DOI: 10.1148/radiol.2443061647
(Radiology 2007;244:907-913.)
© RSNA, 2007
Vascular and Interventional Radiology |
Intraindividual Comparison of High-Spatial-Resolution Abdominal MR Angiography at 1.5 T and 3.0 T: Initial Experience1
Henrik J. Michaely, MD,
Harald Kramer, MD,
Olaf Dietrich, PhD,
Kambiz Nael, MD,
Klaus-Peter Lodemann, PhD,
Maximilian F. Reiser, MD, and
Stefan O. Schoenberg, MD
1 From the Institute of Clinical Radiology, University Hospitals Grosshadern, Ludwig-Maximilians-University Munich, Marchioninistrasse 15, 81377 Munich, Germany (H.J.M., H.K., O.D., M.F.R., S.O.S.); Department of Cardiovascular Radiology, UCLA, Los Angeles, Calif (K.N.); and Bracco-Altana Pharma, Konstanz, Germany (K.P.L.). From the 2006 RSNA Annual Meeting. Received September 24, 2006; revision requested December 5; revision received January 9, 2007; accepted February 20; final version accepted April 9.
Address correspondence to H.J.M. (e-mail: henrik.michaely{at}med.uni-muenchen.de).
 |
ABSTRACT
|
|---|
Purpose: To prospectively compare three-dimensional (3D) contrast material–enhanced abdominal magnetic resonance (MR) angiography at 1.5 and 3.0 T intraindividually in healthy volunteers.
Materials and Methods: After institutional review board approval and informed consent were obtained, 15 healthy male volunteers (age range, 24–41 years) underwent one abdominal 3D contrast-enhanced MR angiographic examination each at 1.5 and 3.0 T in random order. Fast 3D gradient-echo sequence with parallel imaging acceleration factor of three was used for MR angiography; acquired spatial resolutions were 1 x 0.8 x 1 mm3 (imaging time, 19 seconds) at 1.5 T and 0.9 x 0.8 x 0.9 mm3 (imaging time, 18 seconds) at 3.0 T. With the latter, volume of the 3D slab was 8% larger. At 1.5 T, 20-mL bolus of gadobenate dimeglumine was delivered at 2 mL/sec; at 3.0 T, 15-mL bolus was delivered at 2.5 mL/sec. Two blinded radiologists rated image quality of aorta and proximal renal arteries in consensus with five-point scale (4 = very good, 0 = nondiagnostic) according to sequence and in direct intraindividual comparison. Visibility of proximal and segmental renal arteries was rated with three-point scale (3 = completely visible, 1 = nonvisible). Signal-to-noise ratio (SNR) was determined with phantoms. For statistical analysis of the SNRs, t tests were used.
Results: All MR angiographic measurements were diagnostic. Median score for image quality at both field strengths was 4. Depiction of proximal renal arteries was rated 3 at both field strengths. The visibility of the distal renal arteries was better at 3.0 T (median score, 3) than at 1.5 T (median score, 2). With direct comparison, 3.0-T MR angiography was better in 14 of 15 cases; no field strength was preferred in the other case. Mean SNR was significantly (P < .001) higher at 3.0 T (17.8 ± 0.09 [standard deviation]) than at 1.5 T (11.9 ± 0.10).
Conclusion: MR angiography at 3.0 T provided better vessel visibility and SNR than did that at 1.5 T, although voxel size and imaging time were reduced.
© RSNA, 2007
|
INTRODUCTION
|
|---|
Magnetic resonance (MR) angiography has become a standard technique for examination of the abdominal aorta and the renal arteries (1,2), particularly in patients with renal disease in whom the application of potentially nephrotoxic iodinated contrast agents is contraindicated (3). A recent multicenter study, however, has challenged the diagnostic accuracy of MR angiography in the detection and grading of renal artery stenoses (4). In that study, the reported sensitivity was 62%. The poor results of that multicenter trial were attributed to the outdated MR angiographic technique used, a technique that included large voxel volumes and relatively long imaging times.
A dedicated MR angiographic sequence for the renal arteries has to fulfill three fundamental requirements: high spatial resolution with isotropic voxels acquired during a minimal imaging time. In 1998, Hoogeveen et al (5) investigated the theoretical requirements for spatial resolution of MR angiography and concluded that at least 3 pixels constitute the luminal diameter; this finding implies an acquired spatial resolution of at least 1.5 mm for the proximal renal arteries. Schoenberg et al (6) found that measurement of the vessel area instead of the vessel diameter allows for a more precise grading of renal arterial stenoses and correlates well with findings at intravascular ultrasonography. For lossless reformations, isotropic data sets are required. Vasbinder and colleagues (7) addressed the issue of distal renal arterial motion during breath-hold MR angiography and concluded that the distal parts of the renal vessels are always subject to random diaphragmatic motion. Therefore, fast MR angiographic acquisition is mandatory for depiction of the distal renal arteries, as in the case of suspected fibromuscular dysplasia (8).
In clinical practice, high-spatial-resolution isotropic MR angiography with minimized imaging time is limited by the available signal-to-noise-ratio (SNR). Because of the almost doubled SNR (9) and the improved background suppression (9,10) with the prolonged T1 times of most tissues and the increased effectiveness of contrast agents at higher field strengths (11), 3.0-T imaging seems very promising for abdominal MR angiography. Investigators of an initial feasibility study have reported promising results for abdominal MR angiography at 3.0 T (12). However, to our knowledge, in no studies have abdominal MR angiography at 1.5 T and that at 3.0 T been intraindividually compared. Thus, the aim of this study was to prospectively compare three-dimensional (3D) contrast material–enhanced abdominal MR angiography at 1.5 T and at 3.0 T intraindividually in healthy volunteers.
|
MATERIALS AND METHODS
|
|---|
For this study, contrast agent was provided for free by Bracco-Altana Pharma (Konstanz, Germany). Bracco-Altana Pharma also provided financial support. The authors who are not employees of Bracco-Altana Pharma had full control of inclusion of any data and information that might present a conflict of interest for authors who are employees of Bracco-Altana Pharma.
Volunteers
After institutional review board approval and informed consent were obtained, 15 healthy male volunteers (mean age, 30.8 years ± 4.9 [standard deviation]; range, 24–41 years) with no history of renal or vascular disease were included. The institutional review board allowed only male volunteers for this study because the age range of 24–41 years would correspond to childbearing years for women. Thus, the potential for risks of administering the contrast agent could exist for women and fetuses. The volunteers underwent two MR angiographic examinations—one at 1.5 T and one at 3.0 T—in random order at least 2 days apart (mean, 8.9 days; range, 2–21 days). For administration of the contrast agent, a 20-gauge needle was placed in an antecubital vein in all volunteers.
MR Imaging
All measurements were performed with two imaging units: (a) a 32-channel 1.5-T whole-body unit (Magnetom Avanto; Siemens Medical Solutions, Erlangen, Germany) with a gradient strength of 45 mT/m and slew rate of 200 mT/m/msec and (b) a 32-channel 3.0-T whole-body unit (Magnetom Tim Trio; Siemens Medical Solutions) with the same gradient performance and same coil design. For signal reception, one body matrix with six independent receiver elements and six elements of the spine matrix were used—that is, a total of 12 independent receiver elements. To allow for correct positioning, steady-state free precession (true fast imaging with steady-state precession) localizing sequences in coronal and transverse orientation were performed first.
For MR angiography, a fast 3D spoiled gradient-echo sequence with Cartesian k-space readout was used (Table). The MR angiographic sequence was centered to the aorta and acquired in a slightly oblique plane to follow the abdominal aorta at both field strengths. The MR angiographic image was acquired during breath hold in inspiration. To guarantee proper timing of MR angiography, a test bolus technique was applied: A 1-mL bolus of gadobenate dimeglumine (Multihance; Bracco Pharma, Milan, Italy), which is characterized by higher relaxivity and higher enhancement compared with standard gadobenate chelates (11), was injected at rates of 2 mL/sec at 1.5 T and 2.5 mL/sec at 3.0 T, followed by a 30-mL saline chaser. An automated injector pump (Spectris Solaris; Medrad, Indianola, Pa) was used at 1.5 T; for 3.0-T imaging, a different model (Spectris Solaris EP; Medrad) was employed. The transverse section of the test bolus was positioned to cut the aorta perpendicular at the level of the renal arteries.
The time to the center of the MR angiographic sequence was subtracted from the time of peak enhancement in the test bolus measurement to determine the MR angiographic start delay (MR angiographic start delay was calculated by subtracting time to center from time of peak enhancement, all of which were in seconds). The delay timing was not adjusted at either field strength to reflect a potential later peak for the main bolus. For 1.5-T MR angiography, 20 mL of gadobenate dimeglumine was injected at a rate of 2 mL/sec. Because of the greater effectiveness of contrast agents at 3.0 T (13,14) and the shorter imaging time needed with 3.0-T MR angiography in our study, only 15 mL of gadobenate dimeglumine was administered at a rate of 2.5 mL/sec. Injections of contrast agent were followed by a 30-mL saline chaser at the same flow rate. All volunteer examinations were successfully performed without technical problems.
Phantom Measurements
Because the noise is not uniformly distributed when parallel imaging is used, conventional approaches for SNR measurement are not appropriate. For this reason, a method to objectively measure the SNR (15) was used: A cylindrical water phantom (Siemens Medical Solutions), with volume of approximately 9 L and T1 time of 275 msec, was measured 20 times in an identical position with no parallel imaging for the MR angiographic sequence and then again with an acceleration factor of three at 1.5 and 3.0 T (a total of 40 measurements were performed at 1.5 and 3.0 T). For all phantom measurements, a region of interest of 200 x 260 pixels was placed at the same location in the image center by one author (O.D.). The mean signal intensity of the 20 measurements was used as the signal intensity of the phantom, and the pixelwise calculated mean standard deviation of the 20 measurements was used to describe the noise. SNR was then calculated as follows: SNR = SImrep/NmSDrep, where SImrep is mean signal intensity of repeated measurements and NmSDrep is the mean standard deviation of the repeated measurements. Color-coded SNR
parameter maps were calculated. From these phantom measurements, the geometry factor was also determined for both field strengths. The geometry factor is a parameter describing the spatially inhomogeneous noise distribution in parallel imaging reconstructions.
Image Evaluation
The quality of the images obtained with the MR angiographic sequence was rated in consensus with a four-point scale by two radiologists (H.J.M., H.K.) who were blinded to the field strength used. The comparisons were performed in random order (1.5 or 3.0 T first) on 2 consecutive days by using the viewing interface of the system workstation (Leonardo; Siemens Medical Solutions). Both radiologists have 4 years of experience in vascular MR imaging. For this purpose, the source data were used. A five-point ordinal scale was applied: A score of 4 was very good (diagnostic image quality, with homogeneous enhancement of the aorta and of the proximal renal arteries); a score of 3, good (still diagnostic, with inhomogeneous enhancement of the aorta and of the proximal renal arteries); a score of 2, moderate (partly diagnostic, with inhomogeneous enhancement of the aorta and renal arteries only partly enhanced); a score of 1, poor (barely diagnostic, with partial enhancement of the aorta and no enhancement of the proximal renal arteries); and a score of 0, nondiagnostic (lacking enhancement of the aorta and of the proximal renal arteries). In addition, the visibility of the proximal and segmental renal arteries was rated by the two radiologists with a three-point scale: A score of 3 was good (completely visible proximal or segmental renal arteries); a score of 2, moderate (partly visible proximal or segmental renal arteries); and a score of 1, poor (nonvisible proximal and segmental renal arteries). Potential causes for degraded vessel visibility were disturbing noise; disturbing venous overlay; and presence of disturbing artifacts, which included parallel imaging reconstruction artifacts and ringing artifacts.
The radiologists then performed an intraindividual head-to-head comparison between the 1.5- and 3.0-T images, again in random order. In this head-to-head comparison, they were asked to determine whether one sequence was superior to the other or whether they were equal in producing images with the same quality.
Statistical Analysis
By using a statistical program (Power and Sample Size Calculations; http://biostat.mc.vanderbilt.edu/twiki/bin/view/Main/PowerSampleSize), a sample size estimation was performed for the minimum number of volunteers required based on the following assumptions: power of 90%, 
level of .5, estimated measured difference of 50%, and estimated standard deviation of the measured values of 40%. For these assumptions, a minimal sample size of 13 volunteers is required.
SNR values are given as the mean ± standard deviation and 95% confidence interval (CI). We performed t tests. A difference with a P value less than .05 was considered statistically significant. Because of the ordinal character of the visual assessment, the results of these assessments are given as medians. All statistical analyses were performed by using software (SPSS, version 13.0; SPSS, Chicago, Ill).
|
RESULTS
|
|---|
Geometry Factor and SNR
The mean geometry factors were 1.14 ± 0.11 at 1.5 T and 1.40 ± 0.16 at 3.0 T, were not uniformly distributed, and varied over the imaging section (Fig 1a). In the phantom measurements, a significantly (P < .001) higher mean SNR was found at 3.0 T without parallel imaging (35.0 ± 0.2 [95% CI: 34.9, 35.1]) and with a parallel imaging acceleration factor of three (17.8 ± 0.1 [95% CI: 17.8, 17.9]) compared with images obtained at 1.5 T without parallel imaging (19.2 ± 0.1 [95% CI: 19.2, 19.3]) and with a parallel imaging acceleration factor of three (11.9 ± 0.1 [95% CI: 11.9, 12.0]) (Fig 1b–1d).
View larger version (45K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1a: (a) Color-coded geometry factor map of the phantom measurements. At 3.0 T, a slightly higher geometry factor was measured. Because of the similar coil sensitivities in the center of the measurement volume, the highest geometry factor was found in the center of the phantom at both field strengths. (b, c) Color-coded SNR maps of the phantom measurements acquired at (b) 1.5 T and (c) 3.0 T. The SNR has been determined for acquisitions without parallel imaging (ie, acceleration factor [R], one; left side) and for acquisitions with parallel imaging (acceleration factor, three; right side). At both field strengths, the SNR decreases markedly from an acceleration factor of one to that of three. However, the SNR at 3.0 T with acceleration factor of three is still higher than that for the SNR at 1.5 T without parallel imaging (acceleration factor, one). (d) Bar chart demonstrates mean values and standard deviations of phantom measurements at 1.5 and 3.0 T without parallel imaging (acceleration factor, one) and parallel imaging (acceleration factor, three).
|
|
View larger version (60K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1b: (a) Color-coded geometry factor map of the phantom measurements. At 3.0 T, a slightly higher geometry factor was measured. Because of the similar coil sensitivities in the center of the measurement volume, the highest geometry factor was found in the center of the phantom at both field strengths. (b, c) Color-coded SNR maps of the phantom measurements acquired at (b) 1.5 T and (c) 3.0 T. The SNR has been determined for acquisitions without parallel imaging (ie, acceleration factor [R], one; left side) and for acquisitions with parallel imaging (acceleration factor, three; right side). At both field strengths, the SNR decreases markedly from an acceleration factor of one to that of three. However, the SNR at 3.0 T with acceleration factor of three is still higher than that for the SNR at 1.5 T without parallel imaging (acceleration factor, one). (d) Bar chart demonstrates mean values and standard deviations of phantom measurements at 1.5 and 3.0 T without parallel imaging (acceleration factor, one) and parallel imaging (acceleration factor, three).
|
|
View larger version (69K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1c: (a) Color-coded geometry factor map of the phantom measurements. At 3.0 T, a slightly higher geometry factor was measured. Because of the similar coil sensitivities in the center of the measurement volume, the highest geometry factor was found in the center of the phantom at both field strengths. (b, c) Color-coded SNR maps of the phantom measurements acquired at (b) 1.5 T and (c) 3.0 T. The SNR has been determined for acquisitions without parallel imaging (ie, acceleration factor [R], one; left side) and for acquisitions with parallel imaging (acceleration factor, three; right side). At both field strengths, the SNR decreases markedly from an acceleration factor of one to that of three. However, the SNR at 3.0 T with acceleration factor of three is still higher than that for the SNR at 1.5 T without parallel imaging (acceleration factor, one). (d) Bar chart demonstrates mean values and standard deviations of phantom measurements at 1.5 and 3.0 T without parallel imaging (acceleration factor, one) and parallel imaging (acceleration factor, three).
|
|
View larger version (20K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1d: (a) Color-coded geometry factor map of the phantom measurements. At 3.0 T, a slightly higher geometry factor was measured. Because of the similar coil sensitivities in the center of the measurement volume, the highest geometry factor was found in the center of the phantom at both field strengths. (b, c) Color-coded SNR maps of the phantom measurements acquired at (b) 1.5 T and (c) 3.0 T. The SNR has been determined for acquisitions without parallel imaging (ie, acceleration factor [R], one; left side) and for acquisitions with parallel imaging (acceleration factor, three; right side). At both field strengths, the SNR decreases markedly from an acceleration factor of one to that of three. However, the SNR at 3.0 T with acceleration factor of three is still higher than that for the SNR at 1.5 T without parallel imaging (acceleration factor, one). (d) Bar chart demonstrates mean values and standard deviations of phantom measurements at 1.5 and 3.0 T without parallel imaging (acceleration factor, one) and parallel imaging (acceleration factor, three).
|
|
Image Quality at 1.5 versus 3.0 T
The median image quality was rated as 4 (ie, very good) for both field strengths (Fig 2). No difference in the number and quality of artifacts was seen between the two groups. The conspicuity of small vessels, such as the lumbar arteries or the branches of the liver arteries, was better at 3.0 T (Figs 2 and 3). The depiction of the proximal renal arteries was rated as good in all studies at both field strengths. The median rating for depiction of the distal renal arteries was 3 (ie, good) at 3.0 T compared with 2 (ie, moderate) at 1.5 T. The distal renal arteries were more often found to reveal blurring or only moderate opacification at 1.5 T (Fig 3).
View larger version (152K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a: (a, b) Coronal 40-mm-thin maximum intensity projections of MR angiograms in a volunteer that were acquired at (a) 1.5 T (3.77/1.39; flip angle, 25°) and (b) 3.0 T (3.14/1.11; flip angle, 23°) in a comparable position. The overall image quality is very good with both techniques. Closer examination of vessels with a small caliber, such as the branches of the hepatic artery (arrow), one branch of the inferior mesenteric artery (large arrowhead), or the lumbar arteries (small arrowheads), reveals either better conspicuity at 3.0 T or lack of visibility at 1.5 T. (c, d) The source images in this 30-year-old male volunteer reveal intense opacification of the distal branches of the renal artery (d) at 3.0 T, which enhanced more than (c) at 1.5 T. The vessel borders are also less clearly defined at 1.5 T than at 3.0 T on comparable source image positions. Note that the renal parenchymal enhancement seems to be equal in both examinations.
|
|
View larger version (130K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b: (a, b) Coronal 40-mm-thin maximum intensity projections of MR angiograms in a volunteer that were acquired at (a) 1.5 T (3.77/1.39; flip angle, 25°) and (b) 3.0 T (3.14/1.11; flip angle, 23°) in a comparable position. The overall image quality is very good with both techniques. Closer examination of vessels with a small caliber, such as the branches of the hepatic artery (arrow), one branch of the inferior mesenteric artery (large arrowhead), or the lumbar arteries (small arrowheads), reveals either better conspicuity at 3.0 T or lack of visibility at 1.5 T. (c, d) The source images in this 30-year-old male volunteer reveal intense opacification of the distal branches of the renal artery (d) at 3.0 T, which enhanced more than (c) at 1.5 T. The vessel borders are also less clearly defined at 1.5 T than at 3.0 T on comparable source image positions. Note that the renal parenchymal enhancement seems to be equal in both examinations.
|
|
View larger version (134K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2c: (a, b) Coronal 40-mm-thin maximum intensity projections of MR angiograms in a volunteer that were acquired at (a) 1.5 T (3.77/1.39; flip angle, 25°) and (b) 3.0 T (3.14/1.11; flip angle, 23°) in a comparable position. The overall image quality is very good with both techniques. Closer examination of vessels with a small caliber, such as the branches of the hepatic artery (arrow), one branch of the inferior mesenteric artery (large arrowhead), or the lumbar arteries (small arrowheads), reveals either better conspicuity at 3.0 T or lack of visibility at 1.5 T. (c, d) The source images in this 30-year-old male volunteer reveal intense opacification of the distal branches of the renal artery (d) at 3.0 T, which enhanced more than (c) at 1.5 T. The vessel borders are also less clearly defined at 1.5 T than at 3.0 T on comparable source image positions. Note that the renal parenchymal enhancement seems to be equal in both examinations.
|
|
View larger version (132K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2d: (a, b) Coronal 40-mm-thin maximum intensity projections of MR angiograms in a volunteer that were acquired at (a) 1.5 T (3.77/1.39; flip angle, 25°) and (b) 3.0 T (3.14/1.11; flip angle, 23°) in a comparable position. The overall image quality is very good with both techniques. Closer examination of vessels with a small caliber, such as the branches of the hepatic artery (arrow), one branch of the inferior mesenteric artery (large arrowhead), or the lumbar arteries (small arrowheads), reveals either better conspicuity at 3.0 T or lack of visibility at 1.5 T. (c, d) The source images in this 30-year-old male volunteer reveal intense opacification of the distal branches of the renal artery (d) at 3.0 T, which enhanced more than (c) at 1.5 T. The vessel borders are also less clearly defined at 1.5 T than at 3.0 T on comparable source image positions. Note that the renal parenchymal enhancement seems to be equal in both examinations.
|
|
View larger version (110K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3a: (a, b) Coronal 20-mm-thin maximum intensity projections and source data acquired at (a) 1.5 T (3.77/1.39; flip angle, 25°) and (b) 3.0 T (3.14/1.11; flip angle, 23°) in a 34-year-old male volunteer. In the maximum intensity projection images, some motion-induced haziness, particularly of the accessory renal artery (arrows), can be appreciated at 1.5 T, whereas the vessel is clearly depicted at 3.0 T.
|
|
View larger version (97K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3b: (a, b) Coronal 20-mm-thin maximum intensity projections and source data acquired at (a) 1.5 T (3.77/1.39; flip angle, 25°) and (b) 3.0 T (3.14/1.11; flip angle, 23°) in a 34-year-old male volunteer. In the maximum intensity projection images, some motion-induced haziness, particularly of the accessory renal artery (arrows), can be appreciated at 1.5 T, whereas the vessel is clearly depicted at 3.0 T.
|
|
In the direct comparison, 3.0-T MR angiography was rated better than 1.5-T imaging in 14 of the 15 volunteer examinations and as equal in the other study.
|
DISCUSSION
|
|---|
To our knowledge, our study is the first to demonstrate, in an intraindividual comparison, the superiority of 3.0-T abdominal MR angiography over 1.5-T abdominal MR angiography. With current MR imaging units, a high image quality can also be achieved at 1.5 T, as has been reported in previous publications (6,16,17) and has been confirmed by results of blinded reading in our study. Further increasing the spatial resolution or decreasing the acquisition time by increasing the readout bandwidth is limited by the available SNR at 1.5 T, so that image quality is increasingly degraded by noise. With our sequence parameters, there was visible image noise at 1.5 T.
Extended image acquisition times for an increased SNR, as feasible with dedicated techniques such as contrast-enhanced timing-robust angiography (CENTRA; Philips, Eindhoven, the Netherlands) (18), may be less affected by motion-introduced artifacts. Higher factors of parallel imaging would lead to an additional loss of signal and also require dedicated coils with a greater number of simultaneously available receiver elements (19).
In contrast, high-field-strength imaging at 3.0 T provides SNR sufficient for further improvement of MR angiographic technique. In our study, we chose a higher readout bandwidth to shorten repetition time and echo time and thereby achieve faster image acquisition, which led to a reduction of the SNR by 18%. The field of view was also reduced, and the matrix was slightly increased to improve the in-plane spatial resolution, which again led to a decrease in the SNR by 6%. Finally, the decreased section thickness led to an 8% decrease in SNR. On the other hand, by adding phase oversampling and by increasing the number of partitions, the SNR was increased by 13%. Because the SNR is proportional to the voxel volume and the inverse square root of the readout bandwidth, these modifications effectively decreased the maximally achievable SNR by about 20%. Hence, a 60% increase could be theoretically predicted for 3.0-T MR angiography. The observed difference in SNR with the use of parallel imaging was 50%. This deviation from the theoretically achievable value can be explained by imperfect measurements and the higher geometry factor we measured at 3.0 T.
It is well known that the relaxivity of all contrast agents decreases with increasing field strength (11). In our study, we used gadobenate dimeglumine to counteract the roughly 40% reduction in SNR with the application of a parallel imaging acceleration factor of three. This agent is characterized by a higher relaxivity compared with standard gadobenate chelates and hence was shown to yield better enhancement and a higher contrast-to-noise ratio (20). The relaxivity of gadobenate dimeglumine is as follows: r1 = 6.3 L·mmol–1·sec–1 at 1.5 T and r1 = 5.5 L·mmol–1·sec–1 at 3.0 T (11). These relaxivity values should actually lead to weaker enhancement. At the same time, the T1 values of muscle and liver, for example, increase between 5% and 38% with the transition to 3.0 T (10); thus, the ratio of the signal of contrast-enhanced blood to the signal of fat is increased.
The prolonged T1 times at 3.0 T facilitate the background suppression for fast sequences such as MR angiography and, hence, increase the effectiveness of contrast agents, eventually allowing a 20%–30% reduction in contrast agent dose (13,14). Because of the chosen opposed-phase echo time of 1.1 msec at 3.0 T, additional background suppression may be suspected even though this phenomenon was not further investigated in our study.
Unlike the protocol for common abdominal MR angiographic applications, which involve bolus injection rates of 1–2 mL/sec, we decided to increase the injection rate and to decrease the volume of the contrast agent for two reasons. First, we expected that the increased effectiveness of the contrast agent at 3.0 T would allow a decrease in the dose. Second, in previous MR angiographic examinations that involved an infusion rate of 2 mL/sec, there was still enhancement of the right atrium and ventricle. With an injection rate of 2.5 mL/sec, a 15-mL bolus lasts 6 seconds. After the pulmonary passage, it can be assumed to last 12–13 seconds. With the Cartesian k-space trajectories used in our study and with a total acquisition time of 18 seconds, the center of the k-space is reached 7 seconds after the start of the sequence.
From these data, the optimal length of the bolus can be calculated to be 11 seconds. Of course, such bolus timing and the high injection speed are prone to errors, particularly with the test-bolus technique. However, our results show that, with correct execution of the bolus-timing calculation, very good results can be achieved. Cartesian k-space sampling is also known to be very robust in terms of timing errors (21). In none of our volunteers was substantial enhancement of the right atrium or ventricle seen; thus, better use was made of the injected contrast agent. The faster injection rate did not lead to an increased number of venous artifacts. Because of the short acquisition time of only 18 seconds, high-spatial-resolution MR angiography would also be suitable for fluoro-triggering (22).
The good opacification, particularly of the distal parts of small-caliber arteries, that we could observe in our study is partly a consequence of the compact bolus of contrast agent. Other factors are, of course, the higher SNR at 3.0 T and the smaller voxel size, which should reduce partial volume effects. At 3.0 T, the Lamor frequencies of fat and water are not as far apart, and this closeness results in altered in-phase and opposed-phase times. Even though the echo times were only slightly altered compared with 1.5 T, we could detect a slightly more pronounced chemical shift artifact. However, this artifact was not disturbing.
Our study had some limitations. First, we examined only volunteers who were cooperative and whose circulation parameters were normal. These characteristics represent a best-case scenario with which to prove the feasibility of the method. Of course, the crucial question to be answered is whether this improved MR angiographic sequence leads to improved detection of renal arterial lesions in patients. Second, the improved image quality cannot be attributed to the higher field strength alone. One has to take into account the different coils and geometry factors and the purposely altered administration of the contrast agent. Third, because the measurement of SNR and contrast-to-noise ratio in parallel imaging is not reliably feasible with conventional approaches, we performed repeated phantom measurements. These measurements do not, however, allow measurement of the contrast-to-noise ratio after the injection of contrast agent. Last, the examinations were performed at end inspiration, a position in which the diaphragm is more susceptible to motion than at end expiration.
In conclusion, our study demonstrates the superiority of abdominal MR angiography performed at 3.0 T compared with state-of-the-art MR angiography performed at 1.5 T. Abdominal MR angiography at 3.0 T is feasible and provides better image contrast, higher spatial resolution, and broader coverage in a shorter imaging time than does MR angiography performed at 1.5 T.
|
ADVANCES IN KNOWLEDGE
|
|---|
- Even with smaller voxel sizes, shortened imaging times, and improved spatial coverage, 3.0-T abdominal MR angiography is superior to 1.5-T abdominal MR angiography.
- Because of the greater effectiveness of the contrast agent at 3.0 T, the contrast agent dose can be reduced by 25% compared with that used for 1.5-T imaging, with good to very good image quality maintained.
|
IMPLICATIONS FOR PATIENT CARE
|
|---|
- Abdominal MR angiographic examinations performed at 3.0 T in healthy volunteers were superior to those performed at 1.5 T in terms of vessel depiction and conspicuity and therefore have potential for preferential use in patient care.
- At 3.0 T, abdominal MR angiography can be performed with at least a 25% reduction in the contrast agent dose without loss of image quality or vessel contrast.
|
ACKNOWLEDGMENTS
|
|---|
Bracco-Altana Pharma provided volunteer insurance and contrast agent, and Verein MagnetresonanzForschung provided the measurement slots at the 3.0-T imaging unit.
|
FOOTNOTES
|
|---|
Abbreviations: CI = confidence interval • SNR = signal-to-noise ratio • 3D = three-dimensional
Author contributions:Guarantors of integrity of entire study, H.J.M., S.O.S.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, H.J.M., O.D., K.N., M.F.R., S.O.S.; clinical studies, H.J.M., H.K., K.P.L., S.O.S.; experimental studies, O.D.; statistical analysis, H.J.M., O.D., K.N., M.F.R., S.O.S.; and manuscript editing, all authors
See Materials and Methods for pertinent disclosures.
|
References
|
|---|
- Leiner T. Magnetic resonance angiography of abdominal and lower extremity vasculature. Top Magn Reson Imaging 2005;16:21–66.[CrossRef][Medline]
- Schoenberg SO, Rieger J, Nittka M, Dietrich O, Johannson LO, Reiser MF. Renal MR angiography: current debates and developments in imaging of renal artery stenosis. Semin Ultrasound CT MR 2003;24:255–267.[CrossRef][Medline]
- Michaely HJ, Schoenberg SO, Rieger JR, Reiser MF. MR angiography in patients with renal disease. Magn Reson Imaging Clin North Am 2005;13:131–151.[CrossRef]
- Vasbinder GB, Nelemans PJ, Kessels AG, et al. Accuracy of computed tomographic angiography and magnetic resonance angiography for diagnosing renal artery stenosis. Ann Intern Med 2004;141:674–682.[Abstract/Free Full Text]
- Hoogeveen RM, Bakker CJ, Viergever MA. Limits to the accuracy of vessel diameter measurement in MR angiography. J Magn Reson Imaging 1998;8:1228–1235.[Medline]
- Schoenberg SO, Rieger J, Weber CH, et al. High-spatial-resolution MR angiography of renal arteries with integrated parallel acquisitions: comparison with digital subtraction angiography and US. Radiology 2005;235:687–698.[Abstract/Free Full Text]
- Vasbinder GB, Maki JH, Nijenhuis RJ, et al. Motion of the distal renal artery during three-dimensional contrast-enhanced breath-hold MRA. J Magn Reson Imaging 2002;16:685–696.[CrossRef][Medline]
- Slovut DP, Olin JW. Fibromuscular dysplasia. N Engl J Med 2004;350:1862–1871.[Free Full Text]
- Campeau NG, Huston J 3rd, Bernstein MA, Lin C, Gibbs GF. Magnetic resonance angiography at 3.0 Tesla: initial clinical experience. Top Magn Reson Imaging 2001;12:183–204.[CrossRef][Medline]
- de Bazelaire CM, Duhamel GD, Rofsky NM, Alsop DC. MR imaging relaxation times of abdominal and pelvic tissues measured in vivo at 3.0 T: preliminary results. Radiology 2004; 230: 652–659.[Abstract/Free Full Text]
- Rohrer M, Bauer H, Mintorovitch J, Requardt M, Weinmann HJ. Comparison of magnetic properties of MRI contrast media solutions at different magnetic field strengths. Invest Radiol 2005;40:715–724.[CrossRef][Medline]
- Michaely HJ, Nael K, Schoenberg SO, et al. The feasibility of spatial high-resolution magnetic resonance angiography (MRA) of the renal arteries at 3.0 T. Rofo 2005;177:800–804.[Medline]
- Krautmacher C, Willinek WA, Tschampa HJ, et al. Brain tumors: full- and half-dose contrast-enhanced MR imaging at 3.0 T compared with 1.5 T—initial experience. Radiology 2005;237:1014–1019.[Abstract/Free Full Text]
- Araoz PA, Glockner JF, McGee KP, et al. 3 Tesla MR imaging provides improved contrast in first-pass myocardial perfusion imaging over a range of gadolinium doses. J Cardiovasc Magn Reson 2005;7:559–564.[Medline]
- Zech CJ, Herrmann KA, Huber A, et al. High-resolution MR-imaging of the liver with T2-weighted sequences using integrated parallel imaging: comparison of prospective motion correction and respiratory triggering. J Magn Reson Imaging 2004;20:443–450.[CrossRef][Medline]
- Lee VS, Rofsky NM, Krinsky GA, Stemerman DH, Weinreb JC. Single-dose breath-hold gadolinium-enhanced three-dimensional MR angiography of the renal arteries. Radiology 1999;211:69–78.[Abstract/Free Full Text]
- Nael K, Laub G, Finn JP. Three-dimensional contrast-enhanced MR angiography of the thoraco-abdominal vessels. Magn Reson Imaging Clin N Am 2005;13:359–380.[CrossRef][Medline]
- Willinek WA, Gieseke J, Conrad R, et al. Randomly segmented central k-space ordering in high-spatial-resolution contrast-enhanced MR angiography of the supraaortic arteries: initial experience. Radiology 2002;225:583–588.[Abstract/Free Full Text]
- Chen Q, Quijano CV, Mai VM, et al. On improving temporal and spatial resolution of 3D contrast-enhanced body MR angiography with parallel imaging. Radiology 2004;231:893–899.[Abstract/Free Full Text]
- Goyen M, Debatin JF. Gadobenate dimeglumine (MultiHance) for magnetic resonance angiography: review of the literature. Eur Radiol 2003; 13(suppl 3): N19–N27.[CrossRef][Medline]
- Schmidt MA, Britten AJ, Tomlinson MA, et al. Contrast-enhanced magnetic resonance angiography of the iliac arteries: optimization by injection simulation. Magn Reson Med 2001;46:365–373.[CrossRef][Medline]
- Wilman AH, Riederer SJ, King BF, Debbins JP, Rossman PJ, Ehman RL. Fluoroscopically triggered contrast-enhanced three-dimensional MR angiography with elliptical centric view order: application to the renal arteries. Radiology 1997;205:137–146.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
I. C. van den Bos, S. M. Hussain, G. P. Krestin, and P. A. Wielopolski
Liver Imaging at 3.0 T: Diffusion-induced Black-Blood Echo-planar Imaging with Large Anatomic Volumetric Coverage as an Alternative for Specific Absorption Rate-intensive Echo-Train Spin-Echo Sequences: Feasibility Study
Radiology,
July 1, 2008;
248(1):
264 - 271.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
C. K. Kuhl, F. Traber, and H. H. Schild
Whole-Body High-Field-Strength (3.0-T) MR Imaging in Clinical Practice * Part I. Technical Considerations and Clinical Applications
Radiology,
March 1, 2008;
246(3):
675 - 696.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
