| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Neuroradiology |
1 From the Department of Radiology, University of Bonn, Sigmund-Freud-Strasse 25, D-53105 Bonn, Germany. Received September 29, 2004; revision requested December 2; revision received December 30; accepted January 25, 2005. Address correspondence to C.K. (Carsten.Krautmacher{at}ukb.uni-bonn.de).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Institutional review board approval and informed consent were obtained for all patient and volunteer studies. Twelve patients (six women, six men; mean age, 58 years; range, 29–76 years) with 12 enhancing brain lesions (11 patients with primary brain tumors and one with a solitary cerebral metastasis) underwent contrast material–enhanced MR imaging three times, on three separate days: once at 1.5 T with a full dose of 0.10 mmol/kg gadopentetate dimeglumine, once at 3.0 T with a full dose, and once at 3.0 T with half that dose, 0.05 mmol/kg. The same contrast-enhanced T1-weighted spin-echo images (repetition time msec/echo time msec, 500/12; section thickness, 5 mm; matrix, 256 x 205) were obtained at both 3.0 T and 1.5 T after prior optimization of parameters at 3.0 T. The number and conspicuity of enhancing brain lesions were assessed with blinded clinical image reading. Signal-to-noise ratio and CNR were determined with region of interest analysis of enhancing lesions and normal contralateral white matter. For 3.0 T with half the standard dose and with the full dose, CNR of lesions was intraindividually compared with CNR at 1.5 T with the full dose by using the Wilcoxon matched-pairs signed rank test.
RESULTS: At 3.0 T and full dose, CNR was 2.8-fold higher than that at 1.5 T and full dose (P < .001). At the same time, higher lesion conspicuity at clinical image reading was observed. With only half the standard dose, MR imaging at 3.0 T still yielded higher CNR (1.3-fold higher) than that with full dose at 1.5 T (P < .01).
CONCLUSION: With the same amount of contrast agent, MR imaging at 3.0 T offered a significantly higher CNR of enhancing cerebral lesions, compared with that at 1.5 T; even with the dose reduced by half, CNR was still higher at 3.0 T.
© RSNA, 2005
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Over the past decade, most clinical experience in the field of cerebral MR imaging has been with 1.5-T systems with a dose of 0.10 mmol/kg gadolinium chelate, as this combination seems to be an acceptable compromise between imaging expense and diagnostic sensitivity (5,6). Recently, the number of 3.0-T systems in clinical settings has been increasing, and systems operating at even higher field strengths are being considered for clinical applications.
One of the main features of MR imaging at 3.0 T is the general gain in signal-to-noise ratio (SNR) compared with that at lower field strengths. Therefore, one may anticipate that the increased SNR associated with a higher magnetic field will translate, at least to a certain degree, into a higher contrast-to-noise ratio (CNR) between enhancing and nonenhancing tissues. The increased CNR should improve the delineation of contrast agent–induced changes and, thus, could increase the sensitivity of detection of such signal intensity changes.
In addition, the effectiveness of the T1-shortening effect of a gadolinium-based contrast agent depends on the baseline T1 relaxation time of local tissue. With the longer baseline T1 relaxation times brought about by a higher magnetic field strength, the T1-shortening effect of gadolinium-based contrast agents will be greater, as the relaxivity of such contrast agents changes only marginally between 1.5-T MR imaging and 3.0-T MR imaging (7,8). Accordingly, the signal intensity changes caused by contrast enhancement observable in T1-weighted images should generally be stronger at 3.0 T than they are at 1.5 T.
The purpose of our study, therefore, was to prospectively and intraindividually compare the effect of MR imaging at a higher magnetic field strength (3.0 T) on CNR at different doses of a T1-shortening contrast agent in patients with contrast-enhancing brain lesions, with 1.5 T MR imaging as a reference standard.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Study Design
In vitro and in vivo preparatory studies were performed to identify the optimum pulse sequence for contrast-enhanced T1-weighted MR imaging at 3.0 T. With the pulse sequence parameters identified with these initial in vitro and in vivo studies, the actual patient study was performed. The study group consisted of 12 patients with contrast-enhancing cerebral lesions. Patients were imaged three times on three separate days: once at 1.5 T with the full standard contrast agent dose of 0.10 mmol/kg, once at 3.0 T with the full standard contrast agent dose of 0.10 mmol/kg, and once at 3.0 T with half of the standard contrast agent dose, 0.05 mmol/kg.
Study Protocol
All studies were performed with clinical 1.5-T and 3.0-T MR imaging units (Intera; Philips Medical Systems, Best, the Netherlands). With both units, birdcage head coils were used. Both units were equipped with actively shielded gradient systems (maximal gradient amplitude, 30 mT/m; slew rate, 150 mT/m/msec). For analysis, image sets were transferred to a postprocessing unit (EasyVision; Philips Medical Systems).
In vitro and in vivo studies for optimization of parameters. —To define a suitable T1-weighted spin-echo (SE) pulse sequence for contrast-enhanced MR imaging at 3.0 T, a phantom with tubes containing a serial dilution of gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ) in distilled water ranging from 0.01 mmol/L (1:50 000) to 50 mmol/L (1:10) was imaged at 3.0 T. T1-weighted SE sequences with repetition times (TRs) of 350, 500, and 800 msec were used. The other imaging parameters were kept the same: echo time, 12 msec; matrix, 256 x 205; section thickness, 5 mm; and intersection gap, 1 mm. Measurements that were based on regions of interest (ROIs) were performed in tube centers and in ghosting-free background noise (diameter range of ROIs, 0.8–1.0 cm). Throughout the entire study, ROIs were manually placed by the same author (C.K.). The pulse sequence that delivered the highest SNR was considered most suitable for contrast-enhanced MR imaging.
To identify a pulse sequence with appropriate T1 contrast for cerebral MR imaging at 3.0 T, we compared three T1-weighted SE pulse sequences with increasing TRs of 350, 500, and 800 msec in five volunteers (one woman, four men; mean age, 33 years; range, 31–37). Inclusion criteria for this volunteer study were absence of a personal or family history of neurologic disorders and absence of relevant clinical symptoms. Imaging parameters were kept constant and equivalent to those used for the in vitro experiments. SNR of white matter and gray matter (cortex) were measured with ROI-based analysis. ROIs were 0.8–1.7 cm2 and were manually placed by one author (C.K.). The means of three independent signal intensity measurements of white matter and cortex, and measurements of noise (standard deviation) in ghosting-free background were used to calculate SNR in white matter (SNRWM) and SNR in the cortex (SNRCX). CNR between white matter and cortex (CNRWM/CX) was calculated as follows: CNRWM/CX = SNRWM – SNRCX.
Images from the three different pulse sequences of different patients were intermixed and presented in random order. Afterward, all images were read independently by two neuroradiologists (C.K.K. and H. J. Textor, both with more than 10 years of experience with cerebral MR imaging). Readers were blinded to imaging parameters. Depending on the presence of artifacts and on the readers' subjective impression of image contrast, image quality was rated with a five-point scale (score 5, excellent gray matter–white matter contrast and no artifacts; score 4, acceptable gray matter–white matter contrast and no artifacts or excellent gray matter–white matter contrast and some artifacts; score 3, acceptable gray matter–white matter contrast and some artifacts; score 2, acceptable gray matter–white matter contrast and substantial artifacts; score 1, unacceptable gray matter–white matter contrast and substantial artifacts). Medians of ratings were calculated.
Clinical study.—We included consecutive patients who were referred to our department for contrast-enhanced MR imaging of the brain prior to radiation therapy or surgery. Patients were included only if they had contrast-enhancing lesions, were not to undergo radiation therapy or surgery within the study period of 1 week, were willing to participate in this study, and were ready to return for two further contrast-enhanced MR imaging studies on two separate days. Twelve patients (six women, six men; mean age, 58 years; range, 29–76 years) met the entrance criteria and were included in the study. Different intra- or extraaxial brain tumors were diagnosed: six gliomas, three meningiomas, one acoustic neuroma, one primary central nervous system lymphoma, and one metastasis from a bronchial carcinoma.
MR imaging.—All patients underwent contrast-enhanced cerebral MR imaging three times, once with the 1.5-T system and twice with the 3.0-T system. At 1.5 T, the full standard dose of 0.10 mmol/kg gadopentetate dimeglumine was injected. For the two examinations at 3.0 T, two different doses were administered: the full dose of 0.10 mmol/kg was administered once, and half of that dose, 0.05 mmol/kg, was administered once. The three examinations were performed at least 24 hours apart (mean interval, 42 hours ± 30 [standard deviation]). All three examinations were performed within a period of 1 week. The order in which patients underwent contrast-enhanced MR imaging at 1.5 T and 3.0 T was randomized, and the order in which investigations with the full dose and the half dose of contrast agent were performed at 3.0 T also was randomized. The contrast agent was administered intravenously by using a power injector (Spectris MR Injection System; Medrad, Indianola, Pa) at a flow rate of 2 mL/sec. Image acquisition was uniformly started 1 minute after the end of contrast agent injection.
At 1.5 T, a standard T1-weighted SE pulse sequence that was routinely used for clinical purposes at our institution was performed. Parameters included TR msec/echo time msec, 500/12; field of view, 230 mm; matrix, 256 x 205; section thickness, 5 mm; and intersection gap, 1 mm. Imaging was performed in the transverse plane.
At 3.0-T MR imaging, parameters that define image geometry (thickness, number of sections, matrix, and imaging plane) were kept identical to those that were used for the 1.5-T setting. Contrast parameters were chosen in accordance with the results of the in vitro and in vivo experiments (as described later). These experiments suggested the use of a T1-weighted pulse sequence with a TR of 500 msec, which is the same as the TR we routinely used for contrast-enhanced MR imaging at 1.5 T.
Data Analysis
Calculation of SNR and CNR.—Image sets were transferred to the postprocessing unit. Images from the three studies performed in each patient were loaded onto the screen. Anatomically corresponding sections that best displayed the enhancing lesions were chosen. With close supervision of a neuroradiologist (H. Textor, with more than 10 years of experience in reading cerebral MR images), ROIs were manually placed to selectively contain contrast-enhancing tumor tissue. The means of three independent signal intensity measurements per lesion were calculated. To assess background noise, ROIs were always placed in the same area of ghosting-free image background, and the standard deviations were recorded. Signal intensity and noise values were used to calculate SNR. To determine CNR of enhancing lesions, the SNR of contralateral nondiseased frontal white matter (SNRWM) was subtracted from the SNR of contrast-enhancing lesions (SNRL) with the following equation: CNRL/WM = SNRL – SNRWM, where CNRL/WM is the CNR between enhancing lesions and cerebral white matter.
Comparison of CNR between enhancing lesions and cerebral white matter of contrast-enhancing lesions at 3.0 and 1.5 T.—To quantify the effect of a higher magnetic field strength on lesion-to-brain CNR in each patient, we calculated a ratio of change of CNR between enhancing lesions and cerebral white matter (rCNR). It represents each patient's CNR between enhancing lesions and cerebral white matter at 3.0 T relative to the same patient's CNR between enhancing lesions and cerebral white matter at 1.5 T, according to the following equation: rCNR = CNRL/WM3.0T ÷ CNRL/WM1.5T.
For each of the 12 patients, the ratio of change of CNR was calculated for the study performed at 3.0 T with the full contrast agent dose (ratio of change of CNR with full dose) and for the study performed at 3.0 T with half the standard contrast agent dose (ratio of change of CNR with half dose). The median of all individual ratios of change of CNR was calculated for imaging with the full dose and for imaging with half the standard dose at 3.0 T.
Image interpretation.—In randomized order, images from all patient examinations were presented to two experienced neuroradiologists (C.K.K. and H. J. Textor, both with more than 10 years of experience with cerebral MR imaging) who were blinded to imaging parameters. Images were viewed with standardized window settings (window level of 800 and window width of 1600) and were read in consensus. Readers recorded the number and size of contrast-enhancing cerebral lesions in all studies. Afterward, all images obtained in an individual patient were presented together to compare the three different imaging studies with respect to lesion conspicuity. Readers rated corresponding image quality with a three-point scale (score 1, lesion not or only faintly visible; score 2, lesion visible with satisfactory or good contrast to surrounding brain tissue; and score 3, lesion visible with excellent contrast to surrounding brain tissue).
Statistical analysis.—Statistical analysis was performed by using software (SPSS, version 10.0.7; SPSS, Chicago, Ill). The Wilcoxon matched-pairs signed rank test was used to test the distribution of CNR between enhancing lesions and cerebral white matter for individual lesions in the setting with 1.5-T imaging and the full dose compared with the setting with 3.0-T imaging and half the dose and compared with the setting with 3.0-T imaging and the full dose. A difference with P < .05 was chosen to indicate statistical significance.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Imaging parameters for T1-weighted MR imaging in vivo.— Results of ROI analyses in five healthy volunteers (Figs 1, 2) showed higher SNR in the cortex, SNR in white matter, and CNR between white matter and cortex obtained by using T1-weighted SE imaging with a TR of 500 and 800 msec compared with a TR of 350 msec. CNR between white matter and cortex at imaging with a TR of 500 and 800 msec was of similar magnitude.
|
|
Clinical Study
In 12 patients, a total of 12 contrast-enhancing lesions were detected (mean diameter, 25 mm; range, 8–55 mm). All of these lesions were diagnosed prospectively in each of the three examinations per patient. Lesion conspicuity was rated highest on the images obtained at the 3.0-T examination with the full contrast agent dose (median score of 3), whereas lesion conspicuity on the images obtained at the 1.5-T examination with the full contrast agent dose and on the images obtained at the 3.0-T examination with half the dose of the contrast agent was rated equivalent (median score of 2 for both settings) (Fig 3).
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Observing a relative CNR gain greater than the increase in magnetic field strength (ie, more than twofold) may be explained in two ways. On the basis of our initial in vitro and in vivo experiments, at 3.0-T MR imaging, we used a pulse sequence with a TR of 500 msec. This result was somewhat unexpected, as the same TR is widely used for routine T1-weighted imaging in clinical practice at 1.5 T. Because of the longer tissue T1 relaxation times at 3.0 T, use of the same TR will yield a relatively stronger T1 weighting at 3.0 T compared with that at 1.5 T. Therefore, when imaging with the same TR at both field strengths, nonenhancing tissues exhibit a lower than twofold SNR increase between 1.5 T and 3.0 T (ie, lower than the increase in field strength itself). Owing to the fact that enhancing tissues typically exhibit very short T1 relaxation times, they will relax almost completely even within a short TR similar to the one used in our study (ie, 500 msec). Enhancing tissues should, therefore, demonstrate a twofold increase of SNR when moving from 1.5- to 3.0-T imaging. This less than twofold increase of SNR of nonenhancing tissues along with the twofold increase of SNR of enhancing tissues will add to the signal intensity difference between enhancing and nonenhancing tissues. Consequently, we observed a more than twofold increase of lesion-to-brain CNR at 3.0 T. Another explanation for the observed high CNR increase is the fact that because of the longer baseline T1 relaxation times at 3.0 T, the T1-shortening effect per unit of gadopentetate dimeglumine is stronger than it is with the shorter baseline T1 TRs that are present at 1.5 T. For practical purposes, we may conclude that, with circumstances described here, CNR increases at least twofold with increasing field strength.
There are two different clinical strategies to exploit this CNR increase: First, the higher CNR can be invested to improve the visualization of subtle disruptions of the blood-brain barrier, thus possibly improving the sensitivity for depiction of small metastases or early inflammatory changes. Second, the higher CNR could, in principle, be traded to reduce the dose of contrast agent for contrast-enhanced brain imaging at 3.0 T. Our study design should allow the investigation of both approaches, although there are some important limitations.
In regard to the first approach, that of a possible improvement in sensitivity, this may be said. In standard clinical settings, a dose of 0.10 mmol/kg gadolinium chelate is generally considered sufficient for contrast-enhanced MR imaging of the brain at 1.5 T (5,6,10), as this dose delivers acceptable diagnostic sensitivity at a reasonable cost. Imaging for certain indications, however, has been shown to benefit from double or triple the dose of contrast agent. Typically, these indications include clinical situations in which the detection of even subtle disruptions in the blood-brain barrier has an effect on patient treatment (4,11–17). Yuh et al (3) found that studies at 1.5 T with triple the dose of contrast agent were more effective in the detection of metastatic brain lesions than were studies with the full dose of gadolinium chelate. In our study, the higher CNR in the study arm with 3.0 T and the full dose of the contrast agent did not translate into an improved diagnostic sensitivity for contrast-enhancing brain lesions: All lesions revealed by using 3.0-T studies with the full dose of contrast agent were also prospectively and independently diagnosed by using 1.5-T studies with the full dose of contrast agent, yet the patients in our study cohort had relatively large and mostly primary brain tumors. Only one patient was included who was evaluated for small brain metastases. Primary intra- and extraaxial brain tumors are likely to grow as solitary nodules. Accordingly, the composition of our study cohort was probably not well suited to investigate the ability of a technique to help in the identification of additional foci of enhancing lesions, simply because the majority of our patients had no such additional lesions. More studies on larger subsets of patients, particularly studies that include patients without enhancing lesions at 1.5-T imaging, will be needed to find out whether the higher CNR at 3.0-T imaging with the full dose of the contrast agent will improve our ability to diagnose more subtle disruptions of the blood-brain barrier (4,12–18).
In regard to the second approach, that of a reduction in the dose of contrast agent, this may be said. By using the reduced contrast agent dose at 3.0-T MR imaging, CNR was still higher compared with that at 1.5-T imaging with the full dose of contrast agent. This suggests that even with half the contrast agent dose, the diagnostic sensitivity in the depiction of disruptions of the blood-brain barrier should be maintained at 3.0 T compared with the 1.5-T standard setting. In fact, all lesions depicted at 1.5 T with the full dose of contrast agent were also diagnosed prospectively and independently at 3.0 T with half the dose of contrast agent; lesion conspicuity was rated to be equivalent. The same limitations as mentioned previously, however, hold true in regard to the interpretation of these results: We had only relatively large tumors in our small study cohort; accordingly, our results cannot be used to generally recommend half the dose of contrast agent for imaging at 3.0 T. Nevertheless, the presented data may encourage researchers to perform further studies to examine whether these initial findings can be confirmed in larger groups of patients with disruptions of the blood-brain barrier. Confirmation of our results with results of future studies would mean a relevant reduction of the direct cost of routine contrast-enhanced cerebral MR imaging.
A possible disadvantage of our study design might have been caused by pooling of the contrast agent in an enhancing lesion, such that the contrast agent from an earlier imaging study could confound the results of subsequent imaging studies in the same patient. As the order of the three different contrast-enhanced studies of each patient was randomized, however, a systematic effect (or bias) is hardly conceivable. In addition, we kept an interval of at least 24 hours (mean, 42 hours ± 30) between the three imaging studies. We believe that this constitutes an acceptable compromise between scientific demands and clinical necessities—it was by no means acceptable to postpone treatment to complete a scientific study.
In conclusion, in patients with brain tumors, standard contrast-enhanced MR imaging at 3.0 T compared with that at 1.5 T yielded a more than twofold higher lesion-to-brain CNR; even with half the standard contrast agent dose, lesion-to-brain CNR was higher at 3.0 T compared with that at 1.5 T with the full dose of contrast agent. The results of this preliminary study suggest that imaging at 3.0 T with the full contrast agent dose may possibly be used to improve the visualization of subtle disruptions of the blood-brain barrier—an approach that could prove beneficial in clinical settings in which maximum sensitivity is required. It may also be possible that at 3.0-T imaging, a reduced amount of contrast agent will be used in clinical settings that require a sensitivity level similar to the level provided by current 1.5-T imaging.
|
FOOTNOTES |
|---|
Abbreviations: CNR = contrast-to-noise ratio • ROI = region of interest • SE = spin echo • SNR = signal-to-noise ratio • TR = repetition time
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, H.H.S., C.K.K.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, C.K., W.A.W., H. Tschampa, M.B.; clinical and experimental studies, C.K., W.A.W., H. Tschampa, M.B., F.T., J.G., H. Textor; statistical analysis, C.K., W.A.W., M.B., F.T., C.K.K.; and manuscript editing, C.K., W.A.W., F.T., J.G., H.H.S., C.K.K.
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. K. Kuhl, F. Traber, and H. H. Schild Whole-Body High-Field-Strength (3.0-T) MR Imaging in Clinical Practice * Part I. Technical Considerations and Clinical Applications Radiology, March 1, 2008; 246(3): 675 - 696. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. J. Michaely, H. Kramer, O. Dietrich, K. Nael, K.-P. Lodemann, M. F. Reiser, and S. O. Schoenberg Intraindividual Comparison of High-Spatial-Resolution Abdominal MR Angiography at 1.5 T and 3.0 T: Initial Experience Radiology, September 1, 2007; 244(3): 907 - 913. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. K. Kuhl, P. Jost, N. Morakkabati, O. Zivanovic, H. H. Schild, and J. Gieseke Contrast-enhanced MR Imaging of the Breast at 3.0 and 1.5 T in the Same Patients: Initial Experience Radiology, June 1, 2006; 239(3): 666 - 676. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
