HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kuhl, C. K. Articles by Schild, H. H. Search for Related Content PubMed PubMed Citation Articles by Kuhl, C. K. Articles by Schild, H. H.

Whole-Body High-Field-Strength (3.0-T) MR Imaging in Clinical Practice

Part I. Technical Considerations and Clinical Applications¹

¹ From the Department of Radiology, University of Bonn, Sigmund-Freud-Str 25, 53105 Bonn, Germany. Received May 22, 2006; revision requested July 21; revision received September 11; accepted October 12; final version accepted December 5; final review and update by C.K.K. October 8, 2007. Address correspondence to C.K.K. (e-mail: kuhl{at}uni-bonn.de).

	ABSTRACT

TOP ABSTRACT INTRODUCTION PHYSICAL EFFECTS OF HIGHER... HOW TO DEAL WITH... CLINICAL APPLICATIONS CONCLUSION ESSENTIALS References

 
In the year 2002, magnetic field strength of more than 2 T was cleared for clinical patient care. Since then, an increasing number of magnetic resonance (MR) systems operating at a field strength of 3.0 T (and higher) have been installed worldwide. This article is the first of a two-part series on clinical high-field-strength MR imaging. Some basic physical effects of higher magnetic fields as they pertain to clinical MR imaging and spectroscopy are reviewed, from the perspective of a clinical radiologist, and strategies that are useful to avoid magnetic field–related difficulties and artifacts are discussed. Advantages and downsides, which can be expected for clinical MR, are presented and compared with the current level of evidence based on published data about MR of the brain and MR angiography. In the second part of the series, clinical applications regarding cardiac, breast, musculoskeletal, abdominopelvic, and pediatric MR and MR spectroscopy will be presented.

© RSNA, 2008

	INTRODUCTION

TOP ABSTRACT INTRODUCTION PHYSICAL EFFECTS OF HIGHER... HOW TO DEAL WITH... CLINICAL APPLICATIONS CONCLUSION ESSENTIALS References

 
Until 2001, safety guidelines limited the magnetic field strength that may be used for diagnostic purposes to 2 T. Since 2002, this limit was expanded to 4.0 T (1), and field strengths of up to 8.0 T are now considered safe for use in human subjects (2). Since then, an increasing number of magnetic resonance (MR) systems operating at a field strength of 3.0 T (and higher) have been installed worldwide. What are the physical effects of a higher magnetic field strength, and how do they affect MR imaging and spectroscopic applications? Why should a radiologist want to use higher field strengths for his or her patients? What is the current level of evidence regarding the use of high-field-strength imaging for clinical purposes? Is 3.0 T twice as good as 1.5 T? This article intends to provide some answers to these questions, from the perspective of the clinical radiologist.

At our department, the first 3.0-T system was installed in early 2002; in 2004, a second system was added. Over this period, about 20 000 patients underwent high-field-strength MR imaging. On the basis of this background, an effort is being made to objectively review published data and to interpret the results in the light of daily clinical and academic practice. This article is the first of a two-part series about the current status of high-field-strength (3.0-T) MR imaging. It provides an overview about the physical effects of high magnetic field strengths and explains appropriate approaches to deal with technical difficulties specific to a high field strength. In addition, clinical results regarding neuroradiologic, neurovascular, and vascular applications are also presented. The second part will be published in the next issue of Radiology. It will present the clinical results published for cardiac, breast, abdominopelvic, musculoskeletal, pediatric, and spectroscopic applications (3).

	PHYSICAL EFFECTS OF HIGHER MAGNETIC FIELD STRENGTHS

TOP ABSTRACT INTRODUCTION PHYSICAL EFFECTS OF HIGHER... HOW TO DEAL WITH... CLINICAL APPLICATIONS CONCLUSION ESSENTIALS References

 
With increasing magnetic field strength, several physical parameters change. Some of these changes are clearly useful or beneficial for MR imaging or MR spectroscopy; others are disadvantageous and need to be compensated for. Most of the field-dependent physical changes, however, can be both advantageous and disadvantageous, depending on the type of MR application that is intended. To better understand the challenges of high-field-strength MR, a short review of the basic physical effects and their effect on clinical MR imaging is appropriate.

Signal-to-Noise Ratio
The signal-to-noise ratio (SNR) will increase almost linearly with increasing field strength (4,5). The individual SNR of a given image will not depend on the field strength alone but also on a number of other parameters, such as spatial resolution (number of protons per voxel), receiver bandwidth, availability of radiofrequency (RF) coils, and type of image acquisition (sequential vs parallel). Since there are a number of physical obstacles to overcome at higher magnetic field strengths, and since the correction of these obstacles usually takes SNR, the SNR advantage can be partly offset. Therefore, the simple concept "double field strength equals double SNR" will be difficult to put to practice in most of the clinical applications of high-field-strength MR imaging and spectroscopy.

SNR can be considered the "currency" of MR imaging and constitutes the actual reason for buying and using MR systems with higher magnetic field strengths. As will be discussed in detail, the extra SNR afforded by higher field strength can be used in several different ways: to reduce acquisition time at a given spatial resolution or to improve spatial representation at a given acquisition time and variable combinations of the two. In addition, techniques that may suffer from very low SNR at 1.5 T may become clinically feasible at 3.0 T.

Larmor Frequency and Chemical Shift
Resonance frequencies increase linearly with field strength. This is given by the Larmor equation (6), = B₀, where is the angular velocity of spin precession, is the gyromagnetic ratio, that is, a substance-specific constant, and B₀ is the strength of the magnetic field. The resonance frequency, = /2, will increase from 63.9 MHz at 1.5 T to 127.8 MHz at 3.0 T. This higher (faster) resonance frequency is the reason for many advantages—but also difficulties—that go along with high-field-strength MR and MR spectroscopy.

The first time for a new user of a 3.0-T system to experience the consequences of the different resonance frequencies with the 1.5- and 3.0-T systems will be the time when he or she purchases the necessary equipment: Since all RF coils (both receive-only and transmit-receive surface coils) have to be tuned to the resonance frequency of the system they are used with, one cannot use surface coils of a 1.5-T system to image at 3.0 T.

Another, more imaging-related consequence of higher resonance frequency is the faster phase cycling of spins. The time until fat and water protons precess (ie, rotate) in phase (ie, point in the same direction) will be shorter compared with that at 1.5 T. In other words, the in-phase periodicity is reduced from 4.6 msec at 1.5 T to 2.3 msec at 3.0 T. This may be advantageous for pulse sequences that use opposed-phase echo times for fat suppression (eg, abdominal MR or time-of-flight [TOF] angiography), because they can be used with shorter echo times which, in turn, should be helpful to reduce artifacts and increase acquisition speed. It may however be disadvantageous for pulse sequences that require an exact in-phase setting (eg, dynamic subtracted contrast material–enhanced breast MR), because the time difference between the in-phase and opposed-phase echo time is only 1.15 msec, which means that even very small deviations of echo timing will result in an out-of-phase or even opposed-phase situation, a setting that must be meticulously avoided, for example, for contrast-enhanced breast MR imaging.

The most important consequence of the higher resonance frequencies at 3.0 T, however, has substantial implications for the technical difficulties related to high-field-strength MR: Higher resonance frequencies require excitation RF pulses with matching (ie, higher) frequencies. The higher the frequency of the RF pulse, the shorter the RF wavelength. The shorter the RF wavelength, the greater the amount of RF pulse absorbed by the tissue and hence greater difficulty to achieve a homogeneous penetration through the tissue (think of high-frequency ultrasound for comparison). Because of the increased absorption of the RF pulse, a stronger heating of tissue will occur and it will be more difficult to achieve a homogeneous RF excitation of spins across the field of view—the latter is referred to as B₁ inhomogeneity. Both effects (tissue heating and B₁ inhomogeneity) can cause substantial difficulties and affect the way we use 3.0-T systems clinically.

It is not only the external magnetic field B₀ that determines the resonance frequency of a given proton. The specific molecular environment in which a proton resides will modify the actual magnetic field experienced by this proton. The subtle differences of the magnetic field that are caused by different molecular environments of protons will translate into a proportional change of the respective proton's resonance frequency. This is referred to as chemical shift and is usually expressed in Hertz or parts per million. This chemical shift increases in proportion to the magnetic field; for fat and water protons, it increases from 220 Hz at 1.5 T to 440 Hz at 3.0 T. The broader separation of resonance frequencies between protons bound in different molecular environments may be used to the advantage of spectroscopic applications because it allows a better separation of the different metabolite resonances, that is, an improved spectral resolution—provided that the broader spectral line width secondary to the shorter T2* at higher magnetic field strengths does not offset this advantage. The broader resonance frequency differences between protons bound in fat versus those bound in water should also improve the robustness of fat suppression, provided that magnetic field inhomogeneities secondary to susceptibility effects do not prevail.

For structural MR imaging, the increased chemical shift can be disadvantageous because it can cause pronounced artifacts at fat-tissue interfaces, giving rise to artificial black lines which may, for example, suggest the presence of a capsule around a fluid collection or cause artificial thickening of cortical bones (7–9).

Relaxation Times
Tissue T1 and T2 relaxation rates depend, to a different degree, on the field strength (Fig 1). At 3.0 T, R1 (longitudinal or spin-lattice) relaxation rate is slowed down, whereas R2 (transverse or spin-spin) relaxation occurs faster. Since T1 relaxation is slowed down, the T1 relaxation time will increase by about 30%. Since T2 relaxation is accelerated, the T2 relaxation time will be shorter by about 15% (7–9,11–19). However, this field-dependent change of relaxation rate is not constant across all tissues. For example, cerebral gray and white matter exhibit a somewhat different degree of T1 prolongation, with white matter experiencing a stronger field-dependent effect. Differences in T1 or T2 relaxation times between different tissues constitute the basis of tissue contrast in MR imaging. Gray matter is distinguished from white matter because gray matter has a longer T1 relaxation time (appears darker on T1-weighted images). At 3.0 T, the T1 relaxation time of white matter will be increased more than that of gray matter, which will cause the relaxation rates of the two tissues to converge. The clinically relevant implication is that the gray matter-to-white matter contrast will be reduced at 3.0 T (Fig 2). Similarly, a reduced contrast may also be found between healthy and diseased tissues: Most disease states go along with a T1 prolongation (on T1-weighted images, diseased tissues appear darker than normal). The nonuniform field dependency of T1 relaxation times may, however, also be advantageous for specific MR applications. For instance, unlike solid tissues, the T1 relaxation times of fluids (blood, tissue water, cerebrospinal fluid) seem to be relatively constant across different field strengths (16). This helps improve vessel-to–stationary tissue contrast at TOF and contrast-enhanced MR angiography.

View larger version (25K): [in this window] [in a new window] [Download PPT slide]   Figure 1: T1 relaxation rates in frontal gray and white matter depending on magnetic field, indicated by increasing Larmor frequency. Note that relaxation rates converge with increasing field; this leads to a reduced signal intensity difference (ie, image contrast). At Larmor frequencies of 64 MHz (1.5 T), the contrast is higher compared with that at Larmor frequencies of 128 MHz (3.0 T). (Reprinted, with permission, from reference 10.)

View larger version (107K): [in this window] [in a new window] [Download PPT slide]   Figure 2a: Contrast-enhanced T1-weighted MR images in 46-year-old male patient with right parietal high-grade glioma. Spin-echo (SE) images at (a) 1.5 and (b) 3.0 T. (c) Inversion-recovery turbo SE, (d) gradient-echo, and (e) modified driven-equilibrium Fourier transform 3-T images. Note reduced T1 contrast on SE images (b vs a). Note that with the pulse sequences used in c–e, an excellent T1 contrast is also attainable at 3.0 T. However, the effect of the contrast agent is best in b. Note the stronger enhancement effect (higher contrast-to-noise ratio [CNR]) of the same dose of contrast agent with the T1-weighted SE pulse sequence (b) compared with the same pulse sequence at 1.5 T (a).

View larger version (120K): [in this window] [in a new window] [Download PPT slide]   Figure 2b: Contrast-enhanced T1-weighted MR images in 46-year-old male patient with right parietal high-grade glioma. Spin-echo (SE) images at (a) 1.5 and (b) 3.0 T. (c) Inversion-recovery turbo SE, (d) gradient-echo, and (e) modified driven-equilibrium Fourier transform 3-T images. Note reduced T1 contrast on SE images (b vs a). Note that with the pulse sequences used in c–e, an excellent T1 contrast is also attainable at 3.0 T. However, the effect of the contrast agent is best in b. Note the stronger enhancement effect (higher contrast-to-noise ratio [CNR]) of the same dose of contrast agent with the T1-weighted SE pulse sequence (b) compared with the same pulse sequence at 1.5 T (a).

View larger version (140K): [in this window] [in a new window] [Download PPT slide]   Figure 2c: Contrast-enhanced T1-weighted MR images in 46-year-old male patient with right parietal high-grade glioma. Spin-echo (SE) images at (a) 1.5 and (b) 3.0 T. (c) Inversion-recovery turbo SE, (d) gradient-echo, and (e) modified driven-equilibrium Fourier transform 3-T images. Note reduced T1 contrast on SE images (b vs a). Note that with the pulse sequences used in c–e, an excellent T1 contrast is also attainable at 3.0 T. However, the effect of the contrast agent is best in b. Note the stronger enhancement effect (higher contrast-to-noise ratio [CNR]) of the same dose of contrast agent with the T1-weighted SE pulse sequence (b) compared with the same pulse sequence at 1.5 T (a).

View larger version (113K): [in this window] [in a new window] [Download PPT slide]   Figure 2d: Contrast-enhanced T1-weighted MR images in 46-year-old male patient with right parietal high-grade glioma. Spin-echo (SE) images at (a) 1.5 and (b) 3.0 T. (c) Inversion-recovery turbo SE, (d) gradient-echo, and (e) modified driven-equilibrium Fourier transform 3-T images. Note reduced T1 contrast on SE images (b vs a). Note that with the pulse sequences used in c–e, an excellent T1 contrast is also attainable at 3.0 T. However, the effect of the contrast agent is best in b. Note the stronger enhancement effect (higher contrast-to-noise ratio [CNR]) of the same dose of contrast agent with the T1-weighted SE pulse sequence (b) compared with the same pulse sequence at 1.5 T (a).

View larger version (121K): [in this window] [in a new window] [Download PPT slide]   Figure 2e: Contrast-enhanced T1-weighted MR images in 46-year-old male patient with right parietal high-grade glioma. Spin-echo (SE) images at (a) 1.5 and (b) 3.0 T. (c) Inversion-recovery turbo SE, (d) gradient-echo, and (e) modified driven-equilibrium Fourier transform 3-T images. Note reduced T1 contrast on SE images (b vs a). Note that with the pulse sequences used in c–e, an excellent T1 contrast is also attainable at 3.0 T. However, the effect of the contrast agent is best in b. Note the stronger enhancement effect (higher contrast-to-noise ratio [CNR]) of the same dose of contrast agent with the T1-weighted SE pulse sequence (b) compared with the same pulse sequence at 1.5 T (a).

A stronger T2*-shortening effect should also be expected for contrast agents with a predominantly T2*-shortening effect, such as superparamagnetic iron oxide (SPIO)-based contrast agents (26–28). However, the lower baseline signal intensity at 3.0 T may partly foil the enhanced T2* effect of an SPIO. In fact, so far, a higher contrast between SPIO-enhancing lesions and normal tissue has not been documented in patients with focal liver lesions (28,29).

Susceptibility
Susceptibility effects increase approximately linearly with field strength. Susceptibility effects are caused by the variable magnetic susceptibility. Magnetic susceptibility is a technical term for "magnetizability" of a material (or tissue). Magnetization of a material means that, if that material is placed within the range of a magnet, it will, to an extent dictated by the individual material's magnetic susceptibility, develop a local magnetic field by itself—and, thus, will modify the overall magnetic field. Most tissues of the human body are diamagnetic, that is, they have a very low magnetic susceptibility and will therefore reduce the net magnetic field. Paramagnetic material (many T1-based contrast agents such as gadolinium chelates) will increase the net magnetic field slightly, superparamagnetic substances will increase the net magnetic field somewhat more, and ferromagnetic substances will increase it even more. Within the diamagnetic range, the magnetic susceptibility, or magnetizability, of the different tissues of the human body varies, and this will modify the local magnetic field experienced by protons in different tissues.

As given by the Larmor equation, the variations of the local magnetic field will translate into proportional variations of the respective protons' resonance frequencies. Since spatial encoding in MR imaging assumes that the magnetic field is constant across the field of view and would only vary with the applied gradients, such minute susceptibility-induced resonance frequency differences may lead to misregistrations of the position of a given proton. In particular, in pulse sequences without refocusing 180° RF pulses, such as gradient-echo and echo-planar imaging, this will lead to distortions and signal intensity variations across the field of view. In addition, in the transition zones of tissues with strongly varying susceptibilities, such as bone and soft tissue, intravoxel dephasing may occur and lead to signal cancellations. The susceptibility-induced shift of resonance frequencies may also compromise the success of fat suppression, in particular in tissues close to fat-water or air-tissue interfaces.

There are clinical applications for which susceptibility effects (or T2* effects) are the desired contrast and are exploited for diagnostic purposes—for these applications, the stronger susceptibility effects at 3.0 T can be advantageous and may translate into a higher sensitivity—provided that susceptibility effects are sufficiently controlled to maintain a diagnostic image quality. First-pass perfusion imaging and all applications that use the blood oxygen level–dependent, or BOLD, contrast (mainly functional MR imaging) will benefit from higher magnetic field strength. T2* contrast is also used to allow the identification of hemosiderin (in areas with chronic hemorrhage) or calcifications (Figs 3, 4).

View larger version (126K): [in this window] [in a new window] [Download PPT slide]   Figure 3a: Transverse T2-weighted turbo SE (a) 1.5-T (repetition time msec/echo time msec, 2500/100) and (b) 3.0-T (3000/80) MR images in a 52-year-old male patient with pontine cavernoma. Note the superior image quality of b, with improved depiction of the cranial nerves in the cerebellopontine cistern (white arrowhead) and the auditory canal. Note the stronger susceptibility effects caused by hemosiderin-containing cavernoma, which is much more conspicuous at 3.0 T (arrow). Note the improved visibility of the physiologic iron-containing cerebellar dentate nucleus (black arrowhead).

View larger version (124K): [in this window] [in a new window] [Download PPT slide]   Figure 3b: Transverse T2-weighted turbo SE (a) 1.5-T (repetition time msec/echo time msec, 2500/100) and (b) 3.0-T (3000/80) MR images in a 52-year-old male patient with pontine cavernoma. Note the superior image quality of b, with improved depiction of the cranial nerves in the cerebellopontine cistern (white arrowhead) and the auditory canal. Note the stronger susceptibility effects caused by hemosiderin-containing cavernoma, which is much more conspicuous at 3.0 T (arrow). Note the improved visibility of the physiologic iron-containing cerebellar dentate nucleus (black arrowhead).

View larger version (99K): [in this window] [in a new window] [Download PPT slide]   Figure 4a: Functional brain MR images in the same individuals. (a) 1.5-T (3000/50, 90° flip angle) and (b) 3.0-T (3000/35, 90° flip angle) images obtained with single-shot echo-planar pulse sequences from a group analysis of subjects who underwent a block-designed sequence of simple motor tasks, lexical decision tasks, semantic decision tasks, and verbal fluency tasks and superimposed on a transverse three-dimensional turbo gradient-echo image (3.8/8.2, 8° flip angle). The aim was to identify cortical areas associated with decision processes. At 3.0 T, a stronger blood oxygen level–dependent (BOLD) response was obtained for regions that appeared activated at both field strengths (primary motor cortex); further areas of activation are visible in the secondary motor cortex and the right prefrontal region (independent of the side with which the motor response was given). This suggests that at 3.0 T, higher cognitive functions that are associated with very subtle BOLD effects are depictable. (Reprinted, with permission, from reference 30.)

View larger version (103K): [in this window] [in a new window] [Download PPT slide]   Figure 4b: Functional brain MR images in the same individuals. (a) 1.5-T (3000/50, 90° flip angle) and (b) 3.0-T (3000/35, 90° flip angle) images obtained with single-shot echo-planar pulse sequences from a group analysis of subjects who underwent a block-designed sequence of simple motor tasks, lexical decision tasks, semantic decision tasks, and verbal fluency tasks and superimposed on a transverse three-dimensional turbo gradient-echo image (3.8/8.2, 8° flip angle). The aim was to identify cortical areas associated with decision processes. At 3.0 T, a stronger blood oxygen level–dependent (BOLD) response was obtained for regions that appeared activated at both field strengths (primary motor cortex); further areas of activation are visible in the secondary motor cortex and the right prefrontal region (independent of the side with which the motor response was given). This suggests that at 3.0 T, higher cognitive functions that are associated with very subtle BOLD effects are depictable. (Reprinted, with permission, from reference 30.)

View larger version (143K): [in this window] [in a new window] [Download PPT slide]   Figure 5a: Effect of SAR-related difficulties at 3.0 T without parallel imaging. (a, b) Sagittal thin-section (2-mm) T2-weighted turbo SE MR images (250-mm field of view, 258 x 512 imaging matrix) in 29-year-old female patient with multiple sclerosis investigated at (a) 1.5 T (2780/100) and (b) 3.0 T (4000/80). Number of signals acquired was reduced from four at 1.5 T to two at 3.0 T. Acquisition time was 3 minutes 20 seconds for a versus 2 minutes 40 seconds for b. Because of SAR-limiting regulations at 3.0 T, halving the number of signals acquired does not translate into halved acquisition time. There is still higher SNR at 3.0 T than at 1.5 T despite the halved number of signals acquired. Visibility of callosal plaques (arrowheads) is improved at 3.0 T.

View larger version (127K): [in this window] [in a new window] [Download PPT slide]   Figure 5b: Effect of SAR-related difficulties at 3.0 T without parallel imaging. (a, b) Sagittal thin-section (2-mm) T2-weighted turbo SE MR images (250-mm field of view, 258 x 512 imaging matrix) in 29-year-old female patient with multiple sclerosis investigated at (a) 1.5 T (2780/100) and (b) 3.0 T (4000/80). Number of signals acquired was reduced from four at 1.5 T to two at 3.0 T. Acquisition time was 3 minutes 20 seconds for a versus 2 minutes 40 seconds for b. Because of SAR-limiting regulations at 3.0 T, halving the number of signals acquired does not translate into halved acquisition time. There is still higher SNR at 3.0 T than at 1.5 T despite the halved number of signals acquired. Visibility of callosal plaques (arrowheads) is improved at 3.0 T.

Dielectric Effects
Because of the higher resonance frequency at higher magnetic field strength, the wavelength of RF pulses becomes shorter and approaches the diameter of the human body (at 3.0 T, the RF wavelength in water is only about 26 cm). This has an important effect on the absorption of RF energy and on its distribution in tissue, in other words, the "B₁ homogeneity," especially in body applications. At higher magnetic field strength, the maximum B₁ field will be reached more toward the center of the body ("field focusing"), such that a more pronounced energy deposition in deep, central parts of the body will occur. This field-focusing effect may be reinforced if standing RF waves develop by means of a reflection of RF waves at interfaces with high conductivity gradients such as the chest or body wall (dielectric resonance) (1). While the variable conductivity of tissues should prevent such RF-antenna effects at least in part, in the presence of metallic implants, this may cause excessive heating (31). Apart from safety issues, the main consequence of the B₁ field–focusing effect is that at higher magnetic field strength it is more difficult to establish a homogeneous signal across the field of view. Whereas the field focusing may be "damped" by eddy currents that occur close to the body surface (1), in the intermediate region between the body surface and the central part, the effective flip angles may be lower than their specified values. As a result, rings of concentric signal intensity loss or "shading" of signal intensity may occur in the area between the superficial and central body parts (Fig 6). This is more often observed in abdominal and thoracic large-field-of-view applications, where the tissue of the abdominal (or thoracic) wall seems to exhibit an RF-shielding effect. Also, peritoneal or thoracic fluid collections such as ascites or pleural effusions cause a substantial change in the conductivity and the dielectric properties of the tissue contained in the field of view and may thereby reinforce dielectric effects.

View larger version (49K): [in this window] [in a new window] [Download PPT slide]   Figure 6a: (a, b) Dielectric effects at T1-weighted gradient-echo (T1 FFE) and T2-weighted turbo SE (T2 TSE) MR sequences. (a) Left: Images of the water phantom at 1.5 T. Right: Respective signal intensity histograms obtained over the field of view. (b) Left: Images of the water phantom at 3.0 T. Right: Respective signal intensity histograms obtained over the field of view. Note the concentric signal intensity shadowing at 3.0 T secondary to dielectric effects and the "field-focusing" effect (highest signal intensity obtained in the image center), which is more pronounced at 3.0 than at 1.5 T.

View larger version (69K): [in this window] [in a new window] [Download PPT slide]   Figure 6b: (a, b) Dielectric effects at T1-weighted gradient-echo (T1 FFE) and T2-weighted turbo SE (T2 TSE) MR sequences. (a) Left: Images of the water phantom at 1.5 T. Right: Respective signal intensity histograms obtained over the field of view. (b) Left: Images of the water phantom at 3.0 T. Right: Respective signal intensity histograms obtained over the field of view. Note the concentric signal intensity shadowing at 3.0 T secondary to dielectric effects and the "field-focusing" effect (highest signal intensity obtained in the image center), which is more pronounced at 3.0 than at 1.5 T.

	HOW TO DEAL WITH HIGH-FIELD-STRENGTH–RELATED DIFFICULTIES

TOP ABSTRACT INTRODUCTION PHYSICAL EFFECTS OF HIGHER... HOW TO DEAL WITH... CLINICAL APPLICATIONS CONCLUSION ESSENTIALS References

Compensate for Altered Relaxation Times
The usual recommendation to account for the longer T1 and shorter T2 relaxation times is to increase repetition time for T1-weighted imaging and reduce echo and repetition times for T2-weighted imaging. In fact, however, most working groups use the same contrast-determining parameter settings for imaging at 1.5 and at 3.0 T. One reason is that longer repetition time would unduly increase the acquisition time in T1-weighted imaging. In addition, use of the same parameter settings at 1.5 T and at 3.0 T implies a relatively stronger T1 or T2 weighting at 3.0 T. This may help compensate the lower dynamic range (flatter image contrast) of T1-weighted imaging at 3.0 T—yet, it is also associated with an SNR sacrifice. A smarter way to improve T1 contrast at higher field strength is by using different pulse sequence protocols for T1-weighted imaging (19,22,23) (Fig 2). Last, in many clinical situations, a reduced T1 contrast may not be as problematic as it may appear. The reason is that the role of plain T1-weighted images for depicting anatomic details has been decreasing with the increasing availability of fast T2-weighted pulse sequences. In the majority of contemporary pulse sequence protocols, the morphology of a given structure is investigated with fast T2- or intermediate-weighted SE pulse sequences, whereas T1-weighted pulse sequences are usually only performed for contrast-enhanced imaging, be it for MR angiography or for structural contrast-enhanced imaging. The depiction of these effects is not impaired at higher magnetic field strength—or may even be improved due to the stronger T1-shortening effects of contrast agents at higher magnetic fields.

Avoid Susceptibility Effects
If a voxel is large enough to contain tissues with substantially differing susceptibility (magnetizability), this will lead to a substantially faster phase dispersion of spins (referred to as "intravoxel dephasing") and, thus, a faster decay of the transverse magnetization—which leads to a signal loss referred to as susceptibility artifact. The easiest way to avoid this is by using smaller voxel sizes that are less prone to contain tissues with differing susceptibility, in other words: a higher spatial resolution. This will, however, not work for single-shot echo-planar imaging techniques because an increased imaging matrix would be associated with an increased echo time—which would in turn again increase susceptibility effects. Another way to avoid susceptibility effects is by using the shortest possible echo time with the highest achievable receiver bandwidth. Second, use of SE instead of gradient-echo imaging will substantially reduce unwanted susceptibility effects. Third, engineering tools such as additional shim coils, new shimming algorithms, and/or higher order shimming of the gradient coils (currently up to third-order channels) may be used to improve the magnetic field homogeneity in a specific region of interest. Last, parallel imaging techniques reduce the length of the echo train, and, thus, reduce the accumulation of phase errors, in particular in single-shot echo-planar imaging sequences (Fig 7). As an alternative to parallel imaging, new reconstruction algorithms have been developed that periodically correct phase errors in DWI at higher fields (periodically rotated overlapping parallel lines with enhanced reconstruction, or PROPELLER) (32,33).

View larger version (71K): [in this window] [in a new window] [Download PPT slide]   Figure 7a: Sensitivity encoding (SENSE) for diffusion-weighted imaging (DWI). (a) Four transverse sections of a single-shot echo-planar DWI MR pulse sequence at 3.0 T without parallel imaging (4543/77; acquisition time, 1 minute 37 seconds). (b) The same four sections of the same pulse sequence with parallel imaging and a reduction factor of three (3003/69; acquisition time, 1 minute 9 seconds). Note substantial image distortions that occur in areas close to the skull base (temporal lobe and frontal poles) (arrows). These effects are reduced by using SENSE. Overall SNR is not reduced due to use of parallel imaging but is indeed increased.

View larger version (89K): [in this window] [in a new window] [Download PPT slide]   Figure 7b: Sensitivity encoding (SENSE) for diffusion-weighted imaging (DWI). (a) Four transverse sections of a single-shot echo-planar DWI MR pulse sequence at 3.0 T without parallel imaging (4543/77; acquisition time, 1 minute 37 seconds). (b) The same four sections of the same pulse sequence with parallel imaging and a reduction factor of three (3003/69; acquisition time, 1 minute 9 seconds). Note substantial image distortions that occur in areas close to the skull base (temporal lobe and frontal poles) (arrows). These effects are reduced by using SENSE. Overall SNR is not reduced due to use of parallel imaging but is indeed increased.

Parallel imaging works by exploiting the spatially variable sensitivity of surface coils for encoding spatial information (34–36). It is a daily clinical experience that the sensitivity of surface coils decreases with increasing distance from the coil; in other words, there is a linear sensitivity gradient. The basic idea of parallel imaging is to exploit this coil-inherent sensitivity gradient for spatial encoding. To decode this information, at least two (or more) identical surface coils have to be positioned parallel, across from each other, and each coil's sensitivity profile is then mapped with reference to the body coil. The combined sensitivity maps will be used to assign the correct spatial representation of a measured signal, just depending on the sensitivity with which a given pixel is "seen" by each of the surface coils.

Parallel imaging allows one to omit phase-encoding steps that would usually each require an RF pulse—of which each would take acquisition time and would cause more tissue heating. The number of phase-encoding steps in parallel imaging compared with that in conventional, that is, sequential phase encoding, is given by the reduction factor R. With a reduction factor of two (n), only half (or 1/n) of the phase-encoding steps will be acquired, yielding a twofold (n-fold) acceleration of data sampling, and, accordingly, a twofold reduction of RF burden of a given pulse sequence. This explains why parallel imaging is considered to be one of the most effective means to reduce RF deposition at higher magnetic field strengths.

The downside to parallel imaging in virtually all other applications is that it takes an SNR. The SNR penalty correlates with the degree of undersampling; the net SNR will be inversely proportional to the square root of the reduction factor. Accordingly, at a reduction factor of two, parallel imaging will be associated with an SNR penalty of about 30%–40%. Because of the SNR penalty, at 1.5 T the use of parallel imaging was limited to pulse sequences with inherently high SNR (eg, contrast-enhanced MR angiography). At 3.0 T, the higher baseline SNR allows one to use parallel imaging in many more clinical applications, specifically also for the fast acquisition of high image matrices.

Accordingly, another way to look at the facts is to state that 3.0 T not only requires the use of parallel imaging to avoid SAR and susceptibility-related difficulties, but that it allows the application of parallel imaging more broadly because it compensates for parallel imaging–induced SNR loss. Since parallel imaging, per se, is such a versatile tool, our ability to better exploit this technique may indeed be considered one of the major advantages of higher magnetic field strengths. In short: Parallel imaging is required to fully exploit the 3.0 T potential, and 3.0 T is required to fully exploit the parallel imaging potential. The two techniques work synergistically.

In single-shot pulse sequences, the main incentive to use parallel imaging is to reduce susceptibility effects (37,38). A welcome side effect is that in single-shot sequences, unlike in other applications, parallel imaging will even improve (not decrease) SNR (Fig 7). The reason is that in single-shot protocols with parallel imaging and a reduction factor of, for example, two, only the first half of the echo train will be used, whereas the second half of the echo train—which usually suffers from very low signal due to T2 decay—will not be needed for image generation (38).

Refocusing flip angle modulation techniques such as hyperecho, flip angle sweep, or sampling perfection with application of optimized contrasts using different flip angle evolutions (17,39–44) are another means to reduce SAR deposition in turbo (fast) SE or gradient-echo, that is, in RF-intensive, pulse sequences (Fig 8). Since the flip angle determines the amount of energy deposition, reduction of the refocusing flip angle is a very effective means to reduce the RF burden. In flip angle sweep, this is achieved by successively reducing the refocusing flip angle from its usual 180° to 130°–170° over the echo train. The possible downside is that because of the long T1 relaxation times at 3.0 T and the interference of the so-called stimulated echoes, with very small refocusing flip angles like 130°, flip angle sweep may give rise to a mixed contrast (between T2 and T1), which may be difficult to interpret. In a study that systematically analyzed tissue contrast (including pathologic lesions) with stepwise reduction of the refocusing flip angle, a reduction to 75° was associated with an acceptable 15% loss of image contrast. If refocusing flip angle modulation techniques are combined with parallel imaging, the two strategies work synergistically and allow scan time reduction by a factor of six to nine (44).

View larger version (41K): [in this window] [in a new window] [Download PPT slide]   Figure 8a: MR images in a 19-year-old male patient with multiple sclerosis. Faster image acquisition due to parallel imaging and flip angle modulation techniques. (a) Three sections of a fat-suppressed transverse single-shot T2-weighted turbo SE sequence with a SENSE factor of three plus flip angle sweep of 60°, resulting in an acquisition time of 8 seconds. (b) Same transverse sections with the same spatial resolution acquired with a multishot T2-weighted turbo SE sequence without SENSE and flip angle sweep yield an acquisition time of 109 seconds. Note lower SNR and somewhat reduced image quality in a; however, all demyelinating plaques that are visible in b have also been detected in a.

View larger version (46K): [in this window] [in a new window] [Download PPT slide]   Figure 8b: MR images in a 19-year-old male patient with multiple sclerosis. Faster image acquisition due to parallel imaging and flip angle modulation techniques. (a) Three sections of a fat-suppressed transverse single-shot T2-weighted turbo SE sequence with a SENSE factor of three plus flip angle sweep of 60°, resulting in an acquisition time of 8 seconds. (b) Same transverse sections with the same spatial resolution acquired with a multishot T2-weighted turbo SE sequence without SENSE and flip angle sweep yield an acquisition time of 109 seconds. Note lower SNR and somewhat reduced image quality in a; however, all demyelinating plaques that are visible in b have also been detected in a.

Patient Safety Issues
Clinical symptoms that have been associated with high static magnetic fields are dizziness (in particular with rapid head movements in or close to the magnet bore) and nausea (47). There are a few studies, to our knowledge, that prospectively and systematically investigated the incidence and spectrum of clinical side effects of higher magnetic field strength. In a small intraindividual comparative study, Born and coworkers (48) did not find statistically significant differences regarding side effects experienced at 1.5 T and at 3.0 T. Sommer and co-workers (49) investigated the safety of different neurovascular surgical clips and stents and found the majority to be fully compatible.

Although heating effects and dislocation are conceivable, clinical experience suggests that patients with hip replacement or other types of even large osteosynthetic implants can undergo high-field-strength MR imaging just as they undergo MR at 1.5 T as long as the implant is not included in the field of view. If the material is within the field of view, RF heating may become substantial. Although no such events have been published yet, it is conceivable that RF heating of implanted material may become a safety hazard. No heating effects, torque, or image degradation has been observed in cerebral high-field-strength MR imaging and MR angiographic studies in patients who underwent endovascular coiling procedures (50). Accordingly, it seems that these patients can undergo MR angiography safely, as long as the usual precautions are observed that are also in place for imaging at 1.5 T. This is important because it suggests that the higher spatial resolution afforded by high-field-strength MR angiography can be used to follow patients after endovascular coil placement.

	CLINICAL APPLICATIONS

TOP ABSTRACT INTRODUCTION PHYSICAL EFFECTS OF HIGHER... HOW TO DEAL WITH... CLINICAL APPLICATIONS CONCLUSION ESSENTIALS References

 
The higher expenses that are associated with the purchase, siting, and maintenance of high-field-strength as compared with, for example, 1.5-T systems seem justifiable for the presumed "added value" of improved diagnostic capabilities and/or increased patient throughput. The underlying assumption is "double the field equals double the signal equals doubled speed or doubled resolution." Vendors enhance this assumption with words such as "unprecedented image quality" and testimonials of 3.0-T system users. High field strength seems to have become almost a "must-have" for an academic institution, and even more so for private practices in competitive environments (42).

While acquisition of less noisy or just "prettier images," per se, may be considered an added value in some settings, most radiologists would agree that the ultimate question to be answered is: Does 3.0 T help improve clinical decision making? Does it improve the accuracy with which a diagnosis is established or at least the confidence with which radiologists are able to reach even only the same level of accuracy? Do radiologists diagnose treatable diseases earlier? Can treatment effects or disease progression be more accurately monitored?

In order to evaluate the clinical advantages of 3.0 T compared with 1.5 T, dedicated clinical studies have to be performed that fulfill the following requirements: Studies must (a) have a prospective design, (b) include patients (not only volunteers or cadaveric specimens), (c) offer an intraindividual comparison with 1.5 T, (d) use state-of-the-art imaging and artifact-reduction techniques on both systems, (e) should include a standard of reference, and (f) use a meaningful endpoint.

A meaningful endpoint is one that provides evidence on the clinical utility of high-field-strength MR imaging or MR spectroscopy. Comparing SNR or CNR levels or assessing image quality or image quality–related issues, such as the number of depicted vessel segments, should be considered as part of intermediate stages—necessary, but not sufficient to assess the added clinical value of 3.0 T.

Such "added clinical value" beyond image quality may be achieved by improving existing imaging techniques. For instance, a higher spatial resolution may allow an improved assessment of mass margins in, for example, breast imaging, which should improve the technique's specificity. A higher lesion-to-normal tissue contrast in disease states that would go along with only subtle signal intensity changes at 1.5 T—such as early demyelinating plaques or very small ischemic lesions—may enhance the sensitivity with which these diseases may be diagnosed.

In addition to this evolutionary process of improving existing technologies, there is hope that 3.0 T might pave the way for new approaches that offer new or independent diagnostic information compared to what is available or feasible at 1.5 T. Examples would be arterial spin labeling for perfusion imaging, time-resolved contrast-enhanced MR angiography, diastolic cardiac tagging, DWI of cartilage, or investigating myo-inositol concentrations at cerebral MR spectroscopy.

Conducting intraindividual clinical studies has always been cumbersome, and it is even more difficult for high-field-strength MR imaging. This is not only because of the difficulty to obtain informed consent for medically unnecessary repeat examinations at the two field strengths. Rather, this is because current technical and engineering progress is so fast that the results obtained for a given high-field-strength imaging technique may not be representative of the 3.0-T versus 1.5-T performance already a year or 2 years after data collection was closed. Accordingly, all that is going to be explained in the following paragraphs has to be considered provisional and reflects the current status at the time this article was written.

Structural Brain Imaging
The brain was the first application of high-field-strength MR imaging since transmit-receive head coils were first available. For a number of reasons, the head is a good starting point for high-field-strength MR: The relatively small size helps to achieve decent magnetic and RF field homogeneities across the field of view. RF penetration depth and SAR limitations are less problematic for small parts than they are for imaging large objects. Last, there is only little pulsation and no respiratory or gross motion artifacts, unlike in thoracic or abdominal imaging. In fact, long before the approval of high-field-strength MR for diagnostic imaging purposes, functional MR imaging of brain activation in volunteers has successfully been performed with MR systems operating at 3.0 T and even higher field strength (51).

The detection of demyelinating plaques in the early stages of multiple sclerosis has been one of the first disease states investigated with high-field-strength MR (52–58). Table 1 summarizes the published data. Authors concordantly confirmed the expected higher SNR and lesion-to-background CNR at higher magnetic field strength. Across all published studies, more lesions were identified at higher magnetic field strength, and the additional lesion yield at 3.0 T ranges between 13% and 45%. Whenever contrast-enhanced studies were conducted, a higher sensitivity for enhancing lesions was found. Most of the "3.0-T–only" lesions were small, less than 5 mm in size.

View larger version (83K): [in this window] [in a new window] [Download PPT slide]   Figure 9a: (a, b) Transverse fluid-attenuated inversion recovery images in a 22-year-old female patient with clinically isolated syndrome who underwent cerebral MR at (a) 3.0 T (repetition time msec/echo time msec/inversion time msec, 12 000/140/2850; turbo factor, 38; one signal acquired; acquisition time, 4 minutes) and (b) 1.5 T (6000/110/2000; turbo factor, 29; two signals acquired; acquisition time, 3 minutes). The infratentorial lesion is visible on the high-field-strength image (arrow) but not on image obtained at 1.5 T.

View larger version (101K): [in this window] [in a new window] [Download PPT slide]   Figure 9b: (a, b) Transverse fluid-attenuated inversion recovery images in a 22-year-old female patient with clinically isolated syndrome who underwent cerebral MR at (a) 3.0 T (repetition time msec/echo time msec/inversion time msec, 12 000/140/2850; turbo factor, 38; one signal acquired; acquisition time, 4 minutes) and (b) 1.5 T (6000/110/2000; turbo factor, 29; two signals acquired; acquisition time, 3 minutes). The infratentorial lesion is visible on the high-field-strength image (arrow) but not on image obtained at 1.5 T.

Another example of a clinical setting where high spatial and contrast resolutions are required to identify subtle morphologic changes is the patient with simple or complex partial seizures. Knake et al (61) investigated the use of 3.0-T MR (with phased-array coils) for the presurgical evaluation of 40 patients with focal epilepsy. They found that 3.0-T phased-array coil imaging yielded additional, therapeutically relevant diagnostic information in 48% (19 of 40) of patients. Our experiences parallel those of Knake and co-workers in that a "second-look" MR that is targeted based on clinical symptoms, semiology of seizures, and electroencephalographic findings may often reveal the structural correlate of epileptogenic foci in patients with "negative" outside MR examinations (62). This, however, is achievable even with 1.5-T systems and seems to depend more on the specific expertise that is available in a dedicated epilepsy imaging program rather than on the field strength alone. It is to be expected, but has not yet been documented, that high-spatial-resolution three-dimensional volumetric images that are feasible at 3.0 T should facilitate detection of subtle cortical dysplasia. In addition, very-high-spatial-resolution T2-weighted images help identify the hippocampal subanatomy in patients with temporal lobe epilepsy (Fig 10). It appears that early in the course of mesial temporal sclerosis, this subanatomy is lost, which suggests that this finding may be used as an additional criterion for the early diagnosis of hippocampal involvement in temporal lobe epilepsy—but again, this has not been investigated prospectively.

View larger version (137K): [in this window] [in a new window] [Download PPT slide]   Figure 10a: MR images in 32-year-old female patient with medically refractory complex partial seizures. (a) Coronal high-spatial-resolution T2-weighted turbo SE image (4097/100; field of view, 200; 2-mm section thickness; imaging matrix, 640 x 1024). (b) Close-up images of the hippocampal formation. The signal intensity of hippocampus on the left is increased (arrow) compared with that on the right, indicative of sclerosis. No atrophy has yet occurred (volume comparable to right side). Note visibility of the hippocampal subanatomy with a dark line of white matter infolded in the hippocampus, probably representing the alveus. This internal architecture is lost in the sclerotic hippocampus.

View larger version (75K): [in this window] [in a new window] [Download PPT slide]   Figure 10b: MR images in 32-year-old female patient with medically refractory complex partial seizures. (a) Coronal high-spatial-resolution T2-weighted turbo SE image (4097/100; field of view, 200; 2-mm section thickness; imaging matrix, 640 x 1024). (b) Close-up images of the hippocampal formation. The signal intensity of hippocampus on the left is increased (arrow) compared with that on the right, indicative of sclerosis. No atrophy has yet occurred (volume comparable to right side). Note visibility of the hippocampal subanatomy with a dark line of white matter infolded in the hippocampus, probably representing the alveus. This internal architecture is lost in the sclerotic hippocampus.

In patients suspected of having a stroke, in unconscious or uncooperative patients, and especially in pediatric patients—in short—in patients in whom acquisition time matters, 3.0-T SNR may be used to increase acquisition speed. For reasons mentioned above, this can only be achieved if SAR-compensating techniques are available. Lutterbey and co-workers (63) demonstrated that by combining a turbo SE pulse sequence with flip angle sweep and parallel imaging, they were able to reduce the acquisition time of a T2-weighted turbo SE sequence from 2 minutes 7 seconds down to 8 seconds. This protocol was used in 131 patients with 162 cerebral lesions. The authors found that the diagnostic yield was equivalent, with 96% versus 99% of lesions visualized with the fast versus the standard protocol. Whether this loss of diagnostic information that is associated with fast imaging is indeed acceptable on clinical grounds, however, has not been established.

DWI and Diffusion-Tensor Imaging
DWI is routinely used to allow the earliest possible diagnosis of cerebral infarction and is considered an integral part of all stroke imaging protocols. Mostly used as single-shot echo-planar imaging technique, DWI at 1.5 T offers only borderline SNR; as such, DWI would be a good candidate to benefit from higher magnetic fields. On the other hand, the stronger susceptibility effects may degrade image quality and annihilate any advantages. Our group compared single-shot DWI echo-planar imaging in 25 patients with subacute ischemic stroke at 3.0 T and at 1.5 T (64). As one would expect, the higher susceptibility at 3.0 T caused substantial image distortions and blurring—at the time the study was conducted, there was no parallel imaging available to avoid these artifacts. Still, the significantly higher lesion CNR translated into an improved detection rate of subtle ischemic injuries in eight (42%) of 19 patients. In all eight cases, the "3.0-T–only" ischemic lesions were small embolic infarctions that were found in addition to other, larger ischemic lesions, which had also been depicted at 1.5-T DWI (Fig 11). Therefore, on the basis of a patient-wise analysis, sensitivity was not improved at 3.0 T. The 3.0-T advantage is probably better exploitable if parallel imaging is used to reduce susceptibility artifacts (38).

View larger version (30K): [in this window] [in a new window] [Download PPT slide]   Figure 11a: (a, b) Transverse single-shot MR images (4345/82; 128 x 128 matrix; 20 sections; 5-mm section thickness; and b values of 0, 600, and 1000 sec/mm2) in a 46-year-old female patient with clinical evidence of subacute infarction (72 hours after onset) who underwent DWI at (a) 1.5 and (b) 3.0 T within 10 minutes. Images are isotropic at b value of 1000 sec/mm2. Note the higher apparent SNR and the improved visibility of ischemic lesions at 3.0 T (arrowheads). Imaging studies in b were obtained without parallel imaging (it was not available at the time this study was performed); accordingly, susceptibility-related image distortions are more pronounced in b than a.

View larger version (46K): [in this window] [in a new window] [Download PPT slide]   Figure 11b: (a, b) Transverse single-shot MR images (4345/82; 128 x 128 matrix; 20 sections; 5-mm section thickness; and b values of 0, 600, and 1000 sec/mm2) in a 46-year-old female patient with clinical evidence of subacute infarction (72 hours after onset) who underwent DWI at (a) 1.5 and (b) 3.0 T within 10 minutes. Images are isotropic at b value of 1000 sec/mm2. Note the higher apparent SNR and the improved visibility of ischemic lesions at 3.0 T (arrowheads). Imaging studies in b were obtained without parallel imaging (it was not available at the time this study was performed); accordingly, susceptibility-related image distortions are more pronounced in b than a.

Susceptibility-based and Perfusion Imaging
The stronger susceptibility effects at 3.0 T can be exploited for diagnostic purposes, for example, to identify anatomic structures with physiologically increased iron content, such as the substantia nigra (68). This may be used for the early diagnosis of neurodegenerative disorders of, for example, the Parkinson type—yet so far, there are no published data on the clinical use of such an approach.

The stronger susceptibility effects at 3.0 T may also be used to improve the detection of hemorrhage. In fact, it has been shown that acute and subacute hemorrhage is associated with a substantially stronger signal void at 3.0 T than at 1.5 T. Unfortunately, however, the only study, to our knowledge, that focused on that issue dealt with gross intracerebral hematomas only (69). From a clinical standpoint, it would be much more relevant to investigate whether T2*-weighted 3.0-T MR imaging would help improve detection of lesions invisible at 1.5 T, such as chronic hemorrhage in small cavernomas, after diffuse axonal injury or as a cause of symptomatic focal epilepsy (70).

Another application that exploits susceptibility effects is T2*-weighted perfusion imaging. This technique is used to assess the microvascular perfusion of the brain on a capillary level (71). Among other applications, it is clinically used to identify hypoperfused but still viable tissue in the penumbra around an ischemic infarction, that is, salvageable tissue in patients with acute ischemia. In addition, it is used to assess the relevance of arterial steno-occlusive disease on a tissue perfusion level. Manka and co-workers (72) were able to demonstrate that with an echo-shifted echo-planar imaging pulse sequence (principle of echo shifting with a train of observations, or PRESTO), the relatively short echo time enabled T2*-weighted perfusion imaging with acceptable image quality (72,73) and that because of the higher susceptibility-mediated effects at 3.0 T, half of the recommended 1.5-T contrast agent dose was sufficient. If parallel imaging is used, whole-brain perfusion with a sampling rate of more than 1 Hz (acquisition time of less than 1 second per whole-brain volume) is attainable—which may further improve the accuracy of perfusion measurements per se. Whether or not T2*-weighted perfusion imaging at 3.0 T will translate into an improved detection of perfusion abnormalities is, however, not established.

Arterial spin labeling (ASL) is another approach to assess tissue perfusion. For ASL, magnetically labeled blood serves as an endogenous contrast agent. ASL does not require an injected paramagnetic contrast agent and provides quantitative information of regional brain perfusion (74). Despite these important technical advantages, ASL has not gained broad clinical acceptance, because the limited SNR and the short persistence of tags at 1.5 T made the acquisition cumbersome and highly time consuming and substantially limited the anatomic coverage. It has been shown that at higher magnetic field strengths (3.0–4.0 T), these technical difficulties are sufficiently settled to allow a reliable ASL-based mapping of cerebral perfusion within clinically acceptable acquisition times (75,76) (Fig 12). This is not only due to the higher SNR, but especially due to the longer persistence of labeled spins secondary to the slower R1 relaxation rates and longer T1 relaxation times at 3.0 T. It remains to be seen whether at 3.0 T ASL will gain ground and will be used more broadly, not only in the brain, but also for perfusion studies of thoracic and abdominal organs.

View larger version (69K): [in this window] [in a new window] [Download PPT slide]   Figure 12: Transverse MR images obtained with arterial spin labeling at 3.0 T (4000/37, 17 sections, 5-mm thickness, 64 x 64 matrix, 700-msec labeling delay, 50 dynamic acquisitions, 6-minute acquisition time) in 66-year-old patient with right-sided frontodorsal grade IV glioma. Note evidence of tumor hypervascularity as revealed by the high perfusion signal of the glioblastoma (arrows).

In 15 patients, a systematic comparison was made between state-of-the-art TOF MR angiography at 1.5 T with voxel size of 0.72 mm³ and a 3.0-T protocol with voxel size of 0.03 mm³ (80) (Figs 13, 14). With the high spatial resolution at 3.0 T, the depiction of small vessels was considerably improved. This allowed visualization of the peripheral segments of the main brain-supplying arteries and of small penetrating vessels (like the lenticulostriate arteries) that were not apparent on the respective 1.5-T TOF MR angiographic studies altogether. The drawback was that these very-high-spatial-resolution studies took more than 8 minutes to acquire—a borderline acquisition time that may not be acceptable in clinical situations. In a later publication, the same group used parallel imaging (SENSE) for high-field-strength TOF MR angiography. The SENSE-mediated gain of sampling speed could be traded both for a reduction of acquisition time and for an increase in the number of sections (81). The latter helps increase the usually limited anatomic coverage of TOF MR angiographic studies (Fig 15).

View larger version (125K): [in this window] [in a new window] [Download PPT slide]   Figure 13a: TOF (inflow) MR angiograms at (a) 1.5 T (28/6.9; 50 sections; acquisition time, 2 minutes 34 seconds; section thickness, 2 mm; field of view, 160 x 160 mm2; matrix, 336 x 212; voxel size, 0.92 mm3) and (b) 3.0 T (26/3.5; 100 sections; acquisition time, 7 minutes 57 seconds; section thickness, 1 mm; field of view, 250 x 250 mm2; matrix, 832 x 571; voxel size, 0.13 mm3) in the same individual. Note the improved vessel conspicuity at 3.0 T, including the ophthalmic arteries and the small penetrating thalamo-striate arteries (arrowheads).

View larger version (108K): [in this window] [in a new window] [Download PPT slide]   Figure 13b: TOF (inflow) MR angiograms at (a) 1.5 T (28/6.9; 50 sections; acquisition time, 2 minutes 34 seconds; section thickness, 2 mm; field of view, 160 x 160 mm2; matrix, 336 x 212; voxel size, 0.92 mm3) and (b) 3.0 T (26/3.5; 100 sections; acquisition time, 7 minutes 57 seconds; section thickness, 1 mm; field of view, 250 x 250 mm2; matrix, 832 x 571; voxel size, 0.13 mm3) in the same individual. Note the improved vessel conspicuity at 3.0 T, including the ophthalmic arteries and the small penetrating thalamo-striate arteries (arrowheads).

View larger version (104K): [in this window] [in a new window] [Download PPT slide]   Figure 14a: TOF MR angiograms at (a) 1.5 T and (b) 3.0 T in a patient with high-grade stenosis of the right internal carotid artery. Note the improved visibility of the perfused right-sided middle cerebral artery in b. Imaging parameters correspond to those in Figure 13.

View larger version (114K): [in this window] [in a new window] [Download PPT slide]   Figure 14b: TOF MR angiograms at (a) 1.5 T and (b) 3.0 T in a patient with high-grade stenosis of the right internal carotid artery. Note the improved visibility of the perfused right-sided middle cerebral artery in b. Imaging parameters correspond to those in Figure 13.

View larger version (55K): [in this window] [in a new window] [Download PPT slide]   Figure 15a: Images in a 36-year-old male patient with symptomatic stenosis of the internal carotid artery in the petrous segment (arrow). (a) TOF MR angiogram at 3.0 T (26/3.5; 100 sections; acquisition time, 7 minutes 57 seconds; section thickness, 1 mm; field of view, 250 x 250 mm2; matrix, 832 x 1024; voxel size, 0.13 mm3; number of slabs, one; SAR, 1.6 W/kg) suggests a complete vessel occlusion. (b) TOF MR angiogram at 3.0 T with SENSE (31/3.5; 150 sections; acquisition time, 5 minutes 12 seconds; section thickness, 1 mm; field of view, 250 x 250 mm2; matrix, 832 x 1024; voxel size, 0.13 mm3; SENSE factor, 2.5; number of slabs, three; SAR, 0.3 W/kg) depicts the lumen as at least partially patent, indicating high-grade stenosis. Because of broader anatomic coverage, it also includes the M2 and M3 segments, revealing residual hemorrhagic infarction (arrowheads).

View larger version (83K): [in this window] [in a new window] [Download PPT slide]   Figure 15b: Images in a 36-year-old male patient with symptomatic stenosis of the internal carotid artery in the petrous segment (arrow). (a) TOF MR angiogram at 3.0 T (26/3.5; 100 sections; acquisition time, 7 minutes 57 seconds; section thickness, 1 mm; field of view, 250 x 250 mm2; matrix, 832 x 1024; voxel size, 0.13 mm3; number of slabs, one; SAR, 1.6 W/kg) suggests a complete vessel occlusion. (b) TOF MR angiogram at 3.0 T with SENSE (31/3.5; 150 sections; acquisition time, 5 minutes 12 seconds; section thickness, 1 mm; field of view, 250 x 250 mm2; matrix, 832 x 1024; voxel size, 0.13 mm3; SENSE factor, 2.5; number of slabs, three; SAR, 0.3 W/kg) depicts the lumen as at least partially patent, indicating high-grade stenosis. Because of broader anatomic coverage, it also includes the M2 and M3 segments, revealing residual hemorrhagic infarction (arrowheads).

For contrast-enhanced 3.0-T MR angiography, the higher SNR, the stronger T1-shortening effect of contrast agents (higher tissue T1 before vs after contrast material), and the more efficient suppression of background tissue improve the vessel-to-background CNR to an extent that allows the use of parallel imaging with higher reduction factors (R). Of note, higher reduction factors are in turn only technically feasible if dedicated multielement coil arrays are used, which, as a side effect, also add to the higher preexisting SNR. The substantial reduction of acquisition time allows one to sample higher imaging matrices—especially useful for large-field-of-view applications such as abdominal and head and neck applications—and still image fast enough to stay within the arterial phase of the contrast bolus passage. Parallel image acquisitions can be combined with advanced reordering k-space sampling techniques (84–87) that allow one to extend the data sampling into the late arterial and even the venous phase. A further reduction of acquisition time is possible with techniques that reduce the size of k-space to be sampled, such as keyhole or time-resolved imaging of contrast kinetics, or TRICKS (88). If these different acceleration strategies are used in concert, contrast-enhanced MR angiography can be performed in a dynamic, time-resolved fashion—this is new compared to what has been attainable with 1.5-T MR angiography (Fig 16). Being able to observe the passage of blood flow is required for the assessment of cross-filling or reverse flow in unilateral occlusion syndromes or to assess cerebral arteriovenous malformations. In arteriovenous malformation, adequate treatment planning requires the identification of feeders and deep and superficial drainage pathways, which is only feasible with a time-resolved angiography. The required subsecond temporal resolution has been achieved by using two-dimensional "thick-slab" DSA-like angiograms with sampling rates of up to 6 frames per second (89,90). In a prospective study on 18 patients with cerebral arteriovenous malformation (90), time-resolved three-dimensional contrast-enhanced MR angiography at 3 T was used and compared with DSA for classification of arteriovenous malformation. Authors showed that there was perfect agreement between Spetzler-Martin classification of arteriovenous malformation based on four-dimensional MR angiography compared with DSA. Four-dimensional MR angiography missed, however, additional arterial feeders in three (17%) of 18 patients (Table 3).

View larger version (42K): [in this window] [in a new window] [Download PPT slide]   Figure 16a: Time-resolved three-dimensional dynamic MR angiography with contrast-enhanced timing robust MR angiography, or CENTRA, and four-dimensional time-resolved angiography using keyhole (4D-TRAK) in a patient with large frontal arteriovenous malformation allows depiction of (a, c) arterial feeders and (b, d) deep venous drainage through the internal cerebral vein and the vein of Rosenthal; superficial drainage via the superior sagittal sinus is in agreement with DSA findings (bottom row). (a, b) Coronal maximum intensity projections (top row) and (c, d) sagittal maximum intensity projections of (a, c) early arterial and (b, d) early venous phase of 4D-TRAK (top row). A keyhole diameter of 16% (ie, an acceleration factor of six), SENSE with a reduction factor of eight, and 25% half-Fourier imaging (ie, acceleration factor of 1.25) yielded a total acceleration factor of 60 (6 x 8 x 1.25) and an acquisition time of 608 msec per dynamic frame. Each dynamic frame consisted of 140 sections with an acquired voxel size of 1.69 (1.1 x 1.4 x 1.1) mm3.

View larger version (51K): [in this window] [in a new window] [Download PPT slide]   Figure 16b: Time-resolved three-dimensional dynamic MR angiography with contrast-enhanced timing robust MR angiography, or CENTRA, and four-dimensional time-resolved angiography using keyhole (4D-TRAK) in a patient with large frontal arteriovenous malformation allows depiction of (a, c) arterial feeders and (b, d) deep venous drainage through the internal cerebral vein and the vein of Rosenthal; superficial drainage via the superior sagittal sinus is in agreement with DSA findings (bottom row). (a, b) Coronal maximum intensity projections (top row) and (c, d) sagittal maximum intensity projections of (a, c) early arterial and (b, d) early venous phase of 4D-TRAK (top row). A keyhole diameter of 16% (ie, an acceleration factor of six), SENSE with a reduction factor of eight, and 25% half-Fourier imaging (ie, acceleration factor of 1.25) yielded a total acceleration factor of 60 (6 x 8 x 1.25) and an acquisition time of 608 msec per dynamic frame. Each dynamic frame consisted of 140 sections with an acquired voxel size of 1.69 (1.1 x 1.4 x 1.1) mm3.

View larger version (79K): [in this window] [in a new window] [Download PPT slide]   Figure 16c: Time-resolved three-dimensional dynamic MR angiography with contrast-enhanced timing robust MR angiography, or CENTRA, and four-dimensional time-resolved angiography using keyhole (4D-TRAK) in a patient with large frontal arteriovenous malformation allows depiction of (a, c) arterial feeders and (b, d) deep venous drainage through the internal cerebral vein and the vein of Rosenthal; superficial drainage via the superior sagittal sinus is in agreement with DSA findings (bottom row). (a, b) Coronal maximum intensity projections (top row) and (c, d) sagittal maximum intensity projections of (a, c) early arterial and (b, d) early venous phase of 4D-TRAK (top row). A keyhole diameter of 16% (ie, an acceleration factor of six), SENSE with a reduction factor of eight, and 25% half-Fourier imaging (ie, acceleration factor of 1.25) yielded a total acceleration factor of 60 (6 x 8 x 1.25) and an acquisition time of 608 msec per dynamic frame. Each dynamic frame consisted of 140 sections with an acquired voxel size of 1.69 (1.1 x 1.4 x 1.1) mm3.

View larger version (76K): [in this window] [in a new window] [Download PPT slide]   Figure 16d: Time-resolved three-dimensional dynamic MR angiography with contrast-enhanced timing robust MR angiography, or CENTRA, and four-dimensional time-resolved angiography using keyhole (4D-TRAK) in a patient with large frontal arteriovenous malformation allows depiction of (a, c) arterial feeders and (b, d) deep venous drainage through the internal cerebral vein and the vein of Rosenthal; superficial drainage via the superior sagittal sinus is in agreement with DSA findings (bottom row). (a, b) Coronal maximum intensity projections (top row) and (c, d) sagittal maximum intensity projections of (a, c) early arterial and (b, d) early venous phase of 4D-TRAK (top row). A keyhole diameter of 16% (ie, an acceleration factor of six), SENSE with a reduction factor of eight, and 25% half-Fourier imaging (ie, acceleration factor of 1.25) yielded a total acceleration factor of 60 (6 x 8 x 1.25) and an acquisition time of 608 msec per dynamic frame. Each dynamic frame consisted of 140 sections with an acquired voxel size of 1.69 (1.1 x 1.4 x 1.1) mm3.

Parallel imaging–accelerated contrast-enhanced MR angiography has been used to depict renal arteries in a total 44 healthy volunteers and four patients suspected of having renovascular disease (93–95). Despite susceptibility, motion, and dielectric effects, both groups achieved a high image quality in the majority of subjects. Dielectric effects were observed in six of 24 examinations in the series by Kramer et al (93), but did not impair the assessment of the renal arteries. Since no comparison with 1.5 T was obtained in this study, the added value of renal angiography at 3.0 T remains unclear. This is even more so because the spatial resolution of 3.0-T MR angiography used in two studies was 0.65 mm³ and 1.08 mm³ in 22- and 16-second breath-hold time, respectively (95,96). Although this is a good starting point for high-field-strength MR angiography, this is probably not a substantial improvement compared with what is achievable if the same level of advanced imaging is used at 1.5 T. Also at 1.5 T, parallel imaging allows contrast-enhanced MR angiography of the renal arteries with high spatial resolution (voxel sizes as small as 1.30 mm³) to be obtained within a 22-second breath hold and allows the visualization of renal artery segments up to the third generation (97). Data on the clinical use of high-field-strength contrast-enhanced renal artery MR angiography are lacking.

Peripheral MR angiography of runoff vessels with a moving table acquisition can be performed at 3.0 T even by using the system's built-in body coil for image generation. Diehm and co-workers (98) recently published a well-designed clinical study on 10 patients (15 limbs) with arterial steno-occlusive disease, with 1.5 T comparison and DSA as the standard of reference. They did, however, find no significant difference regarding image quality or the diagnostic accuracy of contrast-enhanced MR angiography of the infrapopliteal vessels—despite a substantial SNR increase at 3.0 T.

	CONCLUSION

TOP ABSTRACT INTRODUCTION PHYSICAL EFFECTS OF HIGHER... HOW TO DEAL WITH... CLINICAL APPLICATIONS CONCLUSION ESSENTIALS References

 
The physical changes that are associated with high-field-strength imaging are heterogeneous with regard to their effect on clinical imaging. The higher SNR at 3.0 T is clearly advantageous and the increased RF absorption in tissues at higher field strengths is disadvantageous—all other effects, however, can be both advantageous or disadvantageous, depending on the desired image contrast and/or the intended clinical application.

Whether or not in a given clinical situation the high-field-specific advantages prevail will depend on the appropriate adaptation of pulse sequence parameters and on the efficiency with which artifacts are avoided or compensated. In the majority of clinical scenarios, the high-field-strength SNR advantage will at least in part be offset by high-field-strength–specific physical effects that reduce SNR per se (eg, longer T1 relaxation times, susceptibility or dielectric effects) or that may require the use of acquisition strategies that will consume part of the SNR advantage (eg, parallel imaging). This means that in most clinical situations, the net SNR gain will be less than a factor of two compared with 1.5 T.

Although for most neurologic and angiographic applications 3.0 T yields technical advantages compared to 1.5 T, the evidence regarding the added clinical value of high-field-strength MR is very limited. There is no paucity of articles that focus on the technical evaluation of neurologic or angiographic applications at 3.0 T. This technology-driven science absorbs a lot of time and energy—energy that is not available for research on the actual clinical utility of high-field-strength MR imaging. One may look at it on the bright side and argue that 3.0-T MR is already mature enough to offer all neuroradiologic and angiographic applications that are required in daily clinical practice and launch it as the new "clinical routine" (99). However, in the absence of scientific proof for the clinical superiority of 3.0 T versus 1.5 T in this era of cost containment, it would be desirable to redirect research efforts to address the many existing clinical problems and investigate whether the additional costs associated with 3.0 T are indeed justifiable.

The second part of this two-part series, to appear in the next issue of Radiology, will present the current level of evidence regarding high-field-strength MR for cardiac, breast, abdominopelvic, musculoskeletal, pediatric, and spectroscopic applications.

	ESSENTIALS

TOP ABSTRACT INTRODUCTION PHYSICAL EFFECTS OF HIGHER... HOW TO DEAL WITH... CLINICAL APPLICATIONS CONCLUSION ESSENTIALS References

 

• With increasing magnetic field strength, some physical features such as signal-to-noise ratio (SNR), resonance frequencies, T1 and T2/T2* relaxation times, chemical shift, in-phase periodicity, radiofrequency (RF) deposition, and dielectric resonance effects change.
  
• It is possible to compensate for high field specific physical features (and the emerging technical difficulties) by using pulse sequences with higher receiver bandwidth, by adapting repetition and echo times settings and/or choosing different types of pulse sequences, by applying high-order shimming, and, most important, by using parallel imaging and flip angle–refocusing techniques.
  
• High field strength and parallel imaging are truly complementary and synergistic techniques; high-field-strength MR requires parallel imaging to correct for high-field-strength–specific technical disadvantages (stronger susceptibility effects, stronger RF absorption); in turn, the broader clinical application of parallel imaging requires a high magnetic field to compensate for the parallel imaging–associated SNR penalty.
  
• Evidence regarding structural MR of the brain, diffusion-weighted and diffusion-tensor imaging, and MR angiography suggest technical superiority of high-field-strength MR compared with that at 1.5 T, but evidence regarding an actual clinical benefit is scarce.
  

	FOOTNOTES

 

Abbreviations: CNR = contrast-to-noise ratio • DSA = digital subtraction angiography • DWI = diffusion-weighted imaging • RF = radiofrequency • SAR = specific absorption rate • SE = spin echo • SENSE = sensitivity encoding • SNR = signal-to-noise ratio • TOF = time of flight

Authors stated no financial relationship to disclose.

	References

TOP ABSTRACT INTRODUCTION PHYSICAL EFFECTS OF HIGHER... HOW TO DEAL WITH... CLINICAL APPLICATIONS CONCLUSION ESSENTIALS References

 

1. U.S. Department of Health and Human Services, Food and Drug Administration, Center for Devices and Radiological Health, Radiological Devices Branch, Division of Reproductive, Abdominal, Ear, Nose and Throat, and Radiological Devices, Office of Device Evaluation. Guidance for industry: guidance for the submission of premarket notifications for magnetic resonance diagnostic devices. http://www.fda.gov/cdrh/ode/mri340.pdf. Updated November 14, 1998. Accessed January 3, 2008.
2. U.S. Department of Health and Human Services, Food and Drug Administration, Center for Devices and Radiological Health, Radiological Devices Branch, Division of Reproductive, Abdominal, and Radiological Devices, Office of Device Evaluation. Guidance for industry and FDA staff: criteria for significant risk investigations of magnetic resonance diagnostic devices. http://www.fda.gov/cdrh/ode/guidance/793.html. Updated July 14, 2003. Accessed January 3, 2008.
3. Kuhl CK, Traeber F, Gieseke J, et al. Whole-body high-field-strength (3.0-T) MR imaging in clinical practice. II. Technical considerations and clinical applications. Radiology (in press).
4. Hoult DI, Phil D. Sensitivity and power deposition in a high-field imaging experiment. J Magn Reson Imaging 2000;12:46–67.[CrossRef][Medline]
5. Wen H, Dension TJ, Singerman RW, et al. The intrinsic signal-to-noise ratio in human cardiac imaging at 1.5, 3 and 4 T. J Magn Reson 1997;125:65–71.[CrossRef][Medline]
6. Larmor SJ. On the essence of physical relativity. Proc Natl Acad Sci U S A 1918;4:334–337.[Free Full Text]
7. Gold GE, Han E, Stainsby J, Wright G, Brittain J, Beaulieu C. Musculoskeletal MRI at 3.0 T: relaxation times and image contrast. AJR Am J Roentgenol 2004;183:343–351.[Abstract/Free Full Text]
8. Duewell SH, Ceckler TL, Ong K, et al. Musculoskeletal MR imaging at 4 T and at 1.5 T: comparison of relaxation times and image contrast. Radiology 1995;196:551–555.[Abstract/Free Full Text]
9. Peh WC, Chan JH. Artifacts in musculoskeletal magnetic resonance imaging: identification and correction. Skeletal Radiol 2001;30:179–191.[CrossRef][Medline]
10. Fischer HW, Rinck PA, Van Haverbeke Y, Muller RN. Nuclear relaxation of human brain gray and white matter: analysis of field dependence and implications for MRI. Magn Reson Med 1990;16(2):317–334.[Medline]
11. Lin C, Bernstein MA, Huston J, Fain SB. In-vivo and in-vitro measurements of T1 relaxation at 3.0T [abstr]. In: Proceedings of the Ninth Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2001; 1685.
12. Traber F, Block W, Lamerichs R, Gieseke J, Schild HH. 1H metabolite relaxation times at 3.0 tesla: measurements of T1 and T2 values in normal brain and determination of regional differences in transverse relaxation. J Magn Reson Imaging 2004;19:537–545.[CrossRef][Medline]
13. 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]
14. Stanisz GJ, Odrobina EE, Pun J, et al. T1, T2 relaxation and magnetization transfer in tissue at 3T. Magn Reson Med 2005;54:507–512.[CrossRef][Medline]
15. Williams LA, Gelman N, Picot PA, et al. Neonatal brain: regional variability of in vivo MR imaging relaxation rates at 3.0 T—initial experience. Radiology 2005;235:595–603.[Abstract/Free Full Text]
16. Lu H, Clingman C, Golay X, van Zijl PC. Determining the longitudinal relaxation time (T1) of blood at 3.0 Tesla. Magn Reson Med 2004;52:679–682.[CrossRef][Medline]
17. Weigel M, Hennig J. Contrast behavior and relaxation effects of conventional and hyperecho-turbo spin echo sequences at 1.5 and 3 T. Magn Reson Med 2006;55:826–835.[CrossRef][Medline]
18. Rakow-Penner R, Daniel B, Yu H, Sawyer-Glover A, Glover GH. Relaxation times of breast tissue at 1.5T and 3T measured using IDEAL. J Magn Reson Imaging 2006;23:87–91.[CrossRef][Medline]
19. Runge VM, Patel MC, Baumann SS, et al. T1-weighted imaging of the brain at 3 tesla using a 2-dimensional spoiled gradient echo technique. Invest Radiol 2006;41:68–75.[CrossRef][Medline]
20. Rinck PA, Muller RN. Field strength and dose dependence of contrast enhancement by gadolinium-based MR contrast agents. Eur Radiol 1999;9:998–1004.[CrossRef][Medline]
21. 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]
22. Ba-Ssalamah A, Nobauer-Huhmann IM, Pinker K, et al. Effect of contrast dose and field strength in the magnetic resonance detection of brain metastases. Invest Radiol 2003;38:415–422.[CrossRef][Medline]
23. 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]
24. Kuhl CK, Jost P, Morakkabati N, Zivanovic O, Schild H, Gieseke J. Contrast-enhanced MR imaging of the breast at 3.0 and 1.5 T in the same patients: initial experience. Radiology 2006;239:666–676.[Abstract/Free Full Text]
25. Kuhl CK, Kooijman H, Gieseke J, Schild HH. Effect of B1 inhomogeneity on breast MR imaging at 3.0 T. Radiology 2007;244:929–930.[Free Full Text]
26. Dahnke H, Schaeffter T. Limits of detection of SPIO at 3.0 T using T2 relaxometry. Magn Reson Med 2005;53:1202–1206.[CrossRef][Medline]
27. Simon GH, Bauer J, Saborovski O, et al. T1 and T2 relaxivity of intracellular and extracellular USPIO at 1.5T and 3T clinical MR scanning. Eur Radiol 2006;16:738–745.[CrossRef][Medline]
28. Chang JM, Lee JM, Lee MW, et al. Superparamagnetic iron oxide-enhanced liver magnetic resonance imaging: comparison of 1.5 T and 3.0 T imaging for detection of focal malignant liver lesions. Invest Radiol 2006;41:168–174.[CrossRef][Medline]
29. von Falkenhausen M, Meyer C, Lutterbey G, et al. Intra-individual comparison of image contrast in SPIO-enhanced liver MRI at 1.5T and 3.0T. Eur Radiol 2007;17:1256–1261.[CrossRef][Medline]
30. Hoenig K, Kuhl CK, Scheef L. Functional 3.0-T MR assessment of higher cognitive function: are there advantages over 1.5-T imaging? Radiology 2005;234(3):860–868.[Abstract/Free Full Text]
31. Girnus R, Träber F, Gieseke J, et al. Individual prediction of a temperature rise of guidewires during MR-examinations using rf-field measurements: a comparison of 1.5 and 3 T [abstr]. In: Proceedings of the 12th Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2004; 663.
32. Wang FN, Huang TY, Lin FH, et al. PROPELLER EPI: an MRI technique suitable for diffusion tensor imaging at high field strength with reduced geometric distortions. Magn Reson Med 2005;54:1232–1240.[CrossRef][Medline]
33. Forbes KP, Pipe JG, Bird CR, Heiserman JE. PROPELLER MRI: clinical testing of a novel technique for quantification and compensation of head motion. J Magn Reson Imaging 2001;14:215–222.[CrossRef][Medline]
34. Sodickson DK, Manning WJ. Simultaneous acquisition of spatial harmonics (SMASH): fast imaging with radiofrequency coil arrays. Magn Reson Med 1997;38:591–603.[Medline]
35. Pruessmann KP, Weiger M, Scheidegger MB, Boesiger P. SENSE: sensitivity encoding for fast MRI. Magn Reson Med 1999;42:952–962.[CrossRef][Medline]
36. Griswold MA, Jakob PM, Heidemann RM, et al. Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn Reson Med 2002;47:1202–1210.[CrossRef][Medline]
37. Bammer R, Keeling SL, Augustin M, et al. Improved diffusion-weighted single-shot echo-planar imaging (EPI) in stroke using sensitivity encoding (SENSE). Magn Reson Med 2001;46:548–554.[CrossRef][Medline]
38. Kuhl CK, Gieseke J, von Falkenhausen M, et al. Sensitivity encoding for diffusion-weighted MR imaging at 3.0 T: intraindividual comparative study. Radiology 2005;234:517–526.[Abstract/Free Full Text]
39. Hennig J, Scheffler K. Hyperechoes. Magn Reson Med 2001;46:6–12.[CrossRef][Medline]
40. Gieseke J, Kuhl CK, von Falkenhausen M, et al. Ultra fast T2-weighted TSE sequences with flip angle sweep and SENSE at 3T [abstr]. In: Proceedings of the 12th Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2004; 74.
41. Mugler JP, Bao S, Mulkern RV, et al. Optimized single-slab three-dimensional spin-echo MR imaging of the brain. Radiology 2000;216:891–899.[Abstract/Free Full Text]
42. Runge VM, Case RS, Sonnier HL. Advances in clinical 3-tesla neuroimaging. Invest Radiol 2006;41:63–67.[CrossRef][Medline]
43. Morakkabati-Spitz N, Gieseke J, Kuhl C, et al. MRI of the pelvis at 3 T: very high spatial resolution with sensitivity encoding and flip-angle sweep technique in clinically acceptable scan time. Eur Radiol 2006;16:634–641.[CrossRef][Medline]
44. Morakkabati-Spitz N, Gieseke J, Kuhl C, et al. Female pelvis: MR imaging at 3.0 T with sensitivity encoding and flip-angle sweep technique. Radiology 2006;241:538–545.[Abstract/Free Full Text]
45. Vaughan JT, Adriany G, Snyder CJ, et al. Efficient high-frequency body coil for high-field MRI. Magn Reson Med 2004;52:851–859.[CrossRef][Medline]
46. Graf H, Lauer UA, Schick F. Eddy-current induction in extended metallic parts as a source of considerable torsional moment. J Magn Reson Imaging 2006;23:585–590.[CrossRef][Medline]
47. Schenck JF. Safety of strong, static magnetic fields. J Magn Reson Imaging 2000;12:2–19.[CrossRef][Medline]
48. Born M, Tschampa HJ, Willinek WA, Textor HJ, Kuhl CK, Schild HH. Prevalence of side effects of MRI at clinical 3 Tesla: intraindividual comparison to 1.5 T [abstr]. Radiology 2002;225(P):375.
49. Sommer T, Maintz D, Schmiedel A, et al. High field MR imaging: magnetic field interactions of aneurysm clips, coronary artery stents and iliac artery stents with a 3.0 Tesla MR system [in German]. Rofo 2004;176:731–738.[Medline]
50. Majoie CB, Sprengers ME, van Rooij WJ, et al. MR angiography at 3T versus digital subtraction angiography in the follow-up of intracranial aneurysms treated with detachable coils. AJNR Am J Neuroradiol 2005;26:1349–1356.[Abstract/Free Full Text]
51. Ugurbil K, Adriany G, Andersen P, et al. Ultrahigh field magnetic resonance imaging and spectroscopy. Magn Reson Imaging 2003;21:1263–1281.[CrossRef][Medline]
52. Keiper MD, Grossman RI, Hirsch JA, et al. MR identification of white matter abnormalities in multiple sclerosis: a comparison between 1.5 T and 4 T. AJNR Am J Neuroradiol 1998;19:1489–1493.[Abstract]
53. Sicotte NL, Voskuhl RR, Bouvier S, Klutch R, Cohen MS, Mazziotta JC. Comparison of multiple sclerosis lesions at 1.5 and 3.0 Tesla. Invest Radiol 2003;38:423–427.[CrossRef][Medline]
54. Bachmann R, Reilmann R, Schwindt W, Kugel H, Heindel W, Kramer S. FLAIR imaging for multiple sclerosis: a comparative MR study at 1.5 and 3.0 Tesla. Eur Radiol 2006;16:915–921.[CrossRef][Medline]
55. Nielsen K, Rostrup E, Frederiksen JL, et al. Magnetic resonance imaging at 3.0 tesla detects more lesions in acute optic neuritis than at 1.5 tesla. Invest Radiol 2006;41:76–82.[CrossRef][Medline]
56. Wattjes MP, Harzheim M, Kuhl CK, et al. Does high-field MR imaging have an influence on the classification of patients with clinically isolated syndromes according to current diagnostic MR imaging criteria for multiple sclerosis? AJNR Am J Neuroradiol 2006;27(8):1794–1798.[Abstract/Free Full Text]
57. Wattjes MP, Lutterbey G, Gieseke J, et al. Double inversion recovery brain imaging at 3T: diagnostic value in the detection of multiple sclerosis lesions. AJNR Am J Neuroradiol 2007;28(1):54–59.[Abstract/Free Full Text]
58. Wattjes MP, Lutterbey G, Harzheim M, et al. Imaging of inflammatory lesions at 3.0 Tesla in patients with clinically isolated syndromes suggestive of multiple sclerosis: a comparison of fluid-attenuated inversion recovery with T2 turbo spin-echo. Eur Radiol 2006;16:2067–2073.[CrossRef][Medline]
59. Barkhof F, Rocca M, Francis G, et al. Validation of diagnostic magnetic resonance imaging criteria for multiple sclerosis and response to interferon beta 1a. Ann Neurol 2003;53:718–724.[CrossRef][Medline]
60. Minneboo A, Barkhof F, Polman CH, Uitdehaag BM, Knol DL, Castelijns JA. Infratentorial lesions predict long-term disability in patients with initial findings suggestive of multiple sclerosis. Arch Neurol 2004;61:217–221.[Abstract/Free Full Text]
61. Knake S, Triantafyllou C, Wald LL, et al. 3T phased array MRI improves the presurgical evaluation in focal epilepsies: a prospective study. Neurology 2005;65:1026–1031.[Abstract/Free Full Text]
62. Von Oertzen J, Urbach H, Jungbluth S, et al. Standard magnetic resonance imaging is inadequate for patients with refractory focal epilepsy. J Neurol Neurosurg Psychiatry 2002;73:643–647.[Abstract/Free Full Text]
63. Lutterbey G, Wattjes MP, Kandyba J, et al. Clinical evaluation of a speed optimized T2 weighted fast spin echo sequence at 3.0 T using variable flip angle refocusing, half-Fourier acquisition and parallel imaging. Br J Radiol 2007;80:668–673.[Abstract/Free Full Text]
64. Kuhl CK, Textor J, Gieseke J, et al. Acute and subacute ischemic stroke at high-field-strength (3.0-T) diffusion-weighted MR imaging: intraindividual comparative study. Radiology 2005;234:509–516.[Abstract/Free Full Text]
65. Melhem ER, Mori S, Mukundan G, Kraut MA, Pomper MG, van Zijl P. Diffusion tensor MR imaging of the brain and white matter tractography. AJR Am J Roentgenol 2002;178:3–16.[Free Full Text]
66. Okada T, Miki Y, Fushimi Y, et al. Diffusion-tensor fiber tractography: intraindividual comparison of 3.0-T and 1.5-T MR imaging. Radiology 2006;238:668–678.[Abstract/Free Full Text]
67. Nelles M, Gieseke J, Lachenmayer L, Schild HH, Urbach H. Diffusion tensor pyramidal tractography in patients with anterior choroidal artery infarcts. AJNR Am J Neuroradiol (in press).
68. Allkemper T, Schwindt W, Maintz D, Heindel W, Tombach B. Sensitivity of T2-weighted FSE sequences towards physiological iron depositions in normal brains at 1.5 and 3.0 T. Eur Radiol 2004;14:1000–1004.[CrossRef][Medline]
69. Allkemper T, Tombach B, Schwindt W, et al. Acute and subacute intracerebral hemorrhages: comparison of MR imaging at 1.5 and 3.0 T—initial experience. Radiology 2004;232:874–881.[Abstract/Free Full Text]
70. Sohn CH, Baik SK, Lee HJ, et al. MR imaging of hyperacute subarachnoid and intraventricular hemorrhage at 3T: a preliminary report of gradient echo T2*-weighted sequences. AJNR Am J Neuroradiol 2005;26:662–665.[Abstract/Free Full Text]
71. Patel MR, Siewert B, Warach S, Edelman RR. Diffusion and perfusion imaging techniques. Magn Reson Imaging Clin N Am 1995;3:425–438.[Medline]
72. Manka C, Traber F, Gieseke J, Schild HH, Kuhl CK. Three-dimensional dynamic susceptibility-weighted perfusion MR imaging at 3.0 T: feasibility and contrast agent dose. Radiology 2005;234:869–877.[Abstract/Free Full Text]
73. Thilmann O, Larsson EM, Bjorkman-Burtscher IM, Stahlberg F, Wirestam R. Effects of echo time variation on perfusion assessment using dynamic susceptibility contrast MR imaging at 3 tesla. Magn Reson Imaging 2004;22:929–935.[CrossRef][Medline]
74. Detre JA, Wang J. Technical aspects and utility of fMRI using BOLD and ASL. Clin Neurophysiol 2002;113:621–634.[CrossRef][Medline]
75. Wang J, Alsop DC, Li L, et al. Comparison of quantitative perfusion imaging using arterial spin labeling at 1.5 and 4.0 Tesla. Magn Reson Med 2002;48:242–254.[CrossRef][Medline]
76. Wang J, Aguirre GK, Kimberg DY, Detre JA. Empirical analyses of null-hypothesis perfusion FMRI data at 1.5 and 4 T. Neuroimage 2003;19:1449–1462.[CrossRef][Medline]
77. Bernstein MA, Huston J 3rd, Lin C, Gibbs GF, Felmlee JP. High-resolution intracranial and cervical MRA at 3.0T: technical considerations and initial experience. Magn Reson Med 2001;46:955–962.[CrossRef][Medline]
78. 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;12:183–120.[CrossRef][Medline]
79. Gibbs GF, Huston J 3rd, Bernstein MA, Riederer SJ, Brown RD Jr. Improved image quality of intracranial aneurysms: 3.0-T versus 1.5-T time-of-flight MR angiography. AJNR Am J Neuroradiol 2004;25:84–87.[Abstract/Free Full Text]
80. Willinek WA, Born M, Simon B, et al. Time-of-flight MR angiography: comparison of 3.0-T imaging and 1.5-T imaging—initial experience. Radiology 2003;229:913–920.[Abstract/Free Full Text]
81. Willinek WA, Gieseke J, von Falkenhausen M, et al. Sensitivity encoding (SENSE) for high spatial resolution time-of-flight MR angiography of the intracranial arteries at 3.0 T. Rofo 2004;176:21–26.[Medline]
82. Fushimi Y, Miki Y, Kikuta K, et al. Comparison of 3.0- and 1.5-T three-dimensional time-of-flight MR angiography in moyamoya disease: preliminary experience. Radiology 2006;239:232–237.[Abstract/Free Full Text]
83. Choi CG, Lee DH, Lee JH, et al. Detection of intracranial atherosclerotic steno-occlusive disease with 3D time-of-flight magnetic resonance angiography with sensitivity encoding at 3T. AJNR Am J Neuroradiol 2007;28:439–446.[Abstract/Free Full Text]
84. Huston J, Fain SB, Wald JT, et al. Carotid artery: elliptic centric contrast-enhanced MR angiography compared with conventional angiography. Radiology 2001;218:138–143.[Abstract/Free Full Text]
85. Huston J, Fain SB, Riederer SJ, Wilman AH, Bernstein MA, Busse RF. Carotid arteries: maximizing arterial to venous contrast in fluoroscopically triggered contrast-enhanced MR angiography with elliptic centric view ordering. Radiology 1999;211:265–273.[Abstract/Free Full Text]
86. Wilman AH, Riederer SJ, Huston J, Wald JT, Debbins JP. Arterial phase carotid and vertebral artery imaging in 3D contrast-enhanced MR angiography by combining fluoroscopic triggering with an elliptical centric acquisition order. Magn Reson Med 1998;40:24–35.[Medline]
87. 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(2):583–588.[Abstract/Free Full Text]
88. Korosec FR, Frayne R, Grist TM, et al. Time-resolved contrast-enhanced 3D MR angiography. Magn Reson Med 1996;36:345–351.[Medline]
89. Summers PE, Kollias SS, Valavanis A. Resolution improvement in thick-slab magnetic resonance digital subtraction angiography using SENSE at 3T. J Magn Reson Imaging 2004;20:662–673.[CrossRef][Medline]
90. Hadizadeh D, von Falkenhausen M, Gieseke J, et al. Cerebral arteriovenous malformation: Spetzler-Martin classification at subsecond-temporal-Rrsolution four-dimensional MR angiography compared with that at DSA. Radiology 2008; 246:205–213.[CrossRef][Medline]
91. Nael K, Ruehm SG, Michaely HJ, et al. High spatial-resolution CE-MRA of the carotid circulation with parallel imaging: comparison of image quality between 2 different acceleration factors at 3.0 Tesla. Invest Radiol 2006;41:391–399.[CrossRef][Medline]
92. Willinek WA, von Falkenhausen M, Born M, et al. Noninvasive detection of steno-occlusive disease of the supra-aortic arteries with three-dimensional contrast-enhanced magnetic resonance angiography: a prospective, intra-individual comparative analysis with digital subtraction angiography. Stroke 2005;36:38–43.[Abstract/Free Full Text]
93. Kramer U, Nael K, Laub G, et al. High-resolution magnetic resonance angiography of the renal arteries using parallel imaging acquisition techniques at 3.0 T: initial experience. Invest Radiol 2006;41:125–132.[CrossRef][Medline]
94. Nael K, Villablanca JP, Pope WB, McNamara TO, Laub G, Finn JP. Supraaortic arteries: contrast-enhanced MR angiography at 3.0 T—highly accelerated parallel acquisition for improved spatial resolution over an extended field of view. Radiology 2007;242:600–609.[Abstract/Free Full Text]
95. 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]
96. Michaely HJ, Kramer H, Dietrich O, et al. Intraindividual comparison of high-spatial-resolution abdominal MR angiography at 1.5 T and 3.0 T: initial experience. Radiology 2007;244:907–913.[Abstract/Free Full Text]
97. Born M, Willinek WA, Gieseke J, von Falkenhausen M, Schild H, Kuhl CK. Sensitivity encoding (SENSE) for contrast-enhanced 3D MR angiography of the abdominal arteries. J Magn Reson Imaging 2005;22:559–565.[CrossRef][Medline]
98. Diehm N, Kickuth R, Baumgartner I, et al. Magnetic resonance angiography in infrapopliteal arterial disease: prospective comparison of 1.5 and 3 Tesla magnetic resonance imaging. Invest Radiol 2007;42:467–476.[CrossRef][Medline]
99. Runge VM, Schonberg SO. Advances in 3 Tesla MRI: redefining imaging for clinical routine. Invest Radiol 2007;42(7):485.[CrossRef]
100. Frydrychowicz A, Bley TA, Winterer JT, et al. Accelerated time-resolved 3D contrast-enhanced MR angiography at 3T: clinical experience in 31 patients. MAGMA 2006:19:187–195.[CrossRef][Medline]
101. Markl M, Uhl M, Wieben O, et al. High resolution 3T MRI for the assessment of cervical and superficial cranial arteries in giant cell arteritis. J Magn Reson Imaging 2006;24:423–427.[CrossRef][Medline]
102. Fenchel M, Nael K, Ruehm S, et al. Isotropic high spatial resolution magnetic resonance angiography of the supra-aortic arteries using two-dimensional parallel imaging (iPAT2) at 3 Tesla: a feasibility study. Invest Radiol 2006;41:545–552.[CrossRef][Medline]
103. Cashen TA, Carr JC, Shin W, et al. Intracranial time-resolved contrast-enhanced MR angiography at 3T. AJNR Am J Neuroradiol 2006;27:822–829.[Abstract/Free Full Text]
104. Fenchel M, Nael K, Deshpande VS, et al. Renal magnetic resonance angiography at 3.0 Tesla using a 32-element phased-array coil system and parallel imaging in 2 directions. Invest Radiol 2006;41:697–703.[CrossRef][Medline]

This article has been cited by other articles:

				F. G. Shellock and A. Spinazzi MRI Safety Update 2008: Part 2, Screening Patients for MRI Am. J. Roentgenol., October 1, 2008; 191(4): 1140 - 1149. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kuhl, C. K. Articles by Schild, H. H. Search for Related Content PubMed PubMed Citation Articles by Kuhl, C. K. Articles by Schild, H. H.