Current State-of-the-Art 1.5 T and 3 T Extracranial Carotid Contrast-Enhanced Magnetic Resonance Angiography




Recent advances in magnetic resonance (MR) hardware and software have improved the resolution and spatial coverage of head and neck first-pass contrast-enhanced (CE) MR angiography. Despite these improvements, high-quality submillimeter-resolution 1.5 T and 3 T carotid CE MR angiography is not consistently available in the general radiology practice. This article reviews the important imaging parameters and potential pitfalls that affect carotid CE MR angiography image quality, and the dose and timing of the gadolinium-based contrast agent, and summarizes vendor-specific protocols for high-quality submillimeter-resolution carotid CE MR angiography at 1.5 and 3 T.








  • Time-resolved carotid MR angiography with a small test bolus of GBCA provides:




    • A good estimation of the GBCA arrival for accurate timing of first-pass CE MR angiography



    • Useful dynamic vascular assessment and functional information in the head and neck region




  • Consistently, good-quality, high-resolution, first-pass CE carotid MR angiography is possible; however:




    • At 1.5 T, strive for resolution of 0.71 mm 3 or less in carotid CE MR angiography



    • At 3 T, strive for resolution of 0.54 mm 3 or less in carotid CE MR angiography, perhaps with decreased GBCA




  • Be aware of potential pitfalls when developing and implementing carotid CE MR angiography protocols:




    • Understand how phase resolution and slice resolution are displayed by your MR manufacturer



    • Understand the advantages/disadvantages of methods to acceleration CE MR angiography:




      • Partial k y /k z and percent sampling



      • Parallel imaging



      • Bandwidth





Key Points






  • Use time-resolved magnetic resonance (MR) angiography with a small test bolus of gadolinium-based contrast agent (GBCA) to identify the individual bolus arrival time and obtain additional useful information about vascular kinetics in the head and neck.



  • With attention to specific imaging parameters, resolution of 0.7 mm 3 or less can be obtained consistently with first-pass contrast-enhanced (CE), MR angiography of the head and neck at 1.5 T with even higher spatial resolution possible with a smaller dose of GBCA at 3 T.



  • Be aware of pitfalls in prescribing first-pass CE carotid MR angiography, as well as the potentially confusing annotations, especially regarding resolution in the phase direction and slice thickness with three-dimensional (3D) sequences.



  • Strong collaboration between the MR technologist and radiologist in understanding the specific techniques used by each MR manufacturer when prescribing CE carotid MR angiography can lead to consistently good image quality.



Diagnostic Checklist
The American College of Cardiology Foundation (ACCF) and the American Heart Association (AHA) have jointly written new guidelines on the management of patients with extracranial carotid and vertebral artery disease. The writing committee assigned by the ACCF-AHA Task Force on Practice Guidelines has developed several recommendations based on expert review of evidence-based methodologies. In this document, the Task Force specifically recommends the use of MR angiography or computed tomography (CT) angiography to detect carotid stenosis when sonography either cannot be obtained or yields equivocal or otherwise nondiagnostic results. When intervention for significant carotid stenosis detected by sonography is planned, the Task Force also concluded that MR angiography, CT angiography, or catheter-based contrast angiography could be useful to evaluate the severity of carotid stenosis. It is clear from these recommendations that high-quality carotid MR angiography that can accurately and reproducibly detect extracranial carotid stenosis is necessary. This article reviews the technical factors necessary to obtain high-quality extracranial carotid MR angiography.


In the 1990s, time-of-flight (TOF) MR angiography was reported to have good sensitivity and specificity in detecting internal carotid artery (ICA) stenosis greater than 70% using North American Symptomatic Carotid Endarterectomy Trial (NASCET) criteria identified on digital subtraction angiography. CE MR angiography offered the opportunity to cover more of the carotid artery distribution in a fraction of the time that TOF MR angiography required. There has been much work presented on various techniques for time-resolved extracranial carotid CE MR angiography. Although there are new techniques that hold much promise for improving spatial resolution of time-resolved CE MR angiography while maintaining adequate temporal resolution, currently available clinical time-resolved CE MR angiography techniques cannot achieve the spatial resolution of less than 1 mm that is probably necessary for accurate depiction of extracranial severe carotid stenosis. The advent of elliptical-centric phase reordering and effective timing of the gadolinium contrast bolus moved first-pass CE MR angiography from research into routine clinical practice at the Mayo Clinic. Other workers reported success with breath-held CE MR angiography, which included the great vessel origins at the aortic arch. Some, but not all, of the literature from 2000 to 2005 concluded that first-pass carotid CE MR angiography is superior to TOF MR angiography in the detection of severe extracranial carotid stenosis. This may in part be because of the effect of spatial resolution of extracranial carotid CE MR angiography. In the past 5 years, as submillimeter-resolution first-pass CE MR angiography has become more routine, a general consensus has emerged that CE MR angiography is accurate in the measurement of extracranial carotid stenosis. In this article, the discussion of carotid CE MR angiography is restricted to the first-pass extracranial carotid CE MR angiography technique. Improvements in the technique that have allowed the routine clinical acquisition of submillimeter-resolution carotid CE MR angiography are reviewed. Specifically, the imaging parameter details that provide submillimeter resolution while speeding up the CE MR angiography acquisition are discussed and both the amount and type of GBCA that is required to support such high-resolution CE MR angiography acquisitions are reviewed, taking into account concerns about nephrogenic systemic fibrosis (NSF) in susceptible populations. Various techniques for adequate timing of the GBCA arrival are also reviewed. In addition, vendor-specific protocols at both 1.5 T and 3 T MR are provided to help obtain consistently high-quality extracranial carotid CE MR angiography.


Resolution and scan time


Hinatiuk and colleagues recently reviewed the effect of increased spatial resolution in depicting carotid stenosis as seen on CE MR angiography at 1.5 T. In patients with carotid artery stenosis, decreasing the voxel volume from 0.9 mm 3 to 0.53 mm 3 by increasing the scan matrix and while keeping the field of view (FOV) constant caused the scan time to increase from 21 to 40 seconds. The resulting extracranial carotid CE MR angiography with improved resolution resulted in sharper depiction of the carotid stenosis. With modern gradient systems, repetition times (TR) for 3D gradient-echo acquisitions are approximately half what they were during the early work on carotid CE MR angiography. The investigators made use of the 50% reduction in TR to nearly double the spatial resolution of elliptical-centric carotid CE MR angiography compared with the initial 0.8-mm 3 to 1.0-mm 3 voxel size while maintaining an imaging time of 40 seconds. This reduction in voxel volume to 0.53 mm 3 resulted in a better depiction of the carotid stenosis. Further decreases in voxel volume at 1.5 T, by extending the acquisition time to 50 to 60 seconds, did not improve the vessel depiction because of both a reduction in signal-to-noise ratio (SNR) and sharpness losses, possibly from motion. Recently, the resolution of large FOV first-pass carotid CE MR angiography has been extended to 0.44 to 0.49 mm 3 by making use of the extra SNR at 3 T. Further reductions in voxel size to 0.28 mm 3 have been made possible by combining dedicated carotid surface coils with 3 T MR imaging.


With modern MR scanners and high-performing gradients, it is possible to achieve a resolution of 1 mm or less in all 3 primary axes with first-pass CE MR angiography. Despite the plethora of articles that show the usefulness of first-pass CE MR angiography with a resolution of 1 mm or less, many centers do not routinely obtain this high resolution. One reason may be confusion about the multiple imaging parameters that affect the final CE MR angiography resolution. This article therefore discusses the potential pitfalls and sources of confusion when attempting to prescribe a CE MR angiography protocol with a resolution of 1 mm or less.


Percent Phase FOV


What happens when a partial or asymmetrical FOV is prescribed? In our example of coronally acquired CE MR angiography, we typically use an FOV of 28 to 40 cm in the superior-inferior direction, which is also the direction of the frequency-encoding gradient. Rarely is such a large FOV needed in the right-left direction. Most modern large FOV clamshell-style neurovascular coils or integrated head and neck coils have the ability to turn off elements far laterally near the shoulders. The wraparound artifact from the shoulders is thereby diminished so that typically only an FOV of around 20 to 27 cm is needed, depending on what other acceleration factors such as parallel imaging are being used. Take for example 28-cm FOV, CE, MR angiography with a prescribed matrix of 384 × 220. If a 0.8 rectangular FOV (80% phase FOV) is chosen, the FOV in the y direction (right-left) is decreased to 22.4 cm and the number of phase-encoding steps is decreased to 176. This choice diminishes the total acquisition time, but increases the size of the voxel in the y direction to greater than 1-mm resolution. Depending on the MR manufacturer, the annotation of the final image may say 28 × 22 cm FOV with 384 × 220 matrix. failure to understand what a specific MR vendor is acquiring may result in failure to achieve the desired resolution. This annotation could be interpreted to mean that an acquired matrix of 0.73 × 1.0 mm has been obtained, which would not be true. To maintain resolution at 1 mm or less in the example, a 28-cm FOV with a matrix of 384 × 288 may be needed. Using a 0.8 rectangular FOV, the MR scanner will then obtain a 28 × 22 cm FOV, CE, MR angiogram with an acquired matrix of 384 × 220 with a pixel size of 0.73 × 0.97 mm while decreasing scan time by 20% compared with the scan time without using an asymmetric or partial FOV. The specifics of 3 MR manufactures (GE, Philips, Siemens) are reviewed later in this article to detail with screen shots from their individual software interfaces to help decrease any potential for confusion or misunderstanding.


Partial k y /k z and Percent Sampling


Another way the radiologist can speed up the scan time of CE MR angiography and maintain resolution at 1 mm or less is to use partial Fourier schemes which enables faster imaging but at the expense of SNR. Only a portion of k-space in the phase encoding direction (ky) or in the slice direction (kz) can be obtained. In addition, only a partial number of averages can be acquired to speed up image acquisition. This technique takes advantage of the inherent symmetry in k-space. Instead of obtaining all of the phase-encoding lines in k-space in either/both the y direction or z direction, most of them can be obtained and technique called conjugate synthesis used to intelligently guess what the remaining lines would have looked like. Number of averages is also called number of excitations (NEX), and represents the number of times the same lines of k-space are acquired, which is done to increase the SNR or to average out motion artifact. When the MR technologist requests a partial NEX, it is the same as doing partial k y . Instead of not obtaining specific sections of k y or k z , some MR manufactures allow radiologists to not obtain the corners of 3D k-space. This method is sometimes called the k-space filter or percent sampling. The potential effects on the resolution of the lumen of vessels coursing obliquely through the 3D imaging slab when applying either/both partial k y and k z or cutting the corners of 3D k-space is beyond the scope of this article. Suffice it to say that, if the corners are not cut too much, good image quality can be maintained. Many investigators suggest that, as long as at least 75% to 80% of k y and k z are obtained, or at least 75% to 80% sampling, any deleterious effects on the depiction of the lumen are probably minimal and acceptable. The radiologist therefore needs to know what values of partial k y and k z or percent sampling are being using in the CE MR angiography prescription. The specific recommendations for GE, Philips, and Siemens are discussed later.


Zero Interpolation


Zero interpolation means expanding the k-space matrix beyond what was acquired by assigning zeroes to the outer (nonacquired) parts, before performing image reconstruction by the Fourier transform. This technique was originally described in 1994 as a way to decrease partial volume artifacts, especially in small vessels in MR angiography. It can improve image quality, especially if the acquired k-space data are asymmetric, and it can be applied in all 3 primary axes. A version of this technique has been applied by most MR manufactures. Again, it is important for radiologists to understand whether the matrix values displayed on their MR angiography images are the acquired resolution or the interpolated resolution. Although zero interpolation helps decrease partial volume artifacts and is useful, when discussing the optimal voxel resolution for CE MR angiography, this article refers specifically to the acquired resolution. The potential for misunderstanding zero interpolation is especially a problem when reviewing the number of slices and slice thickness of a CE MR angiography. Many MR manufactures display the number of interpolated coronal slices and the spacing between these interpolated slices. For instance, if a 60-mm thick coronal CE MR angiography slab in the anterior-posterior (AP) direction is prescribed and 44 slices are requested, a set of images each with 1.4-mm thickness is obtained. The annotation can read that 88 slices were obtained with a thickness of 0.7 mm, which represented the interpolated number of slices and the distance between each slice. Although this is accurate, it can lead radiologists to interpret that they have achieved a resolution of 1 mm or less. In this example, the radiologist needs to prescribe at least 60 slices to achieve 1 mm or better resolution in the slice direction with a 60-mm thick CE MR angiography coronal slab acquisition.


Parallel Imaging


Parallel imaging is a major advance in CE MR angiography. It has allowed acquisition to speed up and helped support higher acquired matrices with better resolution. Parallel imaging techniques take advantage of the signal originating from a given location in the body being detected with different sensitivity by various surface coil elements, depending on where the elements are located relative to the structure generating the signal. In this way, surface coils contain spatial information that is used to substitute for lines in k-space, saving time. Depending on the specific arrangement of the coils and patient anatomy, it might be possible to skip every second line, 2 out of every 3 lines, 3 out of every 4 lines, or even more. By skipping lines in k-space, the distance between adjacent acquired lines is increased (by a factor equal to the acceleration factor), and this effectively reduces the FOV by the same factor. An acceleration factor of 4 in a specified phase-encoding direction reduces the FOV in that direction by a factor of 4. Without accurate surface coil calibration, the resulting wraparound artifact would be disastrous, but accurate coil sensitivity information can be used to completely unwrap the signal. The price that is paid for the accelerated performance is in SNR. As discussed later, there are ways to mitigate the SNR loss from parallel imaging by judicious use of GBCA as well as accurate timing of the GBCA bolus arrival. There are 2 main types of parallel imaging: sensitivity encoding (SENSE) and SMASH (SiMultaneous Acquisition of Spatial Harmonics). These original sequence names morphed into a plethora of specific names as they were adopted for use by most MR manufactures. Variants of 1 or both are used in modern CE MR angiography sequences by MR manufactures. SENSE techniques acquire a separate scan to obtain sensitivity maps for each coil in the FOV and the unwrapping is performed in the image domain (after image reconstruction), whereas SMASH techniques have a calibration scan built into the CE MR angiography sequence and the unwrapping is performed in the k-space domain (before image reconstruction). For any parallel imaging technique to work well, whether image space–based or k-space–based, coils separated in space along the acceleration direction are needed. In the typical 1-directional (1D) acceleration of CE MR angiography using parallel imaging, the right-left direction is used. Multiple coils are needed along the right and left lateral aspects of the head and neck. If 2 direction (2D) acceleration of CE MR angiography with parallel imaging is required, multiple coils along the AP direction and right-left directions are needed. Many MR manufactures now have new multichannel large FOV coils designed to do this. The better-designed coils minimize the signal loss caused by the position of these various elements. The signal loss can be quantified by measuring the geometry factor (G factor), also known as the noise amplification factor. The optimal 1D or 2D acceleration with parallel imaging depends on many factors, such as the coil design, MR field strength, and gradient performance. Specific recommendations for GE, Philips, and Siemens at 1.5 T and 3 T are given later in this article.


Bandwidth


Another way to speed up the CE MR angiography acquisition while maintaining a resolution of 1 mm or less is to increase the bandwidth (BW). Increasing BW allows shorter repetition time (TR) and shorter echo times (TE). For conventional MR imaging, increasing bandwidth is associated with lower SNR but, for CE MR angiography, this is generally not the case. The added signal available with very short echo times and with condensing the acquisition window into the peak of the contrast bolus can, up to a point, compensate for SNR lost with increased bandwidth. As with parallel imaging, judicious increases in BW can decrease total acquisition time of CE MR angiography while supporting resolution of 1 mm or less with acceptable overall SNR and image quality. In general, higher BW along with more aggressive parallel imaging acceleration is possible at 3 T compared with 1.5 T because of the inherently higher SNR at 3 T. The shorter TR and higher acceleration factors allow more phase encoding in the y direction and z direction as well as more frequency-encoding steps, which results in higher spatial resolution and smaller voxel sizes with CE MR angiography at 3 T MR compared with 1.5 T MR scans.




GBCA


Now that the various technical components to obtaining CE MR angiography with resolution of 1 mm or less have been introduced, the GBCA should be discussed. Workers differ in their approach to how best to coordinate image acquisition with the GBCA contrast bolus. An ideal image acquisition would take advantage of the first pass encompassing the high arterial concentration of GBCA There is more debate about how important the later phases of GBCA distribution are to quality CE MR angiography. Carotid CE MR angiography can be extended to take advantage of the large residual of the initial bolus of contrast injection caused by recirculation. The blood-brain barrier and rapid return of gadolinium contrast through the brain parenchyma and jugular vein back to the heart makes carotid CE MR angiography especially capable of longer acquisitions and potentially higher resolutions. By modeling the elliptical-centric technique, Fain and colleagues showed that there is sufficient contrast for high resolution with increased acquisition time of carotid CE MR angiography. Initial experience with the use of elliptical-centric phase reordering in combination with a bolus arrival scan or fluoroscopic triggering allowed carotid CE MR angiography with voxel size of 1 mm 3 or less with good intra-arterial contrast and little venous contamination. Other workers have focused exclusively on the first arterial peak within a breath-hold acquisition at 1.5 T, also with excellent results.


As with most MR sequence parameters, the choice of GBCA agent and dose varies by MR field strength. In the past 10 years, most investigators achieving CE MR angiography with the desired resolution of 1 mm or less at 1.5 T have used more than a single dose of a variety of the original extracellular GBCA. A high T1 relaxivity GBCA, gadobenate dimeglumine (Multihance, Bracco Diagnostic, Inc.), was recently introduced. Intraindividual crossover comparison of 2 GBCAs showed superior depiction of arterial lumen with a single dose of gadobenate dimeglumine (Multihance, Bracco Diagnostic, Inc.) compared with gadopentetate dimeglumine (Magnevist, Bayer HealthCare Pharmaceuticals) at 1.5 T. Recent work showed that a single dose of gadobenate dimeglumine (Multihance, Bracco Diagnostic, Inc.) can provide similar image quality of CE MR angiography compared with a double dose of other extracellular GBCAs. The use of high T1 relativity GBCA such as gadobenate dimeglumine (Multihance, Bracco Diagnostic, Inc.), gadofosveset (Ablavar, Lantheus Medical Imaging) and gadobutrol (Gadavist, Bayer HealthCare Pharmaceuticals) helps to minimize the total dose of GBCA administration while supporting sufficient SNR to produce carotid CE MR angiography with high image quality. Typically, a single dose (0.1 mmol/kg) of gadobenate dimeglumine is sufficient to maintain high SNR and image quality with a resolution of 0.8-mm 3 or less for carotid CE MR angiography at 1.5 T.


At 3 T, similar intraindividual crossover studies have shown the superiority of gadobenate dimeglumine (Multihance, Bracco Diagnostic, Inc.) compared with other extracellular GBCA in CE MR angiography when using a single dose. Although the initial work on carotid CE MR angiography at 3 T MR used a double dose (0.2 mmol/kg) of GBCA, subsequent dose reduction studies showed that lower doses of GBCA could still support the desired CE MR angiography with resolution of 1 mm or less. There was sufficient SNR with 0.05 mmol/kg of GBCA (half dose) at 3 T in one large study of carotid CE MR angiography. The trade-off in arterial signal when reducing the contrast dose is less noticeable at 3 T. Because of the higher intrinsic blood-to-tissue signal available at 3 T compared with at 1.5 T, half-dose CE MR angiography is particularly successful at 3 T.


The final image quality on carotid CE MR angiography depends in part on the choice of GBCA, the dose of GBCA, and the field strength and configuration of the MR scanner used. The concern of NSF in connection with GBCA injections also needs to be considered. US Food and Drug Administration (FDA) guidelines currently recommend caution when injecting GBCA in patients with severe or end-stage chronic kidney disease (glomerular filtration rate [GFR] <30 mL/min/1.73 m 2 ) and in patients with acute renal or hepatic failure. These symptoms are uncommon in outpatients presenting with carotid stenosis. The risk/benefit ratio of GBCA in outpatients with moderate chronic kidney disease (GFR 30–59 mL/min/1.73 m 2 ) is unclear, but no proven cases of NSF have occurred in patients with GFR greater than 30 mL/min/1.73 m 2 , other than in the context of acutely deteriorating renal function. Although practice guidelines continue to evolve in the light of accumulating clinical data, the need to use the minimum established dose of GBCA to achieve consistently good image quality of carotid CE MR angiography needs to be balanced with the need to minimize the risk of NSF according to the FDA guidelines.


Optimal Flip Angle


The flip angle that will result in the highest SNR for carotid CE MR angiography depends on the amount of T1 shortening achieved during the first pass of GBCA, and this is called the Ernst angle. The optimal flip angle to use with carotid CE MR angiography varies with the choice of GBCA, the field strength of the MR scanner, and the TR. As described earlier, gadobenate dimeglumine (Multihance), gadofosveset (Ablavar, Lantheus Medical Imaging), and gadobutrol (Gadavist, Bayer HealthCare Pharmaceuticals) have higher T1 relativity GBCA. Therefore, the Ernst angle or optimal flip angle for first-pass carotid CE MR angiography with, for example, gadobenate dimeglumine (Multihance) is higher than with any of the original extracellular GBCAs. At 3 T, high flip angles may not be possible because of specific absorption ratio (SAR) limits exceeding FDA limits for potential heating. Given the longer T1 times at 3 T compared with 1.5 T, the Ernst angle is smaller at 3 T. As noted earlier, the higher SNR at 3 T allows the radiologist to chose to use less GBCA, which results in slightly longer T1 times for CE MR angiography. The optimal flip angle for carotid CE MR angiography therefore varies with the choice of GBCA, field strength, and TR. Table 1 lists specific recommendations. In general, the flip angle should be as large as is allowed by the SAR monitor on the MR scanner.



Table 1

Imaging parameters for CE MR angiography of the extracranial carotid arteries







































































































































































1.5 T 3 T
Avanto Intera HDx TIM Trio Achieva HD750
TR (ms) 2.6 4.8 5.0 3.0 4.8 4.4
TE (ms) 1.2 1.5 1.4 1.2 1.61 1.3
Flip angle (degrees) 25 40 35 20 25 25
BW 440 Hz/pixel 434 Hz/pixel ±50 kHz 720 Hz/pixel 768 Hz/pixel ±90.9 kHz
FOV (cm) 41 35 28 45 35 30
Matrix 512 × 457 432 × 432 346 × 320 640 × 461 584 × 560 376 × 352
RFOV 0.75 0.5 0.9 0.6 1 0.75
Actual FOV 410 × 305 350 × 175 280 × 252 450 × 270 350 × 350 300 × 225
Rx matrix 512 × 340 432 × 260 346 × 288 640 × 346 640 × 640 376 × 264
Slab thickness 108 65 66 102 252 70
No. of slices 120 65 66 128 360 88
Pixel size 0.80 × 0.90 0.81 × 0.81 0.80 × 0.88 0.70 × 0.78 0.6 × 0.62 0.80 × 0.85
Slice thickness (mm) 0.9 1.0 1.0 0.8 1.4 0.8
Voxel volume (mm 3 ) 0.65 0.66 0.71 0.44 0.52 0.54
Parallel image 0 16×
Partial k y 0.75 1.0 1.0 0.75 1.0 1.0
Partial k z 0.75 1.0. 1.0 0.75 1.0 1.0
Percent scan No 78 75 No No 75
Scan time (s) 24 71 36 20 67 40

Sonata BW = 440 Hz/pixel × 512 pixel/1000 = 225.3 kHz or ±112.6 kHz.

Interia BW = 434 Hz/pixel × 432 pixel/1000 = 187.5 kHz or ±93.7 kHz.

GE HD × BW = ±50 kHz or 100,000 Hz/346 = 289 Hz/pixel.

Trio BW = 720 Hz/pixel × 640 pixel/1000 = 460.8 kHz or ±230.4 kHz.

Achieva BW = 768 Hz/pixel × 584 pixel/1000 = 448.5 kHz or ±224.3 kHz.

GE MR750 BW = ±90.9 kHz or 181,800 Hz/376 pixel = 484 Hx/pixel.

Abbreviations: FOV, field of view; RFOV, rectangular field of view; Rx, prescribed.




GBCA


Now that the various technical components to obtaining CE MR angiography with resolution of 1 mm or less have been introduced, the GBCA should be discussed. Workers differ in their approach to how best to coordinate image acquisition with the GBCA contrast bolus. An ideal image acquisition would take advantage of the first pass encompassing the high arterial concentration of GBCA There is more debate about how important the later phases of GBCA distribution are to quality CE MR angiography. Carotid CE MR angiography can be extended to take advantage of the large residual of the initial bolus of contrast injection caused by recirculation. The blood-brain barrier and rapid return of gadolinium contrast through the brain parenchyma and jugular vein back to the heart makes carotid CE MR angiography especially capable of longer acquisitions and potentially higher resolutions. By modeling the elliptical-centric technique, Fain and colleagues showed that there is sufficient contrast for high resolution with increased acquisition time of carotid CE MR angiography. Initial experience with the use of elliptical-centric phase reordering in combination with a bolus arrival scan or fluoroscopic triggering allowed carotid CE MR angiography with voxel size of 1 mm 3 or less with good intra-arterial contrast and little venous contamination. Other workers have focused exclusively on the first arterial peak within a breath-hold acquisition at 1.5 T, also with excellent results.


As with most MR sequence parameters, the choice of GBCA agent and dose varies by MR field strength. In the past 10 years, most investigators achieving CE MR angiography with the desired resolution of 1 mm or less at 1.5 T have used more than a single dose of a variety of the original extracellular GBCA. A high T1 relaxivity GBCA, gadobenate dimeglumine (Multihance, Bracco Diagnostic, Inc.), was recently introduced. Intraindividual crossover comparison of 2 GBCAs showed superior depiction of arterial lumen with a single dose of gadobenate dimeglumine (Multihance, Bracco Diagnostic, Inc.) compared with gadopentetate dimeglumine (Magnevist, Bayer HealthCare Pharmaceuticals) at 1.5 T. Recent work showed that a single dose of gadobenate dimeglumine (Multihance, Bracco Diagnostic, Inc.) can provide similar image quality of CE MR angiography compared with a double dose of other extracellular GBCAs. The use of high T1 relativity GBCA such as gadobenate dimeglumine (Multihance, Bracco Diagnostic, Inc.), gadofosveset (Ablavar, Lantheus Medical Imaging) and gadobutrol (Gadavist, Bayer HealthCare Pharmaceuticals) helps to minimize the total dose of GBCA administration while supporting sufficient SNR to produce carotid CE MR angiography with high image quality. Typically, a single dose (0.1 mmol/kg) of gadobenate dimeglumine is sufficient to maintain high SNR and image quality with a resolution of 0.8-mm 3 or less for carotid CE MR angiography at 1.5 T.


At 3 T, similar intraindividual crossover studies have shown the superiority of gadobenate dimeglumine (Multihance, Bracco Diagnostic, Inc.) compared with other extracellular GBCA in CE MR angiography when using a single dose. Although the initial work on carotid CE MR angiography at 3 T MR used a double dose (0.2 mmol/kg) of GBCA, subsequent dose reduction studies showed that lower doses of GBCA could still support the desired CE MR angiography with resolution of 1 mm or less. There was sufficient SNR with 0.05 mmol/kg of GBCA (half dose) at 3 T in one large study of carotid CE MR angiography. The trade-off in arterial signal when reducing the contrast dose is less noticeable at 3 T. Because of the higher intrinsic blood-to-tissue signal available at 3 T compared with at 1.5 T, half-dose CE MR angiography is particularly successful at 3 T.


The final image quality on carotid CE MR angiography depends in part on the choice of GBCA, the dose of GBCA, and the field strength and configuration of the MR scanner used. The concern of NSF in connection with GBCA injections also needs to be considered. US Food and Drug Administration (FDA) guidelines currently recommend caution when injecting GBCA in patients with severe or end-stage chronic kidney disease (glomerular filtration rate [GFR] <30 mL/min/1.73 m 2 ) and in patients with acute renal or hepatic failure. These symptoms are uncommon in outpatients presenting with carotid stenosis. The risk/benefit ratio of GBCA in outpatients with moderate chronic kidney disease (GFR 30–59 mL/min/1.73 m 2 ) is unclear, but no proven cases of NSF have occurred in patients with GFR greater than 30 mL/min/1.73 m 2 , other than in the context of acutely deteriorating renal function. Although practice guidelines continue to evolve in the light of accumulating clinical data, the need to use the minimum established dose of GBCA to achieve consistently good image quality of carotid CE MR angiography needs to be balanced with the need to minimize the risk of NSF according to the FDA guidelines.


Optimal Flip Angle


The flip angle that will result in the highest SNR for carotid CE MR angiography depends on the amount of T1 shortening achieved during the first pass of GBCA, and this is called the Ernst angle. The optimal flip angle to use with carotid CE MR angiography varies with the choice of GBCA, the field strength of the MR scanner, and the TR. As described earlier, gadobenate dimeglumine (Multihance), gadofosveset (Ablavar, Lantheus Medical Imaging), and gadobutrol (Gadavist, Bayer HealthCare Pharmaceuticals) have higher T1 relativity GBCA. Therefore, the Ernst angle or optimal flip angle for first-pass carotid CE MR angiography with, for example, gadobenate dimeglumine (Multihance) is higher than with any of the original extracellular GBCAs. At 3 T, high flip angles may not be possible because of specific absorption ratio (SAR) limits exceeding FDA limits for potential heating. Given the longer T1 times at 3 T compared with 1.5 T, the Ernst angle is smaller at 3 T. As noted earlier, the higher SNR at 3 T allows the radiologist to chose to use less GBCA, which results in slightly longer T1 times for CE MR angiography. The optimal flip angle for carotid CE MR angiography therefore varies with the choice of GBCA, field strength, and TR. Table 1 lists specific recommendations. In general, the flip angle should be as large as is allowed by the SAR monitor on the MR scanner.


Mar 20, 2017 | Posted by in NEUROLOGICAL IMAGING | Comments Off on Current State-of-the-Art 1.5 T and 3 T Extracranial Carotid Contrast-Enhanced Magnetic Resonance Angiography

Full access? Get Clinical Tree

Get Clinical Tree app for offline access