(Radiographics. 2000;20:135-152.)
© RSNA, 2000
Dynamic Subtraction Contrast-enhanced MR Angiography: Technique, Clinical Applications, and Pitfalls1
Yuji Watanabe, MD, PhD ,
Masako Dohke, MD ,
Akira Okumura, MD ,
Yoshiki Amoh, MD,
Takayoshi Ishimori, MD ,
Kazushige Oda, MD ,
Takafumi Hayashi, MD,
Atsuto Hiyama, MD and
Yoshihiro Dodo, MD, PhD
1 From the Department of Radiology, Kurashiki Central Hospital, 1-1-1 Miwa, Kurashiki 710-8602, Japan. Presented as a scientific exhibit at the 1998 RSNA scientific assembly. Received March 8, 1999; revision requested April 20 and received May 17; accepted May 18. Address reprint requests to Y.W.
 |
Abstract
|
|---|
Rapid advances in techniques of contrast material–enhanced magnetic resonance (MR) angiography have enabled evaluation of the entire aorta and the main arteries. Dynamic subtraction MR angiography consists of first-pass imaging of long segments of arteries by using a three-dimensional fast field echo sequence with multiple rapid bolus injections of a small dose of gadopentetate dimeglumine. Subtraction enables clear demonstration of the enhanced vascular lumen by eliminating background signal. Improved temporal resolution and repeated sequences after gadopentetate dimeglumine administration allow demonstration of arteries and veins separately. Double subtraction postprocessing can be used to eliminate arterial enhancement in demonstration of the portal and systemic veins. Additional postprocessing can be used to demonstrate arteries in a single image in patients with aortic dissection or a prolonged circulation time. To optimize the examination, the pulse sequence, injection dose, injection rate, timing of the start of data acquisition, imaging time, breath holding, section thickness, and coil selection should be considered. This technique is flexible enough to be applied in a variety of clinical settings, including atherosclerotic occlusive disease, aneurysm of aortoiliac arteries, bypass graft, Takayasu arteritis, aortic dissection, antiphospholipid antibody syndrome, renal artery disease, pelvic vascular disease, and the portomesenteric venous system.
Index Terms: Arteries, MR, 56.12142, 9*2.12942 • Blood vessels, MR, 56.12142, 9*2
.12942 • Magnetic resonance (MR), vascular studies, 56.12142, 9*2
.12942 • Veins, MR, 56.12142, 9*2
.12942
|
Introduction
|
|---|
Symptomatic patients with vascular disease often require precise evaluation of long segments of the main arteries. A variety of techniques have been used to perform magnetic resonance (MR) angiography for this purpose (1–29). Conventional time-of-flight MR angiography and phase-contrast MR angiography have been limited by motion artifacts and the need to demonstrate long segments of arteries in a short imaging time (1,2). Contrast material–enhanced MR angiography has been developed as a noninvasive and useful fast MR angiography technique to avoid the limitations of conventional time-of-flight and phase-contrast MR angiography (3,4). Administration of a paramagnetic contrast agent shortens the T1 of blood. Acquisition of data during the first pass of the contrast agent with an appropriate pulse sequence enables demonstration of arteries before the contrast agent reaches veins. This method has several advantages over conventional digital subtraction angiography (5): (a) no need for sedation or catheterization, (b) fewer complications, (c) fewer contrast agent–induced nephrotoxic effects, and (d) capability for production of multiangular reprojection images.
A variety of techniques of contrast-enhanced MR angiography, including various injection methods, pulse sequences, methods of data acquisition, and postprocessing methods, have been developed (3–29). The technical elements of dynamic subtraction MR angiography are rapid sequential volume acquisitions after multiple low-dose injections of gadopentetate dimeglumine with postacquisition subtraction and maximum intensity projection reconstruction. When performed properly, dynamic subtraction MR angiography provides high-quality MR angiograms in various clinical settings. The goal of our method is to demonstrate long segments of the main arteries with a small dose of gadopentetate dimeglumine.
In this article, we describe and illustrate the basic principle, technique, clinical applications, and diagnostic pitfalls of dynamic subtraction MR angiography.
|
Basic Principle
|
|---|
Dynamic subtraction MR angiography provides a picture of the vascular lumen by tracking a bolus of paramagnetic contrast agent (gadopentetate dimeglumine), which is administered intravenously. This method is based on the T1 shortening of blood associated with intravenously administered gadopentetate dimeglumine (3–7). Consequently, the vascular lumen, especially the arterial lumen, demonstrates high signal intensity relative to that of solid organs and other tissues on T1-weighted gradient-echo (GRE) images. Use of subtraction enables clear demonstration of the enhanced vascular lumen by eliminating background signal (Fig 1).
View larger version (39K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1. Normal appearance in a healthy volunteer. Dynamic subtraction MR angiogram obtained with three separate injections of gadopentetate dimeglumine shows the lower body. The injection doses were 0.04, 0.03, and 0.03 mmol/kg for imaging the abdomen, thighs, and calves, respectively. (Reprinted, with permission, from reference 10.)
|
|
In dynamic subtraction MR angiography, fast arterial imaging is achieved by using a three-dimensional (3D) spoiled T1-weighted GRE sequence with a short echo time and acquiring data during the first arterial pass of a rapidly administered bolus of gadopentetate dimeglumine (8–10). Our basic method of dynamic subtraction MR angiography is shown in Figure 2 (10). By repeating the sequence after gadopentetate dimeglumine injection, the gadopentetate dimeglumine bolus can be tracked, which helps demonstrate circulation dynamics (Fig 3).
View larger version (133K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3a. Abdominal aortic coarctation. Dynamic subtraction MR angiograms from four of five phases acquired show vascular enhancement by tracking a 0.04-mmol/kg bolus of gadopentetate dimeglumine. In the arterial phase (a), abdominal aortic coarctation (arrow) and the renal arteries are well visualized with minimal venous enhancement. The portal veins (arrowheads in b), hepatic veins (arrows in c), and inferior vena cava (arrowheads in d) are demonstrated in the different phases.
|
|
View larger version (134K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3b. Abdominal aortic coarctation. Dynamic subtraction MR angiograms from four of five phases acquired show vascular enhancement by tracking a 0.04-mmol/kg bolus of gadopentetate dimeglumine. In the arterial phase (a), abdominal aortic coarctation (arrow) and the renal arteries are well visualized with minimal venous enhancement. The portal veins (arrowheads in b), hepatic veins (arrows in c), and inferior vena cava (arrowheads in d) are demonstrated in the different phases.
|
|
View larger version (138K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3c. Abdominal aortic coarctation. Dynamic subtraction MR angiograms from four of five phases acquired show vascular enhancement by tracking a 0.04-mmol/kg bolus of gadopentetate dimeglumine. In the arterial phase (a), abdominal aortic coarctation (arrow) and the renal arteries are well visualized with minimal venous enhancement. The portal veins (arrowheads in b), hepatic veins (arrows in c), and inferior vena cava (arrowheads in d) are demonstrated in the different phases.
|
|
View larger version (135K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3d. Abdominal aortic coarctation. Dynamic subtraction MR angiograms from four of five phases acquired show vascular enhancement by tracking a 0.04-mmol/kg bolus of gadopentetate dimeglumine. In the arterial phase (a), abdominal aortic coarctation (arrow) and the renal arteries are well visualized with minimal venous enhancement. The portal veins (arrowheads in b), hepatic veins (arrows in c), and inferior vena cava (arrowheads in d) are demonstrated in the different phases.
|
|
|
Technique
|
|---|
We perform all dynamic subtraction MR angiography examinations with a 1.5-T MR imager (Gyroscan ACS-NT; Philips Medical Systems, Best, The Netherlands) with 15-mT/m gradients. Three-dimensional fast field echo and 3D turbo field echo sequences are used in dynamic subtraction MR angiography. Imaging parameters are shown in the Table.
Each patient is positioned supine on the patient table, which is equipped with a spine synergy coil so that both calves can be imaged with the spine syneregy coil and other anatomic areas can be imaged with a body coil. A peripheral intravenous line with a 19-gauge needle is placed into subcutaneous veins of the forearm or antecubital fossa. Before and after a rapid bolus intravenous injection of 0.03–0.1 mmol/kg of gadopentetate dimeglumine, coronal images of the lower half of the body and oblique sagittal images of the thoracic aorta are obtained (Fig 2). A rapid manual injection of gadopentetate dimeglumine is performed within 5 seconds during an intentionally prolonged inspiration and is followed by flushing with 20 mL of physiologic saline solution. Five imaging sets are consecutively acquired after the gadopentetate dimeglumine injection. After imaging of the first anatomic area (eg, the abdomen), the following two anatomic areas (eg, the thighs, calves, and chest) are imaged as described earlier.
The 3D data set obtained immediately before administration of gadopentetate dimeglumine is used as a mask for subsequent image subtraction and is subtracted section by section from each of the five original 3D data sets acquired after administration of gadopentetate dimeglumine. The subtraction is performed by using commercially available software (Philips Medical Systems). A dynamic subtraction contrast-enhanced MR angiogram is created by compressing subtracted images of each phase by means of maximum intensity projection postprocessing.
To achieve high-quality arterial imaging with a small dose of gadopentetate dimeglumine, dynamic subtraction MR angiography needs to be optimized in terms of a variety of clinical settings. There are several points to be considered: pulse sequence, injection dose, injection rate, timing of the start of data acquisition, imaging time and breath holding, section thickness, coil selection, subtraction, and postprocessing with subtraction and addition.
Pulse Sequence
Spoiled T1-weighted GRE sequences (3D fast field echo, 3D turbo field echo) are more suitable for dynamic subtraction MR angiography than a nonspoiled GRE sequence. With a spoiled GRE sequence, the signal intensity increases as the blood level of gadopentetate dimeglumine increases. However, with a nonspoiled GRE sequence, the signal intensity decreases at a blood level of gadopentetate dimeglumine higher than 2 mmol/L due to reduction of T2 relaxation (actually T2* decay) time (6,11) (Fig 4).
View larger version (20K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4. Signal intensity as a function of gadopentetate dimeglumine (Gd-DTPA) concentration for various pulse sequences: spin echo (SE) (500/15); turbo spin echo (TSE) (500/15, echo train length of three); gradient and spin echo (GraSE) (500/15, echo train length of three, echo-planar imaging factor of three); 3D spoiled fast field echo (FFE) (8.8/2.8, 35° flip angle); 3D spoiled turbo field echo (TFE) (8/2.5, 20° flip angle); and two-dimensional (2D) nonspoiled fast field echo (111/2.3, 70° flip angle). mM = millimoles per liter.
|
|
Among spoiled GRE sequences, the signal-to-noise ratio and vascular contrast are higher in 3D fast field echo images than in 3D turbo field echo images, whereas the artifact from cardiac motion is much less severe in 3D turbo field echo images than in 3D fast field echo images. Therefore, we use the 3D fast field echo sequence for imaging any anatomic areas other than the chest and the 3D turbo field echo sequence for imaging the chest (Table).
Sequence parameters that influence the signal-to-noise ratio of vessels at 3D fast field echo or 3D turbo field echo imaging are the repetition time and flip angle (6,11). The short repetition time reduces the acquisition time, thus permitting breath-hold data acquisition and enabling imaging of arteries before contrast material reaches the veins. However, as the repetition time is decreased, the signal-to-noise ratio of enhanced vessels decreases in proportion to the square root of the repetition time. Thus, high-quality vascular imaging can be achieved even with a moderately short repetition time when the timing of data acquisition is appropriate.
The flip angle should be optimized according to the repetition time to achieve a high signal-to-noise ratio in vessels. For contrast material–induced T1 values of vessels enhanced with bolus injection, the optimal flip angle for a repetition time of 10 msec is 25°–40° (6).
Intravoxel dephasing can be reduced by using short echo times and velocity-compensating gradients.
Injection Dose
The arterial contrast achieved in dynamic subtraction MR angiography depends on the dose of gadopentetate dimeglumine injected. However, the dose of gadopentetate dimeglumine can be minimized when the injection method is optimal. Our method consists of a rapid manual injection performed within 5 seconds during an intentionally prolonged inspiration, which maintains a negative intrathoracic pressure and helps the injected gadopentetate dimeglumine bolus pass smoothly into the heart. With our injection method, the minimum dose of gadopentetate dimeglumine necessary to demonstrate main arteries is 0.03 mmol/kg (Fig 5). A larger dose (0.05–0.1 mmol/kg) is necessary to demonstrate small branches of main arteries. Therefore, the injection dose should vary according to the clinical setting.
View larger version (126K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5a. Dose escalation study. Dynamic subtraction MR angiograms obtained after a bolus injection of 0.01 (a), 0.02 (b), 0.03 (c), 0.05 (d), and 0.1 (e) mmol/kg of gadopentetate dimeglumine show that 0.03 mmol/kg is the minimum dose necessary to image the main arteries of the thighs, such as the superficial and deep femoral arteries.
|
|
View larger version (122K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5b. Dose escalation study. Dynamic subtraction MR angiograms obtained after a bolus injection of 0.01 (a), 0.02 (b), 0.03 (c), 0.05 (d), and 0.1 (e) mmol/kg of gadopentetate dimeglumine show that 0.03 mmol/kg is the minimum dose necessary to image the main arteries of the thighs, such as the superficial and deep femoral arteries.
|
|
View larger version (122K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5c. Dose escalation study. Dynamic subtraction MR angiograms obtained after a bolus injection of 0.01 (a), 0.02 (b), 0.03 (c), 0.05 (d), and 0.1 (e) mmol/kg of gadopentetate dimeglumine show that 0.03 mmol/kg is the minimum dose necessary to image the main arteries of the thighs, such as the superficial and deep femoral arteries.
|
|
View larger version (128K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5d. Dose escalation study. Dynamic subtraction MR angiograms obtained after a bolus injection of 0.01 (a), 0.02 (b), 0.03 (c), 0.05 (d), and 0.1 (e) mmol/kg of gadopentetate dimeglumine show that 0.03 mmol/kg is the minimum dose necessary to image the main arteries of the thighs, such as the superficial and deep femoral arteries.
|
|
View larger version (114K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5e. Dose escalation study. Dynamic subtraction MR angiograms obtained after a bolus injection of 0.01 (a), 0.02 (b), 0.03 (c), 0.05 (d), and 0.1 (e) mmol/kg of gadopentetate dimeglumine show that 0.03 mmol/kg is the minimum dose necessary to image the main arteries of the thighs, such as the superficial and deep femoral arteries.
|
|
In a complete study of the main arteries of the entire lower body, three separate injections of 0.03–0.05 mmol/kg (total dose, 0.1–0.15 mmol/kg) are given in each anatomic area (abdomen, both thighs, both calves) (10). In a study of the thoracic and abdominal aorta and the iliac and femoral arteries, three separate injections of 0.03–0.05 mmol/kg (total dose, 0.1–0.15 mmol/kg) are also given in each anatomic area (abdomen, both thighs, chest). In a study of the renal artery and portomesenteric vein, a single injection of 0.1 mmol/kg is used to image only a single anatomic area.
Injection Rate
The peak blood concentration of injected gadopentetate dimeglumine depends on the bolus injection rate (12–14). The signal intensity obtained with the spoiled GRE sequences (3D fast field echo and 3D turbo field echo) increases according to the gadopentetate dimeglumine concentration (Fig 4). Therefore, with a spoiled GRE sequence, a fast injection rate (2–5 mL/sec) is used. However, a nonspoiled GRE sequence does not show high signal intensity when the maximum gadopentetate dimeglumine concentration reaches more than 2 mmol/L. Therefore, with a nonspoiled GRE sequence, slow injection of nondiluted gadopentetate dimeglumine or rapid injection of diluted gadopentetate dimeglumine would be recommended.
Timing of Start of Data Acquisition
The timing of the start of data acquisition after a bolus gadopentetate dimeglumine injection has a major effect on optimal arterial enhancement and separate imaging of arteries and veins. Three methods for proper timing of the start of data acquisition have been reported: test bolus injection, automatic triggering software (eg, Bolus Trak [Philips Medical Systems], Smart Prep [GE Medical Systems, Milwaukee, Wis]), and repeating fast sequences (6,10–18). Timing the data acquisition to coincide with the first pass of contrast material provides excellent arterial imaging. We use the method of repeating fast sequences (Fig 2).
In a complete study of the lower half of the body, the first sequence is started 5–10 seconds after the start of the injection for imaging of the abdomen. One of the five repeated sequences performed after the injection coincides with the first pass of administered gadopentetate dimeglumine. In the subsequent imaging of the thighs and calves, the injection used in imaging of the abdomen acts as a test injection and indicates the best timing of the start of data acquisition. Thus, the timing of the start of data acquisition in the other anatomic areas can be adjusted on the basis of the original source images from the previous data acquisition. For example, when the first pass of contrast material coincides with the second or third sequence performed after the injection for the first anatomic area imaged, the start of data acquisition in the next anatomic area is delayed for a single or double acquisition time so that the first pass of contrast material coincides with the first sequence in the next anatomic area.
Imaging Time and Breath Holding
Use of a breath-hold technique avoids the problems of motion artifacts from respiration and peristaltic bowel movement in imaging of the abdomen. The imaging time is required to be short enough even for a debilitated patient who is unable to breath hold for very long. For this reason, use of several techniques including half-Fourier acquisition, partial echo acquisition, reduced number of sections, rectangular field of view, and zero-filled interpolation should be considered (19–21). A special k-space technique called 3D time-resolved imaging of contrast kinetics is also an extremely fast 3D MR angiography technique (6,7). Another benefit of a short imaging time is that venous enhancement can be minimized in imaging of arteries (Fig 6). We set the imaging time at about 9–15 seconds for the abdomen by using partial echo acquisition, reduced number of sections, rectangular field of view, and zero-filled interpolation (Table). If patients are unable to breath hold for the entire 3D GRE acquisition, they are instructed to breathe in a shallow manner after holding their breath for as long as possible.
View larger version (11K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6. Signal intensity of the lower abdominal aorta as a function of time after a bolus injection of 0.04 mmol/kg of gadopentetate dimeglumine in a healthy young
volunteer. The first pass of the injected gadopentetate dimeglumine occurs 13 seconds after injection and peaks at 17 seconds. The duration of the first pass is about 12 seconds.
|
|
In imaging the thighs, a breath-hold technique is not always necessary, although we use a breath-hold technique to minimize motion artifacts. In imaging the calves, breath holding is not used. The imaging time is set at about 10–19 seconds and 18–21 seconds for imaging the thighs and calves, respectively. In particular, a moderately short repetition time (12–13 msec) is used in combination with half-Fourier acquisition to obtain a high signal-to-noise ratio in the calves. The imaging time is set at about 12–15 msec for imaging the chest.
Section Thickness
The section thickness should vary depending on the anatomic sites and clinical setting. Image resolution can be increased with zero interpolation, which requires no additional time for data collection. Overlapping sections in which the section spacing is set at half of the section thickness can be obtained by means of through-plane zero interpolation. A spatial resolution of approximately 4–5 mm is sufficient to image the abdominal aorta and iliac arteries, and the section thickness can be set at 8–10 mm with 4–5-mm spacing by using overlapping sections (Table). In contrast, a spatial resolution of 1.5–3.0 mm, which can be obtained when the section thickness is set at 3–6 mm with overlapping sections, is required for imaging the main renal arteries and tibioperoneal arteries and allows clinically useful grading of arterial stenosis or occlusion.
Coil Selection
The decision about which coil to use for dynamic subtraction MR angiography is influenced by trade-offs between signal-to-noise ratio, imaging time, spatial resolution, and anatomic coverage. In imaging the abdomen, thighs, and chest, use of a body coil can produce good-quality images of the abdominal aorta, iliac arteries, and femoral arteries (Table). However, in the popliteal and tibio
peroneal arteries, which range in diameter from nearly 1 cm to 1 mm, use of a specialized phased-array coil such as a synergy spine coil will be advantageous to obtain a high signal-to-noise ratio in the vessels (11,22).
Subtraction
Subtraction of the 3D data sets is necessary to eliminate background signal, including a hyperintense (on T1-weighted images) mural thrombus of an aneurysm (Fig 7) and fatty tissue (Fig 8) (3,6,8,23). Especially for imaging additional anatomic areas such as the thighs and calves, subtraction is absolutely necessary to eliminate signal from veins enhanced by already circulating gadopentetate dimeglumine from the previous injection. Another advantage of subtraction is removal of nonenhancing aliased information (Fig 7), thus providing an option to reducing the imaging time by using a rectangular field of view. However, misregistration artifacts from motion and peristaltic bowel movement may degrade image quality. Therefore, to reduce such artifacts as much as possible, the abdomen should be imaged first when multiple anatomic areas including the abdomen are to be imaged. In imaging of the chest, subtraction is not always necessary.
View larger version (150K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7a. Iliac artery aneurysm. (a) Nonenhanced source image shows an aneurysm of the right common iliac artery with a hyperintense mural thrombus (arrow). (b) Arterial-phase source image shows enhancement of the aneurysmal lumen (arrowhead) at the level of the thrombus. (c) Subtracted source image shows only the enhanced aneurysmal lumen; the thrombus is no longer seen. Note that aliased information (arrows in b) is eliminated with subtraction.
|
|
View larger version (145K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7b. Iliac artery aneurysm. (a) Nonenhanced source image shows an aneurysm of the right common iliac artery with a hyperintense mural thrombus (arrow). (b) Arterial-phase source image shows enhancement of the aneurysmal lumen (arrowhead) at the level of the thrombus. (c) Subtracted source image shows only the enhanced aneurysmal lumen; the thrombus is no longer seen. Note that aliased information (arrows in b) is eliminated with subtraction.
|
|
View larger version (140K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7c. Iliac artery aneurysm. (a) Nonenhanced source image shows an aneurysm of the right common iliac artery with a hyperintense mural thrombus (arrow). (b) Arterial-phase source image shows enhancement of the aneurysmal lumen (arrowhead) at the level of the thrombus. (c) Subtracted source image shows only the enhanced aneurysmal lumen; the thrombus is no longer seen. Note that aliased information (arrows in b) is eliminated with subtraction.
|
|
View larger version (99K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8a. Abdominal aortic aneurysm. (a) Contrast-enhanced MR angiogram shows an abdominal aortic aneurysm (arrows), but bilateral iliac arteries are not demonstrated. (b) Dynamic subtraction MR angiogram clearly shows the aneurysm (arrows) and iliac arteries (arrowheads) owing to subtraction of background signal, such as that from fatty tissue.
|
|
View larger version (93K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8b. Abdominal aortic aneurysm. (a) Contrast-enhanced MR angiogram shows an abdominal aortic aneurysm (arrows), but bilateral iliac arteries are not demonstrated. (b) Dynamic subtraction MR angiogram clearly shows the aneurysm (arrows) and iliac arteries (arrowheads) owing to subtraction of background signal, such as that from fatty tissue.
|
|
Postprocessing with Subtraction and Addition
Further postprocessing including double subtraction and addition is performed by using commercially available software (Philips Medical Systems). Double subtraction enables demonstration of arteries and veins separately (Figs 9, 10). When necessary, the subtracted images of the arterial phase are further subtracted section by section from the subtracted images of the venous phase before reconstruction with maximum intensity projection postprocessing (6,24). This technique is very useful in imaging deep veins of the lower extremity (Fig 10) and the portomesenteric veins (Fig 11).
View larger version (55K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10a. Deep vein thrombosis due to antiphospholipid antibody syndrome. (a) Arterial-phase dynamic subtraction MR angiogram obtained with two separate injections of 0.05 mmol/kg of gadopentetate dimeglumine shows no occlusion of the main arteries of the lower extremity. (b) Venous image from dynamic subtraction MR angiography obtained by using double subtraction shows occlusion of the left popliteal and femoral veins (arrows).
|
|
View larger version (65K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10b. Deep vein thrombosis due to antiphospholipid antibody syndrome. (a) Arterial-phase dynamic subtraction MR angiogram obtained with two separate injections of 0.05 mmol/kg of gadopentetate dimeglumine shows no occlusion of the main arteries of the lower extremity. (b) Venous image from dynamic subtraction MR angiography obtained by using double subtraction shows occlusion of the left popliteal and femoral veins (arrows).
|
|
View larger version (140K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11a. Esophageal varix and splenorenal shunt in a patient with liver cirrhosis. (a) Portal-phase dynamic subtraction MR angiogram obtained with a single injection of 0.1 mmol/kg of gadopentetate dimeglumine shows an esophageal varix (arrows) and splenorenal shunt (arrowheads), which are not clearly seen due to superimposition of the enhanced abdominal aorta. (b) MR portogram obtained by using double subtraction clearly shows the portal venous system (open arrows), along with the esophageal varix (solid arrows) and splenorenal shunt (arrowheads). (Reprinted, with permission, from reference 24.)
|
|
View larger version (145K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11b. Esophageal varix and splenorenal shunt in a patient with liver cirrhosis. (a) Portal-phase dynamic subtraction MR angiogram obtained with a single injection of 0.1 mmol/kg of gadopentetate dimeglumine shows an esophageal varix (arrows) and splenorenal shunt (arrowheads), which are not clearly seen due to superimposition of the enhanced abdominal aorta. (b) MR portogram obtained by using double subtraction clearly shows the portal venous system (open arrows), along with the esophageal varix (solid arrows) and splenorenal shunt (arrowheads). (Reprinted, with permission, from reference 24.)
|
|
Postprocessing by addition of the subtracted images enables demonstration of arteries in a single image when the upper and lower segments of the arteries are shown in different acquisition phases (Figs 12, 13). This technique is especially useful in demonstrating aortic dissection with a late-filling pseudolumen.
View larger version (110K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13a. Prolonged circulation time in a 78-year-old patient. (a) Early-phase dynamic subtraction MR angiogram shows enhancement of the proximal abdominal aorta (arrows). (b) Image from the next phase of dynamic subtraction MR angiography shows enhancement of the distal abdominal aorta (arrowheads). (c) Arterial image from dynamic subtraction MR angiography obtained by adding a and b shows the entire abdominal aorta and iliac arteries.
|
|
View larger version (122K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13b. Prolonged circulation time in a 78-year-old patient. (a) Early-phase dynamic subtraction MR angiogram shows enhancement of the proximal abdominal aorta (arrows). (b) Image from the next phase of dynamic subtraction MR angiography shows enhancement of the distal abdominal aorta (arrowheads). (c) Arterial image from dynamic subtraction MR angiography obtained by adding a and b shows the entire abdominal aorta and iliac arteries.
|
|
View larger version (131K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13c. Prolonged circulation time in a 78-year-old patient. (a) Early-phase dynamic subtraction MR angiogram shows enhancement of the proximal abdominal aorta (arrows). (b) Image from the next phase of dynamic subtraction MR angiography shows enhancement of the distal abdominal aorta (arrowheads). (c) Arterial image from dynamic subtraction MR angiography obtained by adding a and b shows the entire abdominal aorta and iliac arteries.
|
|
|
Clinical Applications
|
|---|
Atherosclerotic Occlusive Disease
Atherosclerosis is a systemic process that causes arterial stenosis, occlusion, and aneurysmal dilatation. Symptomatic patients with atherosclerotic disease often require precise evaluation of long segments of arteries. For this purpose, the lower half of the body is imaged by using three separate gadopentetate dimeglumine injections (Fig 2) (8–10). To image the lower body, 0.04–0.06, 0.03–0.05, and 0.03–0.05 mmol/kg (total, 0.1–0.15 mmol/kg) is used for imaging of the abdomen, thighs, and calves, respectively.
In our evaluation of atherosclerotic occlusive disease, we have found that the sensitivity and specificity of dynamic subtraction MR angiography for the detection of severe stenosis (>50%) and occlusion of the aorta and main runoff arteries are 91% and 92%, respectively. In a large-scale study, a higher sensitivity and specificity of 97.1% and 99.2%, respectively, were achieved (25). No occlusions were overlooked, and collateral arteries and reconstitution were clearly demonstrated. Dynamic subtraction MR angiography provides an overview of the main arteries of the lower half of the body (Fig 14) and can be a suitable screening method for atherosclerotic disease.
View larger version (49K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 14a. Atherosclerotic occlusive disease. (a) Dynamic subtraction MR angiogram obtained with three separate injections of gadopentetate dimeglumine shows long segmental occlusions of the right superficial femoral artery (solid arrows) and the left external iliac and superficial femoral arteries (solid arrowheads). Note the reconstitution of the bilateral distal superficial femoral arteries via small collateral arteries. Also note the false-positive occlusions of the right external iliac artery (open arrow) and left popliteal artery (open arrowhead) resulting from inappropriate positioning of the imaging volume. (b) Conventional angiogram also shows reconstitution of the bilateral distal superficial femoral arteries via small collateral arteries. (Reprinted, with permission, from reference 10.)
|
|
View larger version (57K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 14b. Atherosclerotic occlusive disease. (a) Dynamic subtraction MR angiogram obtained with three separate injections of gadopentetate dimeglumine shows long segmental occlusions of the right superficial femoral artery (solid arrows) and the left external iliac and superficial femoral arteries (solid arrowheads). Note the reconstitution of the bilateral distal superficial femoral arteries via small collateral arteries. Also note the false-positive occlusions of the right external iliac artery (open arrow) and left popliteal artery (open arrowhead) resulting from inappropriate positioning of the imaging volume. (b) Conventional angiogram also shows reconstitution of the bilateral distal superficial femoral arteries via small collateral arteries. (Reprinted, with permission, from reference 10.)
|
|
Aneurysm of Aortoiliac Arteries
In our evaluation of aneurysms of aortoiliac arteries, we have found that the accuracy of dynamic subtraction MR angiography is 100%. The original source images demonstrate associated extra-luminal components such as a mural thrombus of the aneurysm, whereas conventional angiography does not (Fig 15). Conversely, the subtracted source images demonstrate only the vascular lumen, even when a mural thrombus has high signal intensity on the nonenhanced mask images (Fig 7). Therefore, careful interpretation of dynamic subtraction MR angiograms as well as both the original and subtracted source images can produce a high accuracy in evaluation of aneurysms.
View larger version (68K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 15a. Iliac artery aneurysm. (a) Dynamic subtraction MR angiogram obtained with three separate injections of gadopentetate dimeglumine shows an aneurysm of the right common iliac artery (arrows) and distal abdominal aorta (arrowhead). Note the irregular lumen of the bilateral femoral arteries. (b) Coronal source image shows both the patent lumen and a mural thrombus of the aneurysm (arrowheads). (c) Conventional angiogram show only the lumen of the aneurysm.
|
|
View larger version (182K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 15b. Iliac artery aneurysm. (a) Dynamic subtraction MR angiogram obtained with three separate injections of gadopentetate dimeglumine shows an aneurysm of the right common iliac artery (arrows) and distal abdominal aorta (arrowhead). Note the irregular lumen of the bilateral femoral arteries. (b) Coronal source image shows both the patent lumen and a mural thrombus of the aneurysm (arrowheads). (c) Conventional angiogram show only the lumen of the aneurysm.
|
|
View larger version (48K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 15c. Iliac artery aneurysm. (a) Dynamic subtraction MR angiogram obtained with three separate injections of gadopentetate dimeglumine shows an aneurysm of the right common iliac artery (arrows) and distal abdominal aorta (arrowhead). Note the irregular lumen of the bilateral femoral arteries. (b) Coronal source image shows both the patent lumen and a mural thrombus of the aneurysm (arrowheads). (c) Conventional angiogram show only the lumen of the aneurysm.
|
|
Bypass Graft
The patency of a bypass graft can be easily determined with dynamic subtraction MR angiography. The anastomoses between the graft and the proximal and distal arteries can be seen as weblike stenoses (Figs 16, 17).
View larger version (41K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 16. Patent Y graft. Dynamic subtraction MR angiogram obtained with three separate injections of gadopentetate dimeglumine shows patency of a Y graft (arrows). The stricturelike regions correspond to the anastomoses (arrowheads).
|
|
View larger version (50K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 17. Patent femorofemoral bypass graft. Dynamic subtraction MR angiogram obtained with three separate injections of gadopentetate dimeglumine shows patency of a femorofemoral bypass graft (arrows). Note the occlusion of the right iliac artery (arrowheads).
|
|
Takayasu Arteritis
In Takayasu arteritis, marked intimal proliferation and fibrosis lead to occlusion and narrowing of the aorta and involved arteries (Figs 18, 19). Aneurysms may also be found in the involved segment of the aorta and its major branches (Fig 19). Therefore, dynamic subtraction MR angiography should include the entire aorta, arch vessels, and major abdominal aortic branches (26). We usually image three anatomic sites in the abdomen, thighs, and chest including the lower neck in the order of data acquisition.
View larger version (57K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 18. Takayasu arteritis. Dynamic subtraction MR angiogram obtained with two separate injections of gadopentetate dimeglumine shows aortic occlusion from the upper thoracic descending aorta to the abdominal aorta (solid arrows). A long aortoaortic bypass graft is clearly seen (solid arrowheads). The stricturelike region corresponds to an anastomosis (open arrow). Also note the stenosis of the left carotid artery (open arrowhead).
|
|
View larger version (47K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 19. Takayasu arteritis. Dynamic subtraction MR angiogram obtained with three separate injections of gadopentetate dimeglumine shows multiple aortic aneurysms (arrows) and narrowing of the thoracic aorta (solid arrowhead). The right renal artery shows marked dilatation of the proximal part and stenosis of the distal part (open arrowhead).
|
|
Aortic Dissection
In evaluation of aortic dissection, the entire aorta, arch vessels, and major aortic branches should be included in the examination. The intimal flap, enhancement of the false lumen, compression of the true lumen, and obstruction of aortic branches can be demonstrated with dynamic subtraction MR angiography (Figs 20, 21). Simultaneous interpretation of original and subtracted source images can facilitate determination of whether aortic branches communicate with the true or the false lumen.
View larger version (83K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 20. Aortic dissection (DeBakey type III). Dynamic subtraction MR angiogram obtained with three separate injections of gadopentetate dimeglumine shows the true lumen (solid arrows) compressed by the patent pseudolumen (solid arrowheads). Note the thrombus in the pseudolumen (open arrows). The left renal artery (open arrowhead) branches off from the true lumen.
|
|
View larger version (90K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 21a. Dissection of the abdominal aorta. Dynamic subtraction MR angiogram (a) and coronal source image (b) show an abdominal aortic dissection that involves the left renal artery. The true lumen of the left renal artery (arrow) demonstrates severe stenosis due to compression by the pseudolumen (arrowheads).
|
|
View larger version (92K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 21b. Dissection of the abdominal aorta. Dynamic subtraction MR angiogram (a) and coronal source image (b) show an abdominal aortic dissection that involves the left renal artery. The true lumen of the left renal artery (arrow) demonstrates severe stenosis due to compression by the pseudolumen (arrowheads).
|
|
Antiphospholipid Antibody Syndrome
Patients with the antiphospholipid antibody, which is associated with the connective tissue diseases including systemic lupus erythematosus, have recurrent venous and arterial thrombosis in any anatomic regions. In a symptomatic patient, arteries as well as veins need to be examined with dynamic subtraction MR angiography. Further postprocessing by using subtraction allows imaging of arteries and veins separately (Figs 9, 10).
Renal Artery Disease
Special attention should be paid to the renal arteries because renal artery stenosis can lead to renovascular hypertension and impaired renal function. Contrast-enhanced MR angiography seems well suited to this clinical setting because gadopentetate dimeglumine is much less nephrotoxic than the iodinated contrast material used in conventional digital subtraction angiography and computed tomographic angiography (27–30). In evaluation of the renal arteries, 0.1 mmol/kg is used to image small branches of the renal arteries or accessory renal arteries. To obtain optimal arterial contrast with high spatial resolution and to image arteries and veins separately, the acquisition time and section thickness of the sequence should be as short and thin, respectively, as possible. Subtraction allows imaging of a small accessory renal artery, although there may be subtraction misregistration artifacts. Note that renal artery stenosis can be overestimated with dynamic subtraction MR angiography (29). Other renovascular diseases include dissection, occlusion, and aneurysm (Figs 21–23).
View larger version (138K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 22a. Renal artery occlusion causing renal infarction.
(a) Dynamic subtraction MR angiogram shows complete occlusion of the proximal left renal artery (arrow). The distal left renal artery is not enhanced. (b) Conventional angiogram also shows complete occlusion of the left renal artery (arrow).
|
|
View larger version (173K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 22b. Renal artery occlusion causing renal infarction.
(a) Dynamic subtraction MR angiogram shows complete occlusion of the proximal left renal artery (arrow). The distal left renal artery is not enhanced. (b) Conventional angiogram also shows complete occlusion of the left renal artery (arrow).
|
|
View larger version (152K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 23a. Renal artery aneurysm in a patient with renovascular hypertension. Dynamic subtraction MR angiogram (a) and conventional angiogram (b) show a saccular aneurysm of the distal left renal artery (arrow).
|
|
View larger version (174K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 23b. Renal artery aneurysm in a patient with renovascular hypertension. Dynamic subtraction MR angiogram (a) and conventional angiogram (b) show a saccular aneurysm of the distal left renal artery (arrow).
|
|
Pelvic Vascular Disease
Uterine arteriovenous malformation is an acquired vascular lesion that manifests as remarkably tortuous and dilated uterine vessels in the myometrium and may project into the endometrial cavity. Dynamic subtraction MR angiography clearly demonstrates a uterine arteriovenous malformation (Fig 24) (31). A varix of the gonadal veins may be misdiagnosed as a pelvic arteriovenous malformation at dynamic subtraction MR angiography when the acquisition time is not short enough to image the arteries and veins separately.
View larger version (151K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 24. Uterine arteriovenous malformation in a 28-year-old woman with massive uterine bleeding. Dynamic subtraction MR angiogram shows a uterine arteriovenous malformation as tortuous and dilated uterine vessels (arrows). Note the early venous return (arrowheads).
|
|
Portomesenteric Venous System
In evaluation of the portomesenteric veins (32,33), 0.1 mmol/kg of gadopentetate dimeglumine is used to image portal vein obstruction and portal venous collateral pathways (24). A large field of view is necessary to detect a portosystemic shunt (Fig 25). An esophagogastric varix can be clearly shown with dynamic subtraction MR angiography, especially arterial-phase–subtracted dynamic subtraction MR angiography (Figs 9, 11). This technique is also useful for assessment and monitoring of treatment effect from sclerotherapy.
View larger version (141K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 25. Mesenteric-gonadal shunt in a patient with hepatic encephalopathy. Dynamic subtraction MR angiogram shows a portosystemic shunt from the superior mesenteric vein (arrows) to the gonadal vein (arrowheads). (Reprinted, with permission, from reference 24.)
|
|
|
Diagnostic Pitfalls
|
|---|
Dynamic subtraction MR angiography is limited by several diagnostic pitfalls that are not encountered at conventional angiography, including inappropriate image coverage, subtraction misregistration artifact, pseudodissection, and phase artifact (6, 10,13). Inapproprate positioning of the imaging volume may produce false-positive occlusion of tortuous arteries such as the external iliac artery (Fig 14) (10). Pseudodissection appears as a central dark line in a vessel (Fig 26) and can occur when the timing of the data acquisition relative to the first pass of contrast material is suboptimal. That is, an insufficient arterial concentration of gadopentetate dimeglumine is synchronized to the central k-space data sampling, whereas the peak arterial concentration is synchronized to the peripheral k-space data sampling. Sequences with half-Fourier acquisition are especially susceptible to pseudodissection artifact. Phase artifact appears as a ghost artifact parallel to vessels with very high signal intensity (Fig 27).
View larger | |
Top
Abstract
Introduction
