| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Neuroradiology |
1 From the Departments of Neuroradiology (K.K., O.J., K.S.) and Radiation Oncology (T.W.), University of Heidelberg Medical School, Im Neuenheimer Feld 400, 69120 Heidelberg, Germany. Received June 8, 2001; revision requested July 5; final revision received March 4, 2002; accepted March 25. Address correspondence to K.K. (e-mail: klaus_kirchhof@med.uni-heidelberg.de).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: The authors prospectively compared the two-dimensional source images, multiplanar reconstructed images, and maximum intensity projection angiograms obtained at contrast-enhanced 3D radio-frequency–spoiled FLASH MR angiography in 20 patients with those obtained at TOF and phase-contrast MR angiographic examinations. Two neuroradiologists in consensus determined the number of visualized cortical veins and graded the quality of visualization of veins and sinuses as intense and continuous, faint and continuous, or noncontinuous. Statistical analysis was performed with the nonparametric sign test and the Wilcoxon matched pairs sign rank test.
RESULTS: The cortical veins, inferior sagittal sinus, and cavernous sinuses were visualized best with FLASH MR angiography (P < .003). The Trolard and Labbé veins were visualized equally well with the FLASH and TOF sequences. For septal, internal cerebral, and Rosenthal left basal vein visualization, phase-contrast MR angiography was inferior to the FLASH and TOF MR angiographic examinations (P < .05). The quality of visualization of the thalamostriate and Galen veins and of the superior sagittal, rectal, and transverse sinuses was the same at all MR angiographic examinations.
CONCLUSION: Three-dimensional FLASH MR angiography depicts some venous structures better than do TOF and phase-contrast MR angiographic examinations. The depiction of other veins is the same with 3D FLASH and TOF sequences.
© RSNA, 2002
Index terms: Cerebral blood vessels • Cerebral blood vessels, MR, 176.121412, 176.121416, 176.12142, 176.12143 • Magnetic resonance (MR), comparative studies, 176.121412, 176.121416, 176.12142, 176.12143 • Magnetic resonance (MR), vascular studies, 176.12142
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Thrombosis of the cerebral veins and dural sinuses usually evolves slowly such that a venous collateral circulation that prevents venous edema and infarction often develops (3–5). However, additional or isolated thrombosis of the subependymal, cortical, or bridging veins usually results in a severe disturbance of the venous circulation with subsequent development of venous infarction and hemorrhage (3,4,6,7). Therefore, the diagnosis of subependymal, cortical, or bridging venous occlusion—either alone or as a complication of dural sinus thrombosis—is essential to the prevention of venous infarction in patients with cerebral venous thrombosis (8–10).
The results of a study performed by Tsai et al (11) showed that the site of parenchymal change does not necessarily correspond to the site of venous thrombosis, despite the close correlation between collateral venous blood flow and brain edema or hemorrhage. Therefore, it is necessary to determine the site and extent of venous thrombosis with magnetic resonance (MR) angiography.
Time-of-flight (TOF) MR angiography performed with either two- or three-dimensional (3D) gradient-echo sequences has been shown to be suitable for delineation of the cerebral arteries, but the lack of inflow of unsaturated protons and the progressive signal loss caused by slow flow impair visualization of the subependymal, cortical, and bridging veins (12–17). Phase-contrast MR angiography has capability for a limited range of velocities and is further limited by gradient imperfection and eddy currents (18). Additionally, all conventional MR angiographic techniques (including TOF and phase-contrast studies) involve long acquisition times, and, therefore, they often cannot be used because patients with cerebral venous thrombosis commonly have impaired consciousness. To improve the usefulness of cerebral venous MR angiography by using very short acquisition times, we decided to abandon the flow-sensitive imaging techniques and generate venous angiograms by using 3D radio-frequency–spoiled gradient-echo fast low-angle shot (FLASH) MR sequences after intravenous bolus injection of gadopentetate dimeglumine (Magnevist; Schering, Berlin, Germany). The purpose of our study was to compare the visualization quality at contrast material–enhanced 3D FLASH venous MR angiography with that at TOF and phase-contrast MR angiographic examinations.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Imaging Examinations
MR angiography was performed by using a 1.5-T whole-body MR system (Edge; Philips Medizin Systeme, Hamburg, Germany) with 27-mT/m gradient capability and a circular transmit-receive polarized head coil. We performed contrast-enhanced single-volume 3D radio-frequency–spoiled gradient-echo FLASH, nonenhanced TOF, and phase-contrast MR angiographic examinations in each patient for an intraindividual comparison of the three techniques.
Three-dimensional radio-frequency–spoiled gradient-echo FLASH MR angiography.—The following imaging parameters were used: 12.0/3.8 (repetition time msec/echo time msec), one signal acquired, 60° flip angle, 110 x 200-pixel matrix, and 260-mm field of view. In addition, a 130-mm-thick slab with 52 partitions and a sagittal section orientation were used to generate an effective section thickness of 2.5 mm. The total examination time was 57 seconds. A 270° presaturation pulse was applied before each excitation to suppress the arterial signal. The delay between manual bolus injection and data acquisition was 20 seconds. Gadopentetate dimeglumine was administered at a dose of 0.05 mmoL per kilogram of body weight.
TOF MR angiography.—The imaging parameters were as follows: 30/7, one signal acquired, 30° flip angle, 192 x 256-pixel matrix, 220-mm field of view, and total examination time of 8.5 minutes. This protocol yielded 112 coronal sections with a thickness of 1.5 mm.
Phase-contrast MR angiography.—The imaging parameters were as follows: 93.0/13.5, one signal acquired, 30° flip angle, 192 x 256-pixel matrix, and 220-mm field of view. In all, 17 sagittal sections with a thickness of 3 mm were acquired. A 120° presaturation pulse was applied before each excitation to suppress the arterial signal. The total examination time was 5 minutes.
Image Analysis
The 3D MR angiograms were generated by using a maximum intensity projection algorithm and rotated around a vertical axis. Multiplanar reconstructed images were obtained in the transverse and sagittal planes. The two-dimensional source MR image, multiplanar reconstructed images, and 3D maximum intensity projection angiograms were interpreted prospectively by two experienced neuroradiologists (K.K., T.W.). The quality of visualization of the veins was graded as follows: intense and continuous, faint and continuous, or noncontinuous. The grades assigned were the result of a consensus opinion following discussion among the two observers. We also compared the images of the cortical veins obtained with the three MR angiographic techniques by determining the numbers of cortical veins and their first-order branches (ie, tributaries) in both hemispheres that were seen on sagittal images. We followed the courses of the vessels continuously to avoid mistaking the proximal and distal aspects of vessels for distinct branches.
Statistical Analysis
For statistical analysis we used the nonparametric sign test and the Wilcoxon matched pairs sign rank test. Differences in the number of cortical veins visualized with the three MR angiographic techniques were analyzed by using the Wilcoxon matched pairs sign rank test. The nonparametric sign test was used to analyze differences in the quality of visualization of the cerebral veins and dural sinuses. If paired veins were not depicted equally on both sides, the right and left vessels were evaluated separately.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
The contrast-enhanced 3D radio-frequency–spoiled FLASH sequence used in this study is based not on the flow of unsaturated spins but rather on the maximal shortening of the T1 value of blood immediately after bolus injection of gadopentetate dimeglumine. Therefore, 3D radio-frequency–spoiled FLASH is not hindered by the signal loss that is caused by progressive spin saturation and that limits the applicability of phase-contrast and TOF sequences at venous MR angiography, and it provides sufficient contrast of small subependymal and cortical veins.
As with contrast-enhanced TOF MR angiography, with 3D FLASH MR angiography, a bolus injection of gadopentetate dimeglumine is required to maintain a high plasma concentration during the acquisition and to avoid enhancement of acute or subacute thrombus (12,13,19). We therefore used a short delay of 20 seconds between manual bolus injection and data acquisition. Although constant infusion after bolus injection has been recommended (13), we applied only a single bolus because of the negligible effect of a constant infusion with imaging times as short as those that we used. Another advantage of less contrast is the decrease in the arteriovenous overlap that makes it more difficult to distinguish the small veins; however, according to Yano et al (13), this problem can be overcome with target and subvolume maximum intensity projection imaging.
The advantages of bolus contrast material injection are combined with those of FLASH MR imaging to result in a very short examination time of 57 seconds. As a result, patients with impaired consciousness also can be examined. To minimize the number of sections required to image the entire brain and thus minimize the imaging time, we chose a sagittal section orientation. However, if thrombosis of the septal veins, Rosenthal basal vein, or Labbé vein is suspected, a coronal section orientation should be chosen to improve visualization of these vessels on multiplanar images.
One unfavorable effect of using a thick slab that covers the entire brain and a short examination time is the limited spatial resolution of 2.5 x 2.4 x 1.3 mm, which is lower than the 1.5 x 1.1 x 0.9-mm spatial resolution with TOF imaging and the 3.0 x 1.1 x 0.9-mm spatial resolution with phase-contrast imaging and might cause subtle irregularities or partial filling defects to be obscured. Therefore, it is preferable to perform 3D FLASH MR angiography if short examination times are essential; however, conscious patients may benefit from the higher spatial resolution techniques. If either TOF or phase-contrast MR angiography fails to depict suspected thrombosis of the cortical veins, however, 3D FLASH MR angiography should be performed despite the lower spatial resolution.
One disadvantage with contrast-enhanced MR angiography is that meningeal enhancement obscures superficial vessels on 3D maximum intensity projection venous angiograms. Another drawback is that enhancement of thrombotic material makes reading subsequent MR angiograms difficult. Therefore, contrast-enhanced MR angiography should be performed only in addition to and after TOF MR angiography. Because recanalization of chronic thrombosis results in an almost instant uptake of contrast material, none of the contrast-enhanced MR angiographic techniques should be used in such cases.
The value of our statistical analysis was limited by the small study population. However, we did not include more patients because this was a pilot study and the patients underwent an additional 15 minutes of imaging.
For MR venous angiography, two-dimensional or 3D TOF sequences are used most often. Three-dimensional TOF sequences provide high spatial resolution and a high signal-to-noise ratio, but they have the disadvantages of inflowing spin saturation and intravoxel spin-phase dispersion (14). Long repetition times, low flip angles, multiple overlapping thin-slab acquisition (MOTSA), a section orientation perpendicular to blood flow, and contrast materials that reduce the T1 value of blood help to reduce spin saturation (14). However, some saturation remains owing to the low venous velocities that diminish the quality of the angiogram (17). MOTSA not only reduces spin saturation, but it also involves the use of small voxels and thus minimizes intravoxel phase dispersion (14). Compared with single-volume 3D sequence acquisition, MOTSA involves longer examination times because the volumes must overlap so that volume aliasing and wraparound artifacts can be avoided (14). Despite the slab boundary and motion artifacts that are produced, the image quality achieved with MOTSA is superior to that achieved with 3D single-volume TOF angiography, predominantly because of differences in the visualization of third-order, fourth-order, and other small vessels (14,20). Magnetization transfer contrast pulses may result in venous suppression and therefore do not improve TOF venous angiograms (21).
The relative thickness of two-dimensional TOF image sections exceeds that of 3D image sections and thus results in decreased spatial resolution and increased intravoxel phase dispersion (17). As mentioned earlier, spin saturation cannot be eliminated entirely. However, with two-dimensional TOF sequences, sequentially acquired sections are superior to sections acquired with MOTSA sequences for delineation of the subependymal veins and dural sinuses (17,22). Whether this also applies to the cortical veins has not, to our knowledge, been tested yet.
If contrast materials are used to reduce the spin saturation with two-dimensional and 3D TOF sequences, they should be administered as a bolus to avoid enhancement of any nonpatent vessels (12,13). Dormont et al (19) observed enhancement of thrombi in all of their patients with chronic sinus thrombosis, and they attributed this enhancement to thrombus organization. The dose range of 5–10 mL of gadopentetate dimeglumine appears to be optimal for contrast-enhanced cerebral MR angiography, which is superior to nonenhanced techniques with regard to imaging of small vessels, slow flow, aneurysms, arteriovenous malformations, and venous angiomas (12,13,15). However, the quality of two-dimensional TOF MR angiograms can be improved without contrast material enhancement. Lewin et al (23) introduced a sequential oblique section technique that helps to increase the magnitude of unsaturated inflow and reduce the examination time and thus possibly avoid the need for contrast material injection. According to their results, an oblique acquisition plane with an angle of 15°–20° toward the vessel of interest enables optimal visualization of small cerebral veins.
The usefulness of phase-contrast angiography is limited by gradient imperfections, eddy currents, and aliasing artifacts that occur when flow velocities exceed the expected values (18). In the case of developmental asymmetry in particular, a wide range of velocities can be seen within dural sinuses. Long imaging times are a major drawback of phase-contrast angiography, especially 3D phase-contrast angiography, in which phase encoding in three axes is required. Phase-contrast angiography can yield additional information about blood flow directions and absolute flow velocities. It should be used in combination with spin-echo or TOF sequences, because hyperintense thrombi or paradoxical enhancement can mimic flow with gradient-echo sequences.
Contrast-enhanced magnetization-prepared rapid gradient-echo sequences also have been used with MR venous angiography. These sequences provide heavily T1-weighted contrast and high spatial resolution, but they involve long examination times—of about 7 minutes. Depiction of the normal venous anatomy and cerebral venous disease with contrast-enhanced magnetization-prepared rapid gradient-echo sequences has been better than that with two-dimensional TOF MR venous angiography and comparable or even superior to that with digital subtraction angiography (24). Contrast-enhanced magnetization-prepared rapid gradient-echo sequences should not be used in cases of chronic thrombosis, which is difficult to distinguish from normal enhancement of a patent vein or sinus. Furthermore, intravenous fibrotic bands or septa and pacchionian granulations can be misdiagnosed as thrombosis.
In summary, our study results show that 3D FLASH MR angiography after bolus injection of gadopentetate dimeglumine enables better visualization of the cortical veins and their first-order branches (tributaries) than do TOF MR angiography and phase-contrast MR angiography. The quality of visualization of the superior sagittal sinus, Galen vein, straight sinus, and transverse sinuses was the same with all three MR angiographic techniques.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, K.S., K.K.; study concepts and design, O.J., K.K.; literature research, T.W.; clinical studies, K.K.; data acquisition, K.K.; data analysis/interpretation, T.W., K.K.; statistical analysis, T.W.; manuscript preparation, K.K.; manuscript definition of intellectual content, all authors; manuscript editing, K.S., T.W., K.K.; manuscript revision/review, all authors; manuscript final version approval, K.S.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Tomasian, N. Salamon, M.S. Krishnam, J.P. Finn, and J.P. Villablanca 3D High-Spatial-Resolution Cerebral MR Venography at 3T: A Contrast-Dose-Reduction Study AJNR Am. J. Neuroradiol., February 1, 2009; 30(2): 349 - 355. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Lettau, K. Sartor, S. Heiland, and S. Hahnel 3T High-Spatial-Resolution Contrast-Enhanced MR Angiography of the Intracranial Venous System with Parallel Imaging AJNR Am. J. Neuroradiol., January 1, 2009; 30(1): 185 - 187. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. H. Hu, N. G. Campeau, J. Huston III, D. G. Kruger, C. R. Haider, and S. J. Riederer High-Spatial-Resolution Contrast-enhanced MR Angiography of the Intracranial Venous System with Fourfold Accelerated Two-dimensional Sensitivity Encoding Radiology, June 1, 2007; 243(3): 853 - 861. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Linn, B. Ertl-Wagner, K.C. Seelos, M. Strupp, M. Reiser, H. Bruckmann, and R. Bruning Diagnostic Value of Multidetector-Row CT Angiography in the Evaluation of Thrombosis of the Cerebral Venous Sinuses AJNR Am. J. Neuroradiol., May 1, 2007; 28(5): 946 - 952. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. G. Wustenberg, C. Offergeld, T. Zahnert, K.-B. Huttenbrink, and T. Kittner Extension of Intracranial Thrombosis After Unilateral Dissection of the Internal Jugular Vein Arch Otolaryngol Head Neck Surg, May 1, 2005; 131(5): 430 - 433. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. P. Mermuys, P. K. Vanhoenacker, P. Chappel, and L. Van Hoe Three-dimensional Venography of the Brain with a Volumetric Interpolated Sequence Radiology, March 1, 2005; 234(3): 901 - 908. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
