December 2006, VOLUME 187
NUMBER 6

Recommend & Share

December 2006, Volume 187, Number 6

Neuroradiology

Original Research

Comparison of CT Venography with MR Venography in Cerebral Sinovenous Thrombosis

+ Affiliations:
1Department of Radiodiagnosis and Imaging, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh 160012, India.

2Present address: Kesri Bhavan, 2/1/1B Munshi Bazar Rd., Kolkata 700015, India.

3Department of Neurology, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India.

Citation: American Journal of Roentgenology. 2006;187: 1637-1643. 10.2214/AJR.05.1249

ABSTRACT
Next section

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
Previous sectionNext section

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.

figure
View larger version (194K)
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
Previous sectionNext section

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 × 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.

figure
View larger version (187K)
Fig. 2A 45-year-old woman with headache and altered sensorium. Unenhanced CT scan of head in axial plane shows hemorrhagic infarcts in bilateral parietooccipital lobes.

figure
View larger version (143K)
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).

figure
View larger version (156K)
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).

figure
View larger version (166K)
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).

figure
View larger version (147K)
Fig. 3C 20-year-old man with headache. Sagittal maximum-intensity-projection MR venogram shows loss of flow signal in superior sagittal sinus (arrow).

figure
View larger version (112K)
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
Previous sectionNext section

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.

TABLE 1: Unenhanced CT and MRI Findings in 30 Patients with Cerebral Sinovenous Thrombosis

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.

TABLE 2: Frequency of Thrombosis of Major Sinuses on CT Venography and MR Venography in 30 Patients

figure
View larger version (147K)
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.

figure
View larger version (169K)
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.

figure
View larger version (188K)
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.

figure
View larger version (214K)
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).

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
Previous sectionNext section

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.

figure
View larger version (195K)
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).

figure
View larger version (169K)
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).

figure
View larger version (166K)
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).

figure
View larger version (123K)
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).

figure
View larger version (175K)
Fig. 7B 18-year-old woman with sudden onset of headache and straight sinus thrombosis. Corresponding maximum-intensity-projection MR image shows loss of signal in straight sinus (arrow).

figure
View larger version (125K)
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).

figure
View larger version (123K)
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).

figure
View larger version (173K)
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.

Address correspondence to N. Khandelwal ().

References
Previous sectionNext section
1. Krayenbuhl H. Cerebral venous thrombosis: the diagnostic value of cerebral angiography. Schweiz Arch Neurol Psychiatr 1954; 74:261-287 [Google Scholar]
2. Ameri A, Bousser MG. Cerebral venous thrombosis. Neurol Clin 1992; 10:87-111 [Google Scholar]
3. Thron A, Wessel K, Linden D. Superior sagittal sinus thrombosis: neuroradiological evaluation and clinical findings. J Neurol 1986; 233:283-288 [Google Scholar]
4. 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] [Google Scholar]
5. Casey SO, Alberico RA, Patel M, et al. Cerebral CT venography. Radiology 1996; 198:163-170 [Google Scholar]
6. Goldberg AL, Rosenbaum AE, Wang H, et al. Computed tomography of dural sinus thrombosis. J Comput Assist Tomogr 1986; 10:16-20 [Google Scholar]
7. Buonanno FS, Mody DM, Ball MR, et al. Computed cranial tomographic findings in cerebral sinovenous occlusion. J Comput Assist Tomogr 1978; 2:281-290 [Google Scholar]
8. Ford K, Sarwar M. Computed tomography of dural sinus thrombosis. Am J Neuroradiol 1981; 2:539-543 [Google Scholar]
9. Osborn AG, Anderson RE, Wing SD. The false falx sign. Radiology 1980; 134:421-425 [Google Scholar]
10. Cure JK, Van Tassel P, Smith HT. Normal and variant anatomy of the dural venous sinuses. Semin Ultrasound CT MR 1994; 15:499-519 [Google Scholar]
11. Yuh WT, Simonson TM, Wang AM, et al. Venous sinus occlusive disease: MR findings. Am J Neuroradiol 1994; 15:309-316 [Google Scholar]
12. Mattle HP, Wentz KU, Edelman RR, et al. Cerebral venography with MR. Radiology 1991; 178:453-458 [Google Scholar]
13. Liauw L, van Buchem MA, Spilt A, et al. MR angiography of the intracranial venous system. Radiology 2000; 214:678-682 [Google Scholar]
14. Hoston J III, Ehman RL. Comparison of time-of-flight and phase-contrast MR neuroangiographic techniques. RadioGraphics 1993; 13:5-19 [Google Scholar]
15. Ayanzen RH, Bird CR, Keller PS. Cerebral MR venography: normal anatomy and potential pitfalls. Am J Neuroradiol 2000; 21:74-78 [Google Scholar]
16. 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] [Google Scholar]
17. 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 Neuroradiol 1999; 20:249-255 [Google Scholar]

Recommended Articles

Comparison of CT Venography with MR Venography in Cerebral Sinovenous Thrombosis

Full Access, , , ,
American Journal of Roentgenology. 2007;189:S64-S75. 10.2214/AJR.07.7015
Abstract | Full Text | PDF (6409 KB) | PDF Plus (2747 KB) 
Full Access, , , , ,
American Journal of Roentgenology. 1997;169:1699-1707. 10.2214/ajr.169.6.9393193
Abstract | PDF (1705 KB) | PDF Plus (929 KB) 
Full Access, , , ,
American Journal of Roentgenology. 2015;205:1269-1275. 10.2214/AJR.14.14200
Abstract | Full Text | PDF (701 KB) | PDF Plus (715 KB) 
Full Access,
American Journal of Roentgenology. 2011;197:1309-1321. 10.2214/AJR.11.7420
Abstract | Full Text | PDF (1195 KB) | PDF Plus (1215 KB) 
Full Access,
American Journal of Roentgenology. 2011;196:23-31. 10.2214/AJR.10.5323
Abstract | Full Text | PDF (1131 KB) | PDF Plus (1102 KB) | Erratum 1 | Erratum 2 
Full Access, , , , ,
American Journal of Roentgenology. 2014;202:W169-W175. 10.2214/AJR.13.10985
Abstract | Full Text | PDF (1084 KB) | PDF Plus (1003 KB)