DOI:10.2214/AJR.05.1249
AJR 2006; 187:1637-1643
© American Roentgen Ray Society
Comparison of CT Venography with MR Venography in Cerebral Sinovenous Thrombosis
N. Khandelwal1,
Ajay Agarwal1,2,
Rohit Kochhar1,
J. R. Bapuraj1,
Paramjeet Singh1,
S. Prabhakar3 and
S. Suri1
1 Department of Radiodiagnosis and Imaging, Postgraduate Institute of Medical
Education and Research (PGIMER), Chandigarh 160012, India.
2 Present address: Kesri Bhavan, 2/1/1B Munshi Bazar Rd., Kolkata 700015,
India.
3 Department of Neurology, Postgraduate Institute of Medical Education and
Research (PGIMER), Chandigarh, India.
Received July 19, 2005;
accepted after revision October 25, 2005.
Address correspondence to N. Khandelwal
(khandelwaln{at}hotmail.com).
Abstract
OBJECTIVE. The purpose of this study was to compare cerebral CT
venography with MR venography and determine the reliability of CT venography
in the diagnosis of cerebral sinovenous thrombosis.
SUBJECTS AND METHODS. Fifty patients who were clinically suspected
of having cerebral sinovenous thrombosis, irrespective of age and sex,
underwent cerebral CT venography and MR venography. Projection venograms were
displayed using maximum-intensity-projection images for both CT venography and
MR venography. The CT venograms were also displayed using the integral
algorithm, which depicts the average intensity value of the first five voxels
deep in relation to the model surface that is nearest the viewer, allowing
direct visualization of the thrombus in the sinuses. All CT venograms and MR
venograms were independently evaluated by experienced neuroradiologists.
RESULTS. Of these 50 patients, 30 patients were diagnosed as having
cerebral sinovenous thrombosis on both CT venography and MR venography. The
total numbers of sinuses involved were 81 and 77 (CT venography and MR
venography). When MR venography was used as the gold standard, CT venography
was found to have both a sensitivity and a specificity of 75-100%, depending
on the sinus and vein involved.
CONCLUSION. CT venography is as accurate as MR venography for
diagnosing cerebral sinovenous thrombosis.
Keywords: cerebral sinovenous thrombosis • CT • MRI • neuroradiology • venography
Introduction
Cerebral sinovenous thrombosis, or cerebral venous thrombosis (CVT), as a
cause of serious neurologic symptoms and a fatal outcome was first described
in the early 19th century [1].
The true incidence of CVT is unknown because of lack of adequate epidemiologic
studies [2]. Intracranial dural
sinus thrombosis is a relatively common and potentially fatal condition. The
diverse clinical presentations and lack of accurate diagnostic techniques have
made CVT a difficult diagnosis with a grave prognosis. That milder forms of
CVT can now be recognized and that most patients with CVT recover with
recanalization of the thrombosed blood vessel have contributed to the decrease
in mortality [3]. Various
radiologic techniques have been used to visualize the intracranial venous
system. Conventional and digital subtraction cerebral angiography, CT, MRI,
and recently, MR venography and CT venography have increased our ability to
detect this condition.
MRI, including MR venography, is now established as the imaging technique
of choice for the immediate evaluation and follow-up of CVT
[4]. With the advent of helical
CT, 3D vascular imaging has also become possible, and large volumes of tissue
can be scanned during peak arterial and venous enhancement, maintaining high
spatial resolution. The term "CT venography" was first used by
Casey et al. [5], who described
the technique as a rapid method of depicting the intracranial venous
circulation with consistently high quality. CT venography can be instantly
performed as an adjunct to unenhanced CT in patients undergoing an initial
workup for CVT. Because the scanning duration is less than 1 minute, image
quality is hardly impaired by patient motion, and patient monitoring is easier
in critically ill patients as compared with MRI. The purpose of this study was
to determine the reliability of CT venography in diagnosing CVT using MR
venography as the gold standard. MR venography has well-known shortcomings in
diagnosing venous thrombosis, but because no gold standard exists, MR
venography was the reference standard in our series.
View larger version (153K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1 —18-year-old woman with sudden onset of headache and seizures.
Unenhanced CT of head in axial plane shows cord sign in straight sinus
(arrow) and "dense vein" sign in superior sagittal sinus
(arrowhead). No parenchymal changes are present. Both CT venography
and MR venography confirmed presence of venous sinus thrombosis.
|
|
Subjects and Methods
Fifty patients who were clinically suspected of having CVT, irrespective of
age and sex, and who presented to emergency neurology services of our
institution between May 2003 and October 2004, underwent preliminary
unenhanced CT followed by CT venography. Within 24 hours of CT venography, MRI
of the brain followed by an MR venography sequence based on the 2D
time-of-flight technique was performed in the coronal oblique plane. The study
was approved by our institutional review board.
The median age of the patients—30 women and 20 men—was 32 years
(range, 18-57 years). All CT venography was performed on an MDCT scanner
(LightSpeed, GE Healthcare) with an Advantage Windows (version 4.0, GE
Healthcare) 3D workstation. A routine unenhanced CT scan was obtained with
5-mm-thick contiguous axial sections from the base of the skull to the vertex,
followed immediately by CT venography. The scanning direction was from vertex
to skull base. Scans were angled parallel to a line drawn from the posterior
margin of the foramen magnum to the superior margin of the orbit to exclude
the lens.
The scanning parameters used were a slice thickness of 2.5 mm at an
interval of 1.25 mm. The gantry rotation speed was 3.5-7.5 mm per rotation
using 120 kV and 250-300 mA. A prescan delay of 30-40 seconds was used with a
display field of view of 23-25 cm. A total of 70-80 mL of nonionic contrast
material (iodixanol, 270 mg I/mL) was administered at a rate of 3-4 mL/s by a
power injector into an antecubital vein.
Unenhanced CT scans of the patients were analyzed for parenchymal lesions
and for the presence of the cord sign and the "dense vein" sign.
CT venograms were analyzed for filling defects in the dural sinuses and for
indirect evidence of CVT in the form of collaterals and tentorial enhancement.
The source images were displayed with an approximate window setting of 400-450
H and a level of approximately 130-150 H to clearly visualize the dural venous
sinuses separately from adjacent bone. The acquired CT data were processed
using the Advantage workstation. Source images were reconstructed in coronal
and sagittal planes using oblique multiplanar reconstruction (MPR) of
appropriate window settings—a window width of 400-450 H and a level of
120-150 H—to separate sinuses from adjacent bones. Reformatted images
were displayed using maximum intensity projection (MIP).
Using CT SOFT computer software (GE Healthcare) on the Advantage
workstation, 3D CT volume-averaging images were reconstructed after removing
the bone structures by applying a threshold adequate to include only the bone
and to exclude the venous sinus. Projection venograms were displayed using MIP
and integral algorithms. The integral display technique gave a 3D projection
image and allowed direct visualization of the thrombus and the extent of sinus
involvement [5].
All MR venography was performed on a 1.5-T MRI system (Vision, Siemens
Medical Solutions). Preliminary MRI of the brain was performed using the
following MR sequences: T1-weighted spin-echo (axial plane), proton density-
and T2-weighted turbo spin-echo (axial plane), diffusion-weighted MRI (axial
plane), and FLAIR MRI (coronal plane).
These sequences were followed by an MR venography sequence based on the 2D
time-of-flight technique in a coronal oblique plane to minimize the effects of
in-plane saturation by keeping the plane of acquisition at 90° to the
general anteroposterior direction of the venous sinuses. Our parameters were
TR range/TE range, 32-40/8-12; flip angle, 50-70°; slice thickness,
1.5-3.0 mm; and matrix size, 144 x 256. The total number of acquisitions
was 1 or 2, and the total acquisition time was 6-8 minutes. In the
time-of-flight MR venography sequence, an inferior saturation band was applied
at the level of the carotid bifurcation to eliminate signals from the arterial
structures. After the acquisition of all source images of the MR venography
sequence, the images were processed and displayed by means of an MIP algorithm
using computer software.
MR venography images (both source images and MIP images) were analyzed for
direct evidence of CVT, which included a lack of typical high-flow signal from
a sinus that does not appear aplastic or hypoplastic on base images of MR
venographic sequences, and the frayed appearance of the flow signal from a
sinus after recanalization. The indirect evidence of CVT included formation of
collaterals over the extracranial veins, unusually prominent flow signal from
deeper medullary veins, cerebral hemorrhage, visualization of emissary veins,
and signs of increased intracranial pressure.
All CT and MR venograms were independently evaluated by two experienced
neuroradiologists for evidence of CVT.
View larger version (114K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —45-year-old woman with headache and altered sensorium.
T1-weighted MR image in sagittal plane shows hyperintense signal in superior
sagittal sinus and loss of normal flow void (arrows).
|
|
View larger version (123K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A —20-year-old man with headache. Axial T2-weighted (A)
and coronal FLAIR (B) MR images show no parenchymal abnormalities and
normal flow voids in superior sagittal sinus (arrows, A) and
both lateral sinuses (arrows, B).
|
|
View larger version (131K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B —20-year-old man with headache. Axial T2-weighted (A)
and coronal FLAIR (B) MR images show no parenchymal abnormalities and
normal flow voids in superior sagittal sinus (arrows, A) and
both lateral sinuses (arrows, B).
|
|
View larger version (85K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3D —20-year-old man with headache. Three-dimensional CT venogram
using integral algorithm depicts superior sagittal sinus thrombosis as
intraluminal filling defect (arrow) that extends into right
transverse sinus.
|
|
Results
Of 50 patients with clinically suspected CVT, 30 were diagnosed as having
cerebral sinovenous thrombosis on CT venography or MR venography. These 30
positive cases were included in our study for further evaluation. Patients
were divided into three groups depending on the predominant dural sinus
involvement as noted on MR venography. Twenty-four patients (80%) had
thrombosis of the superficial cortical sinus venous system (group 1). One
(3.3%) patient had isolated deep venous system thrombosis (group 2). Five
(16.7%) patients had generalized thrombosis of both the superficial and the
deep venous systems (group 3).
On unenhanced CT, the cord sign was present in two (6.7%) of 30 cases and
the dense vein sign was identified in nine (30%) of 30 cases
(Fig. 1). Venous infarcts were
present in 22 cases (73.3%), which were hemorrhagic in 16 cases (53.3%)
(Fig. 2A). Unenhanced CT
revealed no abnormality in four (13.3%) of 30 cases, but evidence of CVT was
subsequently seen on CT venography and MR venography.
The most common MR finding, seen in 80% of cases, was the presence on
proton density-, T1-, and T2-weighted images, of a hyperintense signal in the
dural sinuses (replacement of normal flow void), which is suggestive of
intraluminal thrombus (Fig.
2B). Hemorrhagic infarcts were present in 60% (18/30) of patients.
Of eight patients with normal parenchyma on MRI, there was evidence of CVT in
the form of replacement of normal flow void by abnormal signal intensity in
six patients. In two patients, routine MRI sequences were completely normal
but evidence was seen of venous sinus thrombus on CT venography and MR
venography (Fig. 3A,
3B,
3C,
3D).
Table 1 summarizes the
unenhanced CT and MRI findings in the defined groups.
The sinus most frequently involved was the superior sagittal sinus in 20
(66.7%) of 30 patients. Of 20 patients with superior sagittal sinus
thrombosis, 15 also had associated thrombosis of either the right or the left
or both transverse sinuses, and 16 patients had associated thrombosis of
either the right or the left or both sigmoid sinuses as well. Overall, the
involvement of various sinuses (on CT venography/MR venography) was as
follows: superior sagittal sinus (20/19), right transverse sinus (11/12), left
transverse sinus (14/14), right sigmoid sinus (12/9), left sigmoid sinus
(11/10), straight sinus (5/5), Galen's vein (4/4), right internal cerebral
vein (2/2), and left internal cerebral vein (2/2).
Table 2 summarizes the
frequency of thrombosis of major sinuses on CT venography and MR
venography.
View larger version (110K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A —30-year-old man with severe headache. CT venography axial
source image shows complete thrombosis of left transverse sinus and partial
thrombosis of right transverse sinus. Arrows indicate filling defect.
|
|
View larger version (131K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4B —30-year-old man with severe headache. CT venography axial
source image shows complete thrombosis of left sigmoid sinus and partial
thrombosis of right sigmoid sinus. Arrows indicate filling defect.
|
|
View larger version (150K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4C —30-year-old man with severe headache. Two-dimensional fast
low-angle shot MR venogram in axial plane shows isointense signal in left
sigmoid sinus and normal flow signal on right side.
|
|
View larger version (167K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4D —30-year-old man with severe headache. FLAIR MR image in
coronal plane shows hyperintense signal in superior sagittal sinus and left
sigmoid sinus and normal flow void in right sigmoid sinus (arrow). MR
venography missed partial thrombosis of right transverse and sigmoid
sinuses.
|
|
Both CT venography and MR venography established the diagnosis of CVT in 30
of 50 patients. However, discrepancies existed regarding the extent of
thrombosis in individual sinuses in these 30 positive cases. CT venography
showed thrombosis in a total of 81 sinuses as compared with 77 for MR
venography, with CT venography depicting additional evidence of venous
thrombosis in the contralateral transverse and sigmoid sinuses of two patients
(Fig. 4A,
4B,
4C,
4D).
Using MR venography as the gold standard, we calculated the sensitivity,
specificity, negative predictive value (NPV), positive predictive value (PPV),
and accuracy of CT venography. The results are summarized in
Table 3. Using Pearson's
correlation coefficient, we found that a highly significant correlation
existed between CT venography and MR venography findings (r = 0.7913,
p < 0.00001).
View this table:
[in this window]
[in a new window]
|
TABLE 3: Sensitivity, Specificity, Positive Predictive Value (PPV), Negative
Predictive Value (NPV), and Accuracy of CT Venography When Using MR Venography
as Gold Standard
|
|
Discussion
Before MRI and MR venography, CT was the best noninvasive method for
diagnosing CVT [6]. Unenhanced
CT signs of CVT include the presence of hemorrhagic venous infarcts, the cord
sign, the dense vein sign, and the "empty delta" sign on
contrast-enhanced CT
[7-9].
These signs are present in a minority of cases and their absence does not
exclude the diagnosis of CVT. In two of our patients, the only evidence of CVT
on unenhanced CT was the dense vein sign, which was confirmed on CT venography
and MR venography.
MRI is intrinsically sensitive to flow phenomena and can be applied in all
three spatial planes. On routine MRI, patent dural venous sinuses show signal
voids on spin-echo images. Both the intravascular thrombus and its
complications (mass effect, edema, infarction, hemorrhage, and hydrocephalus)
may be identified on MRI. With CVT, the expected signal void is replaced by an
abnormal signal, the specific nature of which depends on the sequence
parameters [10]. In a study by
Yuh et al. [11], MRI findings
of venous sinus occlusive disease were grouped into three patterns: brain
swelling without signal abnormalities on T2-weighted imaging, brain swelling
with signal abnormalities on T2-weighted imaging (interstitial edema) but no
hematoma, and brain swelling with signal changes that are compatible with
hematoma and edema on T2-weighted imaging.
In our study, hemorrhagic infarcts were the most frequent parenchymal
lesions, seen in 60% of cases; nonhemorrhagic infarcts were seen in 13.3% of
cases. The replacement of normal flow void by abnormal signal was seen in 80%
of cases. Of eight patients with normal parenchyma on MRI, there was evidence
of CVT in the form of replacement of normal flow void by abnormal signal
intensity on proton density-, T1-, or T2-weighted images in six cases. In the
remaining two cases, normal flow void was seen on routine MR sequences and no
parenchymal changes were seen, but there was evidence of thrombus on CT
venography and MR venography. This finding emphasizes the importance of
performing MR venography in all suspected cases of CVT even when routine MR
findings are apparently normal. Spin-echo MRI may also be misleading, with a
false diagnosis of dural sinus thrombosis resulting from flow-related
enhancement or even from echo rephrasing, or a false impression of vessel
patency resulting from intracellular deoxyhemoglobin/methemoglobin mimicking a
normal signal void on long-TR sequences (T2-weighted images). Subdural clot
lying along a dural sinus may also be confused with an intraluminal
thrombus.
View larger version (153K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5A —28-year-old postpartum woman with sudden loss of
consciousness. Sagittal maximum-intensity-projection image of CT venography
(A) and corresponding MR venography image (B) show loss of flow
signal in superior sagittal sinus (arrows) and increased flow through
collaterals (arrowheads, A and short arrows,
B). Note progression of thrombus to involve posterior portion of
superior sagittal sinus on MR venography (B), which was performed 24
hours after CT venography (A).
|
|
View larger version (129K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5B —28-year-old postpartum woman with sudden loss of
consciousness. Sagittal maximum-intensity-projection image of CT venography
(A) and corresponding MR venography image (B) show loss of flow
signal in superior sagittal sinus (arrows) and increased flow through
collaterals (arrowheads, A and short arrows,
B). Note progression of thrombus to involve posterior portion of
superior sagittal sinus on MR venography (B), which was performed 24
hours after CT venography (A).
|
|
MR venography is regarded as the best noninvasive method for evaluating the
cerebral venous system [4,
12]. The two most common
techniques are based on time-of-flight effects (2D and 3D time-of-flight
techniques) of moving spins and on motion-induced phase shifts (the 3D
phase-contrast technique). At MR venography, flow gives rise to high signal
intensity, and absence of flow is characterized by reduced signal intensity
[13]. Liauw et al.
[13] conducted a study in 12
healthy volunteers using time-of-flight and phase-contrast techniques in the
transverse, sagittal, and coronal planes. They found that visualization of a
normal intracranial venous system was better with 3D phase-contrast and 2D
time-of-flight MR angiographic techniques in the coronal plane than with
transverse or sagittal 2D time-of-flight MR angiography.
The major technical drawbacks of the time-of-flight technique are
saturation effects occurring at in-plane flow and the inclusion of substances,
such as methemoglobin, with a short T1 relaxation time
[14,
15]. In 3D phase-contrast MR
angiography, the acquisition times are longer (even up to 40 minutes), so the
technique is more susceptible to motion artifacts. A prior estimate of blood
flow velocity is required to avoid aliasing. The phase-contrast technique may
be more sensitive to signal loss because of turbulence or intravoxel dephasing
[15]. In this study, most
patients underwent imaging in an acute setting and were too clinically
unstable to undergo MRI with long acquisition times; hence, the 2D
time-of-flight technique was preferred over the 3D phase-contrast venographic
technique.
With the recent advent of helical CT, 3D vascular imaging has also become
possible. A large volume of tissue can be scanned during peak arterial or
venous enhancement while still maintaining high spatial resolution. CT
venography is a rapid and a useful method that yields detailed images of the
intracranial venous circulation with consistently high quality
[5]. Protocols are available
that allow high-resolution scanning of the intracranial vasculature and
postprocessing of scanned data to create MPRs in any desired plane as well as
3D projection images (MIP algorithm).
View larger version (130K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6 —45-year-old man with new onset of seizures. CT venography
multiplanar reconstructed coronal image shows filling defects in superior
sagittal sinus (arrowhead) and left sigmoid sinus
(arrow).
|
|
View larger version (95K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 7A —18-year-old woman with sudden onset of headache and straight
sinus thrombosis. Three-dimensional CT venogram obtained using integral
algorithm shows filling defect in straight sinus (arrowheads).
|
|
View larger version (97K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8A —25-year-old man who presented with sudden loss of
consciousness. Three-dimensional CT venography images with integral algorithm
show filling defect in posterior part of superior sagittal sinus and normal
anterior part (arrow, A) as well as filling defects in
superior sagittal sinus and left transverse sinus (arrows,
B).
|
|
View larger version (98K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8B —25-year-old man who presented with sudden loss of
consciousness. Three-dimensional CT venography images with integral algorithm
show filling defect in posterior part of superior sagittal sinus and normal
anterior part (arrow, A) as well as filling defects in
superior sagittal sinus and left transverse sinus (arrows,
B).
|
|
View larger version (137K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8C —25-year-old man who presented with sudden loss of
consciousness. Corresponding MR venography maximum-intensity-projection image
shows loss of flow signal in superior sagittal sinus and straight sinus and
increased flow through collaterals.
|
|
On CT venography, a thrombosed dural sinus is seen as a filling defect and
is often associated with contrast enhancement of the walls of the dural sinus
as well as abnormal collateral venous drainage and tentorial enhancement. In a
study by Casey et al. [5], 36
CT venograms were obtained in 33 patients after the IV administration of
iodinated contrast material. The superior sagittal sinus, the straight sinus,
and both transverse sinuses as well as Galen's vein and the internal cerebral
veins were seen in 100% of cases, whereas the cavernous sinus, the inferior
sagittal sinus, and the basal Rosenthal's vein were seen in 95% of cases. CT
venograms were easier to interpret, revealed better small-vessel anatomy, and
had fewer artifacts than MR venography.
In a study by Ozsvath et al.
[16], 24 patients underwent
both CT venography and MR venography of the intracranial venous circulation.
Dural sinus thrombosis was diagnosed in eight of the 17 patients with
suspected CVT using MR venography. In these eight patients, the diagnosis was
also made with CT venography. CT venograms revealed greater small-vessel
conspicuity. The superior sagittal sinus, straight sinus, Galen's vein, and
internal cerebral veins were visualized in all CT and MR venograms. Both the
right and left transverse sinuses were present in all CT venograms. However,
MR venography failed to show one transverse sinus in each of the four
patients. CT venography was also more reliable in revealing the basal
Rosenthal's veins, the thalamostriate veins, and the inferior sagittal sinus.
Those authors concluded that CT venography is superior to MR venography in
identifying cerebral veins and dural sinuses and is at least equivalent in
diagnosing dural sinus thrombosis.
In a study by Wetzel et al.
[17], 25 patients underwent
both intraarterial digital subtraction angiography and CT venography to
compare the reliability of the two techniques in imaging cerebral vein anatomy
and pathology. When digital subtraction angiography was used as the standard
of reference, MPR images had an overall sensitivity of 95% (specificity, 19%)
and MIP images, a sensitivity of 80% (specificity, 44%), in depicting the
cerebral venous anatomy. On the basis of an interobserver consensus including
digital subtraction angiography, MPR images, and MIP images, the sensitivity
and specificity were 95% and 91% for MPR, 90% and 100% for digital subtraction
angiography, and 79% and 92% for MIP images. MPR images were superior to those
of digital subtraction angiography in showing the cavernous sinus, the
inferior sagittal sinus, and the basal Rosenthal's vein. Venous occlusive
diseases were correctly recognized on both MPR and MIP images. Those authors
concluded that CT venography is a reliable method for depicting the cerebral
venous structures.
In the acute setting, unenhanced CT, which is usually the initial
technique, can be immediately followed by venography (actual scanning time of
1 minute), thus saving time to diagnosis and treatment
[5]. CT is less impaired by
motion artifacts because of a rapid acquisition time, which is of paramount
importance in sick and uncooperative patients after a seizure. CT venography
shows the contrast material in vessels and directly depicts sinus thrombosis
as a filling defect on 3D MIP images, similar to filling defects that are
known to be shown on conventional cerebral angiography
[1]. CT venography more
frequently depicts sinuses or smaller cerebral veins with low flow than MR
venography does [16]. CT
venography is not affected by flow-related artifacts that affect MR
venography. In situations in which MRI has absolute or relative
contraindications, such as ferromagnetic implants or cardiac pacemakers, CT
venography is the technique of choice.
CT venography has disadvantages, such as significant exposure to ionizing
radiation and the need for IV contrast material
[13]. CT also cannot be used
as a screening technique in pregnant patients or for repeated follow-up
[4].
Venograms are usually displayed using an MIP algorithm that emphasizes the
brightest voxels in a vessel at the expense of less-bright voxels. Thrombosis
may thus be obscured by the surrounding contrast-enhanced blood on CT
venography and by high-flow signal on MR venography
[16]. However, CT venograms
can also be displayed using an integral algorithm that depicts the average
intensity of the first five voxels deep in relation to the model surface that
is nearest the viewer, allowing direct visualization of the thrombus in the
sinuses [5].
In our study we have compared CT venography with MR venography to determine
the reliability of CT venography in diagnosing CVT, using MR venography as the
gold standard. Both CT venography and MR venography diagnosed CVT in 30 of 50
patients in our study group, and a statistically significant correlation was
seen between the two techniques (r = 0.7913, p < 0.00001,
Pearson's correlation coefficient). Venograms were displayed using the MIP
algorithm for both CT venography and MR venography, which gave an overall view
of the status of the venous sinuses and the presence of collaterals (Fig.
5A,
5B). For CT venography, we
also used the MPR (Fig. 6) and
integral algorithms, which directly displayed the intraluminal thrombus,
confirming the diagnosis of CVT (Figs.
7A,
8A, and
8B). When MR venography was
used as the gold standard, CT venography was found to have both a sensitivity
and a specificity of 75-100%, depending on the sinus or vein involved.
The limitations of our study were that CT venography and unenhanced CT were
performed first and were later followed by MR venography. An 8- to 24-hour gap
occurred between the two techniques, which may have affected the results
because of progression or resolution of the thrombus in this interval. Second,
contrast material was not used for MR venography because of cost constraints.
We do recommend further studies comparing CT venography with contrast-enhanced
MR venography. In addition, we used only the 2D time-of-flight technique for
MR venography, which has its limitations when compared with the 3D
phase-contrast technique.
In conclusion, CT venography is as accurate as MR venography in diagnosing
CVT. CT venography can be used as a reliable alternative to MR venography in
evaluating patients with clinically suspected CVT, especially in the acute
setting. An integral algorithm is a useful display tool, especially for the
direct visualization of an intraluminal thrombus.
References
- Krayenbuhl H. Cerebral venous thrombosis: the diagnostic value of
cerebral angiography. Schweiz Arch Neurol Psychiatr1954; 74:261
-287[Medline]
- Ameri A, Bousser MG. Cerebral venous thrombosis. Neurol
Clin 1992; 10:87
-111[Medline]
- Thron A, Wessel K, Linden D. Superior sagittal sinus thrombosis:
neuroradiological evaluation and clinical findings. J
Neurol 1986; 233:283
-288[CrossRef][Medline]
- Vogl TJ, Bergman C, Villringer A, Einhaupl K, Lissner J, Felix R.
Dural sinus thrombosis: value of venous MR angiography for diagnosis and
follow-up. AJR 1994;162
: 1191-1198[Abstract/Free Full Text]
- Casey SO, Alberico RA, Patel M, et al. Cerebral CT venography.
Radiology 1996;198
: 163-170[Abstract/Free Full Text]
- Goldberg AL, Rosenbaum AE, Wang H, et al. Computed tomography of
dural sinus thrombosis. J Comput Assist Tomogr1986; 10:16
-20[Medline]
- Buonanno FS, Mody DM, Ball MR, et al. Computed cranial tomographic
findings in cerebral sinovenous occlusion. J Comput Assist
Tomogr 1978; 2:281
-290[Medline]
- Ford K, Sarwar M. Computed tomography of dural sinus thrombosis.
Am J Neuroradiol 1981;2
: 539-543[Abstract]
- Osborn AG, Anderson RE, Wing SD. The false falx sign.
Radiology 1980;134
: 421-425[Abstract/Free Full Text]
- Cure JK, Van Tassel P, Smith HT. Normal and variant anatomy of the
dural venous sinuses. Semin Ultrasound CT MR1994; 15:499
-519[CrossRef][Medline]
- Yuh WT, Simonson TM, Wang AM, et al. Venous sinus occlusive
disease: MR findings. Am J Neuroradiol1994; 15:309
-316[Abstract]
- Mattle HP, Wentz KU, Edelman RR, et al. Cerebral venography with
MR. Radiology 1991;178
: 453-458[Abstract/Free Full Text]
- Liauw L, van Buchem MA, Spilt A, et al. MR angiography of the
intracranial venous system. Radiology2000; 214:678
-682[Abstract/Free Full Text]
- Hoston J III, Ehman RL. Comparison of time-of-flight and
phase-contrast MR neuroangiographic techniques.
RadioGraphics 1993;13
: 5-19[Abstract]
- Ayanzen RH, Bird CR, Keller PS. Cerebral MR venography: normal
anatomy and potential pitfalls. Am J Neuroradiol2000; 21:74
-78[Abstract/Free Full Text]
- Ozsvath RR, Casey SO, Lustrin ES, Alberico RA, Hassankhani A, Patel
M. Cerebral venography: comparison of CT and MR projection venography.
AJR 1997; 169:1699
-1707[Abstract/Free Full Text]
- Wetzel SG, Kirsch E, Stock KW, Kolbe M, Kaim A, Radue EW. Cerebral
veins: comparative study of CT venography with intraarterial digital
subtraction angiography. Am J Neuroradiol1999; 20:249
-255[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
R Smith and M D Hourihan
Investigating suspected cerebral venous thrombosis
BMJ,
April 14, 2007;
334(7597):
794 - 795.
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
