to obtain
useful information. Apart from patient education and firm fixation of the head
in the head coil by straps, the 3D TOF MR angiogram sequence should be
immediately followed by the avascular 3D TOF MR angiogram sequence so as to
minimize the time gap between the two sequences.
Another MR angiographic technique, the 3D phase contrast method, may also
be used to eliminate signal hyperintensity of the subacute hematoma. Its
principle of action depends on the velocity-induced phase shift, and there is
no blood flow within the hematoma. However, the 3D TOF method possesses a
number of unique advantages over the 3D phase contrast method. For example,
the 3D TOF method is more useful than the 3D phase contrast method for
revealing an aneurysm because of its greater spatial resolution using a larger
matrix size (512 x 512). Furthermore, because the data acquisition time
of the 3D phase contrast method is long, a low-resolution matrix (256 x
128) is commonly used. Consequently, the larger voxel size of the 3D phase
contrast method results in a greater intravoxel phase dispersion, leading to a
larger signal loss from turbulent flow
[4]. The total time for
acquiring two sets of 3D TOF MR angiograms is comparable to that for acquiring
one set of 3D phase contrast MR angiograms, in the range of 10–12 min.
The processing time for maximum-intensity-projection reconstruction is
identical for 3D TOF MR angiography and 3D phase contrast MR angiography, and
the time for the digital subtraction process is less than 30 sec. One
trade-off of this technique is that any thrombus will also be removed even
though the aneurysm remains intact. Conventional 3D TOF MR angiography is
helpful in this regard.
In conclusion, 3D digital subtraction TOF MR angiography can be used to
remove the subacute hematoma from the MR angiograms, and hence improve
visualization of the underlying blood vessels.
References
- Wilcock DJ, Jaspan T, Worthington BS. Problems and pitfalls of 3-D
TOF magnetic resonance an giography in intracranial circulation.
Clin Radiol1995; 50:526
–532[Medline]
- Carriero A, Palumbo L, Tonni AG, D'Angelo C, Magarelli N, Bonomo L.
Diagnostic pitfalls in magnetic resonance angiography [in Italian].
Ra diol Med1995; 90:719
–725
- Takano K, Utsunomiya H, Ono H, Okazaki M, Tanaka A. Dynamic
contrast-enhanced subtraction MR angiography in intracranial vascular
abnormalities. Eur Radiol1999; 9:1909
–1912[Medline]
- Scarabino T, Carriero A, Magarelli N, et al. MR angiography in
carotid stenosis: a comparison of three techniques. Eur J
Radiol 1998;28:117
–125[Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Introduction
Subjects and Methods
