Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print March 20, 2006, 10.1148/radiol.2392050509
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2392050509v1
239/3/666    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kuhl, C. K.
Right arrow Articles by Gieseke, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kuhl, C. K.
Right arrow Articles by Gieseke, J.
(Radiology 2006;239:666-676.)
© RSNA, 2006

Breast Imaging

Contrast-enhanced MR Imaging of the Breast at 3.0 and 1.5 T in the Same Patients: Initial Experience1

Christiane K. Kuhl, MD, Petra Jost, MD, Nuschin Morakkabati, MD, Oliver Zivanovic, MD, Hans H. Schild, MD and Jürgen Gieseke, PhD

1 From the Departments of Radiology (C.K.K., P.J., N.M., H.H.S., J.G.) and Gynecology (O.Z.), University of Bonn, Sigmund-Freud-Str 25, 53105 Bonn, Germany, and Philips Medical Systems, Best, the Netherlands (J.G.). From the 2004 RSNA Annual Meeting. Received March 27, 2005; revision requested May 24; revision received May 26; accepted June 21; final version accepted July 26. Address correspondence to C.K.K. (e-mail: kuhl{at}uni-bonn.de).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Purpose: To establish a pulse sequence for dynamic contrast material–enhanced magnetic resonance (MR) imaging of the breast at 3.0 T and to prospectively compare MR imaging at 3.0 T with MR imaging at 1.5 T in the same patients.

Materials and Methods: A prospective intraindividual internal review board–approved study was performed in 37 women with 53 lesions (25 breast cancers, 28 benign focal lesions) who underwent contrast-enhanced dynamic bilateral subtraction MR imaging twice, once at 1.5 T with a standard technique (voxel size, 1.44 mm3) and once at 3.0 T (voxel size, 0.45–0.72 mm3) with variable repetition time and flip angle settings. Written informed consent was obtained. Sagittal single breast high-spatial-resolution MR imaging was performed with active fat suppression. Image quality, number and features of enhancing lesions, and Breast Imaging Reporting and Data System categories were compared by using the Wilcoxon matched-pairs signed rank test and Student t test for matched pairs. Diagnostic confidence was compared by using a receiver operating characteristic (ROC) analysis.

Results: With repetition time prolonged to account for longer T1 relaxation times at 3.0 T and a flip angle of 60°, enhancement rates at 3.0 T were substantially below those at 1.5 T. In two patients with benign lesions, enhancement was rated as insufficient to establish diagnosis. When parameter settings were kept equivalent, equivalent enhancement rates were observed with both systems. With these settings, 3.0-T MR imaging yielded homogeneous signal intensity over the entire field of view. No dielectric resonance effects were observed. Overall image quality scores for the dynamic series were slightly higher at 3.0 T (P < .01). A total of 49 lesions were prospectively identified with both systems. Owing to substantial patient motion at 1.5 T, two malignant lesions in one patient were visualized at 3.0 T only. At 3.0 T, differential diagnosis of enhancing lesions was possible with higher diagnostic confidence, as reflected by a larger area under the ROC curve (P < .05).

Conclusion: Initial experiences indicate that contrast-enhanced MR imaging at 3.0 T is nearing readiness for clinical use.

© RSNA, 2006


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
It has become increasingly clear that magnetic resonance (MR) imaging should be ideally suited for preoperative staging of the ipsilateral and contralateral breast in patients with Breast Imaging Reporting and Data System (BI-RADS) category 5 and 6 lesions or for screening in women who are at increased risk for breast cancer (1–15). All these applications, however, require that a technique be used that offers not only a high sensitivity but also a high specificity. One way to achieve this is by using pulse sequences that offer high spatial resolution in order to improve the morphologic analysis of enhancing lesions. High-spatial-resolution images must, however, be acquired within a short period of time, which is important in order to obtain optimum arterial phase contrast between the enhancing lesion and the adjacent enhancing fibroglandular tissue. Because of these time constraints and for reasons related to the signal-to-noise ratio (SNR), the maximum achievable spatial resolution at 1.5 T is limited. New image acquisition strategies, such as parallel imaging with sensitivity encoding (SENSE), may be used to reduce image acquisition times. At 1.5 T, however, high matrix imaging with a single signal average is usually associated with an already borderline SNR; if parallel imaging is used, the associated 30% SNR penalty may result in an inappropriately low SNR (16).

MR systems operating at higher magnetic field strengths (eg, 3.0 T) offer a higher SNR—possibly enough to acquire high imaging matrices with only one signal average or to allow fast acquisition strategies (17,18). Higher magnetic field strengths, however, are associated with physical effects that may not be advantageous for breast imaging; stronger susceptibility effects, longer T1 relaxation times, shorter in-phase echo times, and higher radiofrequency (RF) deposition may impair image quality and complicate or even impede image generation (19–23).

Thus, we conducted our study to establish a pulse sequence for dynamic contrast material–enhanced MR imaging of the breast at 3.0 T and to prospectively compare MR imaging at 3.0 T with MR imaging at 1.5 T in the same patients.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Our study was approved by our institutional ethics committee. One author (J.G.) is an employee of Philips Medical Systems and has a faculty position at the Department of Radiology of the University of Bonn. Authors who were not employees of Philips Medical Systems had control over the data and information submitted for publication.

Study Design and Patients
The investigation was conducted as a prospective intraindividual comparative study on consecutive patients who underwent contrast-enhanced MR imaging at 1.5 T for clinical reasons and who fulfilled the following inclusion criteria: 18 years of age or older, clinical indications for contrast-enhanced MR imaging of the breast at 1.5 T, no contraindication for MR imaging in general (no contraindications specific to high-field-strength imaging), proof of contrast-enhanced lesion on 1.5-T MR images, and willingness to volunteer for a second contrast-enhanced MR imaging examination performed on a separate day with the 3.0-T system.

A total 37 women (age range, 33–69 years; mean age, 50 years ± 9.3 [standard deviation]) met the inclusion criteria and volunteered to participate in the study. Patients were referred for contrast-enhanced MR imaging for preoperative staging of core biopsy–proved breast cancer or mammographic BI-RADS category 5 lesions (15 cases). Nine women were referred to clarify equivocal mammographic and/or ultrasonographic (US) findings, six were referred for familial breast cancer screening, five were referred for follow-up after conservation treatment, and two were referred for other reasons. All patients were informed in detail about the scientific purpose of the second imaging examination at high field strength, and written informed consent was obtained.

Care was taken to schedule the second (ie, the high-field-strength) examination after a minimum time period of at least 24 hours to allow for complete contrast agent elimination. After this waiting period had elapsed, the examination was performed as immediately as possible. In premenopausal women, specific care was taken to perform the second examination during the same week of the patient's menstrual cycle. The mean time interval between the examination at 1.5 T and the examination at 3.0 T was 2.4 days, with a median of 1 day. A total 21 patients were examined after 1 day, six after 2 days, three after 3 days, six after 4–6 days, and one after 9 days.

MR Imaging
Examinations were performed by using 1.5- and 3.0-T whole-body MR imaging systems (Intera 1.5 T and Intera 3.0 T; Philips Medical Systems, Best, the Netherlands). Both systems are equipped with high-performance gradients (master gradients), with a maximum slew rate of 150 (mT · m–1)/msec and a maximum gradient strength of 30 mT/m. Dedicated four-element SENSE-compatible breast surface coils (MRI Devices; InVivo Research, Orlando, Fla) were used with both systems. No breast fixation was used with either MR system.

For 1.5-T MR imaging, our standard bilateral dynamic contrast-enhanced protocol, which has been in routine clinical use for many years, was employed (11). This protocol consists of a fast localizer (scout view) gradient-echo sequence, with 25 sections (transverse, coronal, and sagittal) acquired separately for the left and right breasts; this sequence was used to prescribe sections of the subsequent dynamic series to cover the volume of fibroglandular tissue of both breasts exactly. The dynamic series consisted of a T1-weighted two-dimensional (2D) gradient-echo pulse sequence (290/4.6 [repetition time msec/echo time msec], flip angle of 90°), with a total of five dynamic acquisitions—one obtained before and four obtained immediately after a bolus injection of 0.1 mmol per kilogram body weight gadolinium dimeglumine (Magnevist; Schering, Berlin, Germany) at an injection rate of 3 mL/sec. Each dynamic volume consisted of 33 sections (3 mm thick), with an acquisition matrix of 512 x 394 (77% of a 512 matrix) and a field of view of 300–330 mm (adjusted to the size of the breasts). With this parameter setting, spatial resolution was 0.6 x 0.8 x 3 mm (noninterpolated), and temporal resolution was 1 minute 50 seconds per dynamic acquisition. To suppress the signal of fat, image subtraction was performed off-line after the actual imaging session.

For 3.0-T MR imaging, the pulse sequence for the dynamic series had to be adjusted to account for the physical effects of the higher magnetic field strength (Table). Our premise was to keep the spatial and temporal resolution equivalent to (or better than) those obtained with the standard 1.5-T setting. The following modifications were made for all pulse sequences at 3.0 T:


View this table:
[in this window]
[in a new window]

  		Core Pulse Sequence Data for Bilateral Contrast-enhanced Dynamic Series

 
First, for all pulse sequences performed without active fat suppression, the echo time had to be reduced from 4.6 to 2.3 msec to account for the shorter in-phase echo settings at 3.0 T. Second, because of the increased RF deposition in tissues at 3.0 T, the number and duration of RF pulses per unit time had to be reduced to adhere to specific absorption rate (SAR) limitations. Therefore, we used parallel imaging with a SENSE factor of two at 3.0 T in order to reduce the number of RF pulses that were needed for image generation. The imaging matrix was increased to a full 512 x 512 or 1024 x 640 mm, and the field of view was kept equivalent to that used for 1.5-T MR imaging in each patient.

To identify the appropriate pulse sequence parameters that would best demonstrate enhancing lesions (ie, gadolinium-induced T1-shortening effect) at 3.0 T, we used stepwise variations of repetition time and flip angle in the dynamic series. By this we sought to systematically investigate the influence of these parameters on the enhancement pattern of lesions at 3.0 T, with the enhancement pattern of each lesion at 1.5 T serving as the reference for each patient. The exact pulse sequence parameters regarding repetition time and flip angle are given in the Table.

To account for the fact that T1 relaxation times are prolonged by about 30% at 3.0 T and that RF deposition scales with the degree of flip angle, we began with a longer repetition time at 3.0 T compared with our standard repetition time at 1.5 T and reduced the flip angle to 60°. From there, repetition time was progressively reduced and combined with a flip angle of 60° or 73°, the latter being the highest flip angle that was technically achievable for the given in-phase echo time and SAR limitations. A total of seven different parameter groups with distinct parameter settings were defined as follows: Group 1 (two patients) was investigated with a repetition time of 430 msec and a flip angle of 60°; group 2 (three patients), with a repetition time of 400 msec and a flip angle of 60°; group 3 (three patients), with a repetition time of 300–380 msec and a flip angle of 60°; group 4 (three patients), with a repetition time of 300–380 msec and a flip angle of 73°; group 5 (six patients), with a repetition time of 260–290 msec and a flip angle of 60°; group 6 (13 patients), with a repetition time of 260–290 msec and a flip angle of 73°; and group 7 (seven patients), with a repetition time of less than 260 msec and a flip angle of 73°.

Because the parameter setting that was used in the first two patients (group 1) turned out to be inappropriate (see Results), these examinations were not used for further analysis of image quality or for image interpretation. Accordingly, the analysis of image quality and diagnostic confidence was based on data from 35 patients in groups 2–7 only.

In addition to and after the dynamic series, we acquired sagittal T1-weighted radial-segmented three-dimensional gradient-echo MR images at 3.0 T, with active fat suppression and high spatial resolution (9/2.8, 25° flip angle, 80 sections acquired, 1–2-mm section thickness, 250-mm field of view, active fat suppression, 0.5 x 0.5 x 1.0-mm pixel size [noninterpolated], and total acquisition time of 90 seconds). These images were not used for the direct comparison of image quality because no such images were obtained at 1.5 T. They were, however, made available for clinical image interpretation.

Data Analysis
Image interpretation and final diagnosis.—Images obtained at 3.0 T and those obtained at 1.5 T for the 35 patients in groups 2–7 were transferred to hardcopy with standardized window settings. Immediately after the examinations, image interpretation was performed prospectively by different radiologists who had substantial expertise in interpreting MR images of the breast (C.K.K. and N.M., with 11 and 6 years of experience, respectively) and who had each read half of the 1.5-T MR images and half of the 3.0-T MR images. The readers noted the number of lesions and assigned a BI-RADS category for each lesion. The same interpretation criteria were used for the 1.5- and the 3.0-T examinations (see below).

Readers were blinded to the findings and final BI-RADS categories given for the same patient at the other field strength. To avoid interreader variability confounding the results, a conjoint reading of images from both examinations was performed for cases in which the lesions were assigned divergent BI-RADS categories. Conventional images, as well as mammograms, US images, and reports, were made available during the reading session just as they would be in the clinical setting. Final management was decided on the basis of results from the 1.5-T examination. For cases in which higher BI-RADS categories were assigned at 3.0 T, the results were carefully discussed with the patient and her surgeon. Final management was decided only after a repeat explanation of the experimental character of the 3.0-T examination was provided.

Masses that had an irregular lesion shape, borders or spicules, and heterogeneous or rim enhancement or non-masslike enhancement, with asymmetric and segmental or ductal configuration, were considered suspicious for malignancy. Rapid enhancement was used to corroborate suspicious findings, except for lesions that appeared morphologically benign and had persistent enhancement in the entire dynamic series. If rim enhancement was identified within a focal mass, the lesion was considered suspicious, irrespective of other findings. If dark septations were present within an oval mass with smooth borders, the lesion was rated as definitely benign (fibroadenoma), irrespective of other findings. Slow, persistent enhancement was considered to corroborate the diagnosis of a benign lesion, except for non-masslike enhancement with a ductal or segmental configuration or for slow, persistent enhancement in masses with morphologic findings suggestive of malignancy (irregular borders or spicules and/or rim enhancement). If washout was observed, the lesion was considered suspicious, except in cases of well-circumscribed lesions with dark septations or small (<5 mm) non-masslike foci of enhancement.

The final diagnosis was established by means of excisional or core biopsy in 20 patients with 31 lesions and by means of follow-up of at least 12 months, including mammography and MR imaging, in 15 patients with 20 lesions.

Image quality.—For the prospective analysis of image quality, all images were transferred to a workstation and were displayed with standardized window settings. Images were displayed in a pairwise fashion to allow direct comparison of 1.5- and 3.0-T MR images. No attempt was made to blind the reader to the field strength with which the respective images were acquired because it was obvious from the appearance of the images obtained with SENSE at 3.0 T. A breast radiologist with 11 years of experience in reading MR images (C.K.K.) was asked to rate the image quality by visual assessment. Only the images of the dynamic series were used for comparison; the high-spatial-resolution images obtained before and after the dynamic series at 3.0 T (see above) were considered separately because they were obtained with an entirely different acquisition technique that would not allow fair (or meaningful) comparison between 1.5- and 3.0-T MR images.

Image quality was rated on a scale of 1 (nondiagnostic image quality) to 5 (excellent) on the basis of the following criteria:

A rating of 1 (nondiagnostic image quality) was assigned for cases in which the image quality was considered insufficient for diagnosis because of insufficient signal intensity homogeneity or massive dielectric resonance effects across the field of view, with a complete loss of signal intensity in parts of the field of view owing to severe artifacts (complete image degradation by susceptibility or pulsation) and/or poor visual SNR. A dielectric resonance effect was considered to be present if concentric signal intensity shadowing was noted in the center of the image. A susceptibility artifact was considered to be present if image blurring and image distortions were noted at tissue interfaces.

A rating of 2 (poor image quality) was assigned for cases in which there was insufficient signal intensity homogeneity or dielectric resonance effects, with substantial yet incomplete signal intensity variations across the field of view and/or substantial but incomplete image degradation due to susceptibility or pulsation and/or a low visual SNR.

A rating of 3 (acceptable) was assigned for cases in which there were only slight inhomogeneities in signal intensity or dielectric resonance effects across the entire field of view and/or moderate susceptibility or pulsation artifacts and a high visual SNR.

A rating of 4 (good) was assigned for cases in which there were mild heterogeneity changes in signal intensity across the field of view (eg, slight left-to-right differences in signal intensity), no dielectric resonance effects, mild susceptibility or pulsation artifacts, and high visual SNR.

A rating of 5 (excellent) was assigned for cases in which there were no or hardly perceivable signal intensity variations across the field of view, no dielectric resonance effects, no or only mild susceptibility or pulsation artifacts, and high visual SNR.

Quantitative assessment of enhancement rates.—A breast radiologist (C.K.K.) placed a region of interest in all lesions that were identified during hardcopy reading. The size of the region of interest varied with the size of the enhancing lesion and was chosen to selectively include the area with the strongest enhancement, as identified on the first postcontrast subtracted image. The average size of the region of interest was 3 mm2. The enhancement rates (percentage signal intensity increase) of the lesions were calculated by using the following equation: [(SIpost – SIpre)/SIpre] · 100, where SIpre is the signal intensity before contrast injection (first dynamic acquisition) and SIpost is the signal intensity after contrast injection (first postcontrast dynamic acquisition). The enhancement rates of each lesion at 3.0 T were compared with those of each lesion at 1.5 T.

Relative enhancement rates (RERs) (ie, the enhancement rate of a lesion at 3.0 T vs the enhancement rate of the same lesion at 1.5 T) were calculated as ER3.0T/ER1.5T, where ER3.0T is the enhancement rate measured at 3.0 T and ER1.5T is the enhancement rate measured at 1.5 T. A RER of 1 would correspond to equivalent enhancement rates at 3.0 and 1.5 T, whereas a RER of 0.5 would indicate that the enhancement rate at 3.0 T was 50% of the enhancement rate at 1.5 T.

RERs were compared between the seven parameter groups to identify the pulse sequence parameters (specifically regarding repetition time and flip angle) that would yield the strongest enhancement at 3.0 T. For this purpose, all the quantitative enhancement rates of lesions that belonged to the same parameter group (ie, those lesions that were imaged with the same parameter settings) were averaged. The resulting mean values were compared to determine the parameter setting that would yield the highest RER (3.0 vs 1.5 T).

Statistical Analysis
The image quality scores of the dynamic contrast-enhanced series were compared for both field strengths; the Wilcoxon matched-pairs signed rank test was used to test for statistical significance. Enhancement rates of lesions at 3.0 T were compared with those of the same lesions at 1.5 T; the Student t test for matched pairs was used to test for statistical significance. To account for several observations in the same patient (ie, in patients with more than one enhancing lesion), a mean was calculated for the different enhancement rates of the different lesions in each patient, and this mean value alone was used for further analysis. The same method was applied for BI-RADS categories in patients with several lesions in one breast. In these patients, lesion classes were defined for the purpose of the receiver operating characteristic (ROC) analysis. A lesion class was used for further analysis if several lesions or if a cluster of lesions with identical appearance were present on the MR image. A P value of less than .05 was considered to indicate a statistically significant difference. Only malignant lesions were evaluated in patients with breast cancers.

To investigate the differential diagnostic power offered by MR imaging at 1.5 and 3.0 T, an ROC analysis was performed (ROCKIT, version 0.9b; C. E. Metz, University of Chicago, Ill), and the areas under the ROC curves were compared for both examinations.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
A total 53 lesions were identified on 1.5- and 3.0-T MR images in 37 patients, and the following diagnoses were established: A total of 25 breast cancers were diagnosed in 16 patients. Five patients had bilateral breast cancer, and another five had multifocal or multicentric breast cancer. Three patients had pure ductal carcinoma in situ, and another two patients had ductal carcinoma in situ in addition to an invasive lesion. The size of the invasive cancers ranged from 4 to 55 mm. In 21 patients, a total 28 benign lesions were present, including 17 fibroadenomas (11 patients), six areas of adenosis and/or fibrocystic disease (four patients), three intramammary lymph nodes (two patients), one radial scar (one patient), and extensive chronic mastitis (one patient).

The 3.0-T MR images obtained in the two patients from group 1 who had two benign lesions (fibroadenomas) were excluded from analysis because of insufficient enhancement. Accordingly, the analysis of image quality and diagnostic confidence was based on data from 35 patients with a total of 51 enhancing lesions. For the identification of an appropriate imaging protocol, however, data from all 53 lesions in 27 patients were included for analysis.

Lesion Detection and Classification
A total 49 lesions were prospectively identified at 1.5-T MR imaging. These lesions were also prospectively and independently identified at 3.0-T MR imaging. Two additional lesions were diagnosed at 3.0-T MR imaging in a 54-year-old woman who presented for preoperative staging of a biopsy-proved breast cancer in her right breast. During the 1.5-T MR imaging examination, the images of the left breast had been substantially degraded owing to considerable patient motion. At 3.0 T, no motion occurred, and a suspicious mass with adjacent non-masslike linear enhancement was identified in the contralateral left breast. The lesions were scored as BI-RADS category 5 (highly suggestive of breast cancer and ductal carcinoma in situ) at 3.0 T. Preoperative MR-guided localization was performed at 1.5 T; this time, images were obtained without patient motion, and the lesion was clearly visualized. Histologic analysis revealed a ductulolobular invasive cancer (pT1b, TNM classification) with adjacent intraductal cancer.

BI-RADS categories, which were given per lesion and not per breast, were identical in 40 (78%) of 51 lesions. BI-RADS categories differed in 10 patients with a total of 11 different lesions or classes of lesions. In addition to the discrepancies noted in the one patient with two contralateral breast cancers mentioned above, the following discrepancies regarding BI-RADS categories were observed: In one patient (Fig 1) who presented for preoperative staging of a biopsy-proved duct invasive breast cancer, multicentric disease had already been identified at 1.5-T MR imaging. During the 3.0-T MR imaging examination, two additional invasive foci were prospectively identified in a third quadrant and were categorized as BI-RADS category 4 lesions, whereas these foci had been categorized as BI-RADS category 2 lesions at 1.5-T MR imaging. The demonstration of these two additional lesions (which were grouped as a single class of lesions for analysis) did not change management because multicentric disease had already been diagnosed in this patient at 1.5-T MR imaging.


Figure 1
View larger version (82K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a: Multicentric ductal invasive cancer in 48-year-old woman. (a) Maximum intensity projection of subtracted images of first postcontrast subtracted acquisition at 1.5-T dynamic MR imaging (2D gradient echo, 290/4.6, 90° flip angle, transverse orientation, 512 x 394 matrix). (b) Maximum intensity projection of subtracted images of first postcontrast subtracted acquisition at 3.0-T dynamic MR imaging (2D gradient echo, 290/2.3, 73° flip angle, transverse orientation, 512 x 512 matrix). Large ductal invasive cancer in the upper inner quadrant (arrow) and smaller lesion in the upper outer quadrant (arrowhead) are seen in a. These lesions plus additional enhancing lesions (arrows) are seen in b.

 

Figure 1
View larger version (88K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b: Multicentric ductal invasive cancer in 48-year-old woman. (a) Maximum intensity projection of subtracted images of first postcontrast subtracted acquisition at 1.5-T dynamic MR imaging (2D gradient echo, 290/4.6, 90° flip angle, transverse orientation, 512 x 394 matrix). (b) Maximum intensity projection of subtracted images of first postcontrast subtracted acquisition at 3.0-T dynamic MR imaging (2D gradient echo, 290/2.3, 73° flip angle, transverse orientation, 512 x 512 matrix). Large ductal invasive cancer in the upper inner quadrant (arrow) and smaller lesion in the upper outer quadrant (arrowhead) are seen in a. These lesions plus additional enhancing lesions (arrows) are seen in b.

 
In two patients, the 1.5- or 3.0-T examination yielded false-positive (BI-RADS 4) results, whereas the corresponding 3.0- or 1.5-T examination yielded true-negative (BI-RADS 3) results. In four patients with fibroadenomas (Fig 2), lesions were rated as BI-RADS category 3 at 1.5-T MR imaging and as BI-RADS category 2 at 3.0-T MR imaging owing to the visibility of septations that were not resolved at 1.5 T. In another two patients who had fibrocystic disease with focal adenosis at final diagnosis, lesions were considered suspicious (BI-RADS 4) at 1.5 T, whereas the same lesions appeared less conspicuous at 3.0 T.


Figure 2
View larger version (131K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a: Fibroadenoma in 49-year-old woman. (a) Dynamic subtraction MR image of first postcontrast acquisition (2D gradient echo, 300/4.6, 90° flip angle, transverse orientation) at 1.5 T. (b) Dynamic subtraction MR image of first postcontrast acquisition (2D gradient echo, 290/2.3, 73° flip angle, transverse orientation) at 3.0 T. Note the enhancing mass in upper outer quadrant (arrow) in a. Lesion shows slightly heterogeneous enhancement, and no other features of internal architecture are visible. Lesion was classified as BI-RADS category 3 owing to its smooth borders, oval shape, and rapid enhancement. Lesion (arrow) is also visible in b; however, dark septations are resolved and visible. Lesion was classified as BI-RADS category 2 at 3.0 T.

 

Figure 2
View larger version (120K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b: Fibroadenoma in 49-year-old woman. (a) Dynamic subtraction MR image of first postcontrast acquisition (2D gradient echo, 300/4.6, 90° flip angle, transverse orientation) at 1.5 T. (b) Dynamic subtraction MR image of first postcontrast acquisition (2D gradient echo, 290/2.3, 73° flip angle, transverse orientation) at 3.0 T. Note the enhancing mass in upper outer quadrant (arrow) in a. Lesion shows slightly heterogeneous enhancement, and no other features of internal architecture are visible. Lesion was classified as BI-RADS category 3 owing to its smooth borders, oval shape, and rapid enhancement. Lesion (arrow) is also visible in b; however, dark septations are resolved and visible. Lesion was classified as BI-RADS category 2 at 3.0 T.

 
The ROC analysis revealed an area under the ROC curve of 0.98 for images obtained at 3.0-T MR imaging and 0.84 for those obtained at 1.5-T MR imaging (P < .05, two-tailed analysis).

Image Quality
Image quality scores for dynamic 2D gradient-echo MR images obtained with both field strengths differed slightly, with a mean score of 4.6 ± 0.5 for MR images obtained at 3.0 T and 4.2 ± 0.5 for those obtained at 1.5 T. This difference proved to be statistically significant (P < .01) by using the Wilcoxon matched-pairs signed rank test. The median image quality score was 5 (excellent) for images obtained at 3.0 T and 4 (good) for those obtained at 1.5 T. In detail, image quality scores for the dynamic series yielded equivalent values in 13 (37%) of 35 patients; a lower score (by one point) was given at 3.0 T in three (9%) of 35 patients, and a higher score (by one point) was given at 3.0 T in 19 (54%) of 35 patients. None of the image quality ratings differed by more than one point. Among the 10 patients in whom the acquisition matrix was increased to 1024 x 680, seven had images of excellent quality that were superior to those obtained at 1.5 T (Fig 3). In three patients, however, blurring and double contours of enhancing lesions on the subtracted images degraded image quality. This was due to the fact that, with very small pixel sizes, the already subtle motion of the patients encompassed a full pixel thickness and therefore produced subtraction artifacts. On none of the 3.0-T MR images were artifacts resulting from the use of parallel imaging with SENSE observed.


Figure 3
View larger version (94K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a: Lobular invasive pT1a breast cancer (4 mm) in 52-year-old woman. (a) Dynamic MR image of precontrast acquisition (2D gradient echo, 300/2.3, 72° flip angle, transverse orientation, 1024 x 680 matrix) at 3.0 T. (b) First postcontrast acquisition obtained with same technique as in a. (c) Subtracted image of same acquisition as in b. (d) Close-up view of c. Although the lesion (arrow) is smaller than 5 mm (corresponding to a focus of enhancement according to the BI-RADS lexicon), it is possible to delineate morphologic details, such as spicules.

 

Figure 3
View larger version (96K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b: Lobular invasive pT1a breast cancer (4 mm) in 52-year-old woman. (a) Dynamic MR image of precontrast acquisition (2D gradient echo, 300/2.3, 72° flip angle, transverse orientation, 1024 x 680 matrix) at 3.0 T. (b) First postcontrast acquisition obtained with same technique as in a. (c) Subtracted image of same acquisition as in b. (d) Close-up view of c. Although the lesion (arrow) is smaller than 5 mm (corresponding to a focus of enhancement according to the BI-RADS lexicon), it is possible to delineate morphologic details, such as spicules.

 

Figure 3
View larger version (89K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3c: Lobular invasive pT1a breast cancer (4 mm) in 52-year-old woman. (a) Dynamic MR image of precontrast acquisition (2D gradient echo, 300/2.3, 72° flip angle, transverse orientation, 1024 x 680 matrix) at 3.0 T. (b) First postcontrast acquisition obtained with same technique as in a. (c) Subtracted image of same acquisition as in b. (d) Close-up view of c. Although the lesion (arrow) is smaller than 5 mm (corresponding to a focus of enhancement according to the BI-RADS lexicon), it is possible to delineate morphologic details, such as spicules.

 

Figure 3
View larger version (125K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3d: Lobular invasive pT1a breast cancer (4 mm) in 52-year-old woman. (a) Dynamic MR image of precontrast acquisition (2D gradient echo, 300/2.3, 72° flip angle, transverse orientation, 1024 x 680 matrix) at 3.0 T. (b) First postcontrast acquisition obtained with same technique as in a. (c) Subtracted image of same acquisition as in b. (d) Close-up view of c. Although the lesion (arrow) is smaller than 5 mm (corresponding to a focus of enhancement according to the BI-RADS lexicon), it is possible to delineate morphologic details, such as spicules.

 
The high-spatial-resolution sagittal fat-suppressed MR images obtained after the dynamic series at 3.0 T (Fig 4) revealed consistently excellent image quality, with a score of 5 in all but three cases; these cases were scored as 3 owing to substantial patient motion. Fat suppression was homogeneous in all cases.


Figure 4
View larger version (104K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4a: Bilateral fibroadenomas in 58-year-old woman. (a, b) Two sections of a dynamic subtraction MR imaging series (first postcontrast acquisition, 2D gradient echo, 300/4.6, 90° flip angle, transverse orientation, 512 x 394 matrix, acquisition time of 110 seconds) at 1.5 T. (c, d) Two sections of postcontrast fat-suppressed T1-weighted spiral-segmented three-dimensional MR image of (c) left and (d) right breasts at 3.0 T (9/2.8, 25° flip angle, 80 sections acquired, 1–2-mm section thickness, 250-mm field of view, 0.5 x 0.5 x 1.0-mm pixel size [noninterpolated], acquisition time of 90 seconds) demonstrate enhancing mass (arrow). In a and b, some motion occurred that degraded the quality of the subtraction images; note the two enhancing masses in the right and left breasts (arrow). Image quality is superior in c and d. Note the strong enhancement of the parenchyma, which is due to the long time interval after contrast material injection, at 3.0 T. Note the excellent visualization of dark septations in c and (to a lesser extent) in d.

 

Figure 4
View larger version (105K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4b: Bilateral fibroadenomas in 58-year-old woman. (a, b) Two sections of a dynamic subtraction MR imaging series (first postcontrast acquisition, 2D gradient echo, 300/4.6, 90° flip angle, transverse orientation, 512 x 394 matrix, acquisition time of 110 seconds) at 1.5 T. (c, d) Two sections of postcontrast fat-suppressed T1-weighted spiral-segmented three-dimensional MR image of (c) left and (d) right breasts at 3.0 T (9/2.8, 25° flip angle, 80 sections acquired, 1–2-mm section thickness, 250-mm field of view, 0.5 x 0.5 x 1.0-mm pixel size [noninterpolated], acquisition time of 90 seconds) demonstrate enhancing mass (arrow). In a and b, some motion occurred that degraded the quality of the subtraction images; note the two enhancing masses in the right and left breasts (arrow). Image quality is superior in c and d. Note the strong enhancement of the parenchyma, which is due to the long time interval after contrast material injection, at 3.0 T. Note the excellent visualization of dark septations in c and (to a lesser extent) in d.

 

Figure 4
View larger version (131K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4c: Bilateral fibroadenomas in 58-year-old woman. (a, b) Two sections of a dynamic subtraction MR imaging series (first postcontrast acquisition, 2D gradient echo, 300/4.6, 90° flip angle, transverse orientation, 512 x 394 matrix, acquisition time of 110 seconds) at 1.5 T. (c, d) Two sections of postcontrast fat-suppressed T1-weighted spiral-segmented three-dimensional MR image of (c) left and (d) right breasts at 3.0 T (9/2.8, 25° flip angle, 80 sections acquired, 1–2-mm section thickness, 250-mm field of view, 0.5 x 0.5 x 1.0-mm pixel size [noninterpolated], acquisition time of 90 seconds) demonstrate enhancing mass (arrow). In a and b, some motion occurred that degraded the quality of the subtraction images; note the two enhancing masses in the right and left breasts (arrow). Image quality is superior in c and d. Note the strong enhancement of the parenchyma, which is due to the long time interval after contrast material injection, at 3.0 T. Note the excellent visualization of dark septations in c and (to a lesser extent) in d.

 

Figure 4
View larger version (161K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4d: Bilateral fibroadenomas in 58-year-old woman. (a, b) Two sections of a dynamic subtraction MR imaging series (first postcontrast acquisition, 2D gradient echo, 300/4.6, 90° flip angle, transverse orientation, 512 x 394 matrix, acquisition time of 110 seconds) at 1.5 T. (c, d) Two sections of postcontrast fat-suppressed T1-weighted spiral-segmented three-dimensional MR image of (c) left and (d) right breasts at 3.0 T (9/2.8, 25° flip angle, 80 sections acquired, 1–2-mm section thickness, 250-mm field of view, 0.5 x 0.5 x 1.0-mm pixel size [noninterpolated], acquisition time of 90 seconds) demonstrate enhancing mass (arrow). In a and b, some motion occurred that degraded the quality of the subtraction images; note the two enhancing masses in the right and left breasts (arrow). Image quality is superior in c and d. Note the strong enhancement of the parenchyma, which is due to the long time interval after contrast material injection, at 3.0 T. Note the excellent visualization of dark septations in c and (to a lesser extent) in d.

 
Enhancement Rates
Mean enhancement rates for all lesions (and thus for all parameter groups with different repetition times and flip angles) were 88.8 ± 41.2 at 3.0 T and 103.7 ± 45.9 at 1.5 T. The broad standard deviation of the enhancement values is not solely the result of variable acquisition techniques; these broad deviations are mainly due to the fact that the mean value included both benign and malignant lesions. The difference between lesion enhancement at 3.0 T and lesion enhancement at 1.5 T proved statistically significant by using the Student t test for one sample (P < .05) and by using the Wilcoxon matched-pairs signed rank test. This difference was lost (ie, lesions showed equivalent enhancement rates) if only those lesions from parameter groups 3–7 were included in the analysis.

The mean RER for all lesions was 0.89 ± 0.30, indicating that (on average) lesion enhancement at 3.0 T was less than 90% of lesion enhancement at 1.5 T.

Analysis of the influence of variable repetition times and flip angles (Fig 5) showed a correlation between RER, repetition time, and flip angle, with RER decreasing as flip angle decreased and repetition time increased. Enhancement rates at 3.0 T and those at 1.5 T were almost equivalent (ie, RER of about 1) in only those subgroups that were examined with a repetition time at 3.0 T that was at or below the repetition time at 1.5 T, with a flip angle of 73°. In the parameter groups examined with a repetition time of longer than 300 msec (particularly if combined with a flip angle of 60°), enhancement rates obtained at 3.0 T were substantially lower than those obtained at 1.5 T. In the patients imaged with the longest repetition time (460 msec) and a flip angle of 60°, enhancement at 3.0 T was only one-third of the enhancement at 1.5 T (RER of 0.3).


Figure 5
View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5: RERs for different parameter groups. Columns provide average RERs of lesions in each parameter group. RER of 1 indicates equivalent enhancement at 1.5 and 3.0 T. Note that with a longer repetition time (TR) and lower flip angle (FA), RER is well below 1, indicating that enhancement at 3.0 T remained substantially below the corresponding value at 1.5 T. Note that RER increases with shorter repetition times and higher flip angles. However, even if a shorter repetition time was used, the enhancement rates at 3.0 T were merely equivalent to those recorded at 1.5 T.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
 References
 
Some basic technical requirements that are of pivotal importance for MR imaging of the breast include (a) T1-weighted gradient-echo MR images acquired with a high enough spatial resolution to depict the morphologic details of small, enhancing lesions, (b) the use of an acquisition time short enough to ensure that enhancing lesions are not masked by progressively enhancing adjacent fibroglandular tissue, and (c) the use of an acquisition time short enough to allow kinetic analysis of enhancing lesions (24–32). These diverging requirements (for high spatial and high temporal resolution) can be best met by using high-performance MR systems.

Higher magnetic field strengths (eg, 3.0 T) have been shown to be useful for structural and functional neuroimaging, as well as for MR angiography (33–35). The higher SNR afforded by the higher field strength can be successfully translated into improved image quality and higher diagnostic yield. It is clear that high-field-strength imaging for body applications will have to overcome some intrinsic difficulties that are specific to high-field-strength imaging (36–39). Homogeneous signal intensity for a large field of view is more difficult to achieve with increasing field strength. For breast imaging, this may interfere with the clinical demand to image both breasts. All types of motion (breathing, pulsation, and considerable or subtle motion) can result in substantially stronger artifacts at higher field strengths, which may prove problematic in any thoracic or abdominal application, including breast imaging.

Susceptibility effects will also be stronger at 3.0 T and may cause image distortions, particularly on images acquired with a gradient-echo pulse sequence, which is the type of sequence that is typically used for breast MR imaging. Fat-water phase cycling is faster with 3.0-T imaging than with 1.5-T imaging such that opposed-phase echo times are reached with only a 1.1-msec shift in in-phase echo time. This means that exact echo times must be maintained for dynamic images that are acquired without fat suppression. T1 relaxation times are longer at higher field strengths, which means that, theoretically, longer repetition times will have to be used—an effect that could compromise temporal resolution if a 2D gradient-echo sequence is used.

Difficulties that are new for 3.0-T MR imaging compared with 1.5-T MR imaging are RF related. Heterogeneous RF distribution (so-called dielectric resonance effects) can lead to extensive areas of signal void on 3.0-T MR images. Second, RF energy deposition in tissue scales exponentially with increasing field strength, which means that SAR limitations are reached sooner at 3.0 T. If no techniques are available to compensate for this effect (such as SENSE), high-field-strength systems will have to be slowed down (ie, the number of RF pulses per time unit will have to be reduced) in order to avoid excessive heating in the patient (17,18,20).

Despite these potential obstacles, our intraindividual comparative study showed that image quality was slightly but significantly higher at 3.0 T than at 1.5 T. Image quality was rated as excellent throughout the high-spatial-resolution images with active fat suppression that were, for SNR reasons, obtained at 3.0 T only.

For bilateral transverse dynamic imaging, the field of view was large enough to accommodate both breasts in all 37 patients. No relevant signal intensity variations were observed across the field of view. Dielectric resonance effects were not observed at all; this finding was expected because the chest and the abdominal wall contribute most to the shielding of RF distribution. Susceptibility effects, particularly blurring, were not observed on the 2D gradient-echo MR images, possibly because the echo time used at 3.0 T was shorter than the echo time used at 1.5 T. Motion artifacts were observed with equivalent prevalence and to an equivalent degree on 3.0- and 1.5-T MR images. Motion was, however, a problem in patients who were examined with high in-plane imaging matrices (1024 x 640) and in whom subtraction artifacts could degrade image quality. This effect was not directly attributable to the higher field strength but was rather due to the smaller pixel size, which at image subtraction could cause subtraction artifacts with even subtle patient motion. We conclude that if very high matrix imaging is performed for a large (bilateral) field of view, some sort of breast fixation will be needed to avoid micromotion, which causes pixel shifts (and thus, artifacts) on subtraction images.

By using SENSE and a reduction factor of two, we were able to reduce the RF deposition by half (compared with non-SENSE imaging). Still, the one 3.0-T-related physical feature that did in fact have a substantial effect on image acquisition (and therefore on image appearance) was RF absorption and/or SAR. Because RF deposition scales exponentially with field strength and with the size of the RF pulse flip angle, we had to reduce the flip angle from 90° (at 1.5 T) to 60° or 73° (at 3.0 T) to adhere to SAR limitations and to fit the flip angle into the shorter in-phase echo time at 3.0 T. Because the size of the flip angle, together with repetition time, is an important determinant of T1 contrast, this may have a substantial effect on the overall enhancement rates observed.

A 30% prolongation of T1 relaxation times has been reported for healthy tissues at 3.0 T (19–21). Accordingly, we began with a pulse sequence that had a longer repetition time compared with the standard 1.5-T setting. With this parameter setting, however, the enhancement rates of the same lesions at 3.0 T remained substantially below those at 1.5 T, with a RER on the order of 0.3–0.4. This was probably at least in part due to the lower flip angle. Even if the same repetition time was used at both 3.0 and 1.5 T and the flip angle was increased to 73°, enhancement rates at 3.0 T would be merely equivalent to those recorded at 1.5 T. Our results indicate that, for 2D gradient-echo dynamic MR imaging with a flip angle limited to 60°–73°, it seems prudent not to adapt the repetition time to account for T1 prolongation at 3.0 T but to use the shortest possible repetition time. Moreover, and unlike findings for contrast-enhanced MR imaging of the brain at 3.0 T, the contrast agent dose may not be reduced for breast MR imaging (35,40).

The higher spatial resolution that is attainable at 3.0 T helped improve the classification of 10 of the 51 total lesions in nine of 35 patients. The increase in diagnostic confidence was reflected by a significantly larger area under the curve at ROC analysis. We wish to underscore, however, that this difference regarding diagnostic confidence should not be interpreted as proof of higher sensitivity and/or specificity of MR imaging at 3.0 T. A limitation of our study was a substantial selection bias in our cohort because patients were asked to participate in the study if they had an enhancing lesion observed at 1.5-T MR imaging; accordingly, the cohort is probably not representative of the average group of patients seen in breast MR imaging clinics. So there is reason to assume (but no proof yet) that MR imaging of the breast at 3.0 T may lead to an improved diagnostic accuracy compared with 1.5-T MR imaging.

In our study, we used the extra SNR afforded by 3.0-T MR imaging to improve spatial resolution only, with the intention of improving the analysis of subtle morphologic details. Thus, another study limitation is that we did not investigate the use of the extra SNR for breast imaging with high temporal resolution (ie, with the intention to improve the kinetic modeling of enhancing lesions) (36). Also, we did not investigate the use of the extra SNR for functional breast imaging techniques, such as diffusion-weighted MR imaging or MR spectroscopy. There is evidence to suggest that these techniques may be used to further improve the differential diagnosis of enhancing lesions (41–45); at 1.5 T, these techniques have limited SNR and poor spatial resolution. It may be anticipated that the extra signal intensity gained by using 3.0-T systems should also foster the use of these approaches, with another potential increase in specificity.

In conclusion, our data suggest that 3.0-T MR imaging of the breast is nearing readiness for clinical use. Further studies in unselected groups of patients are under way in our department to investigate the effect of high-field-strength imaging on diagnostic accuracy of MR imaging in clinical practice.


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
References
 


   FOOTNOTES
 

Abbreviations: BI-RADS = Breast Imaging Reporting and Data System • RER = relative enhancement rates • RF = radiofrequency • ROC = receiver operating characteristic • SAR = specific absorption rate • SENSE = sensitivity encoding • SNR = signal-to-noise ratio • 2D = two dimensional

See Materials and Methods for pertinent disclosures.

Author contributions: Guarantor of integrity of entire study, 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; manuscript final version approval, all authors; literature research, C.K.K.; clinical studies, C.K.K., P.J., N.M.; statistical analysis, C.K.K.; and manuscript editing, C.K.K.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 

  1. Bedrosian I, Mick R, Orel SG, et al. Changes in the surgical management of patients with breast carcinoma based on preoperative magnetic resonance imaging. Cancer 2003;98:468–473.[CrossRef][Medline]
  2. Harms SE. Breast Cancer Staging Working Group report. J Magn Reson Imaging 1999;10:991–994.[CrossRef][Medline]
  3. Kneeshaw PJ, Turnbull LW, Drew PJ. Role of magnetic resonance imaging in the diagnosis and single-stage surgical resection of invasive lobular carcinoma of the breast. Br J Surg 2002;89: 1296–1301.[CrossRef][Medline]
  4. Liberman L, Morris EA, Kim CM, et al. MR imaging findings in the contralateral breast of women with recently diagnosed breast cancer. AJR Am J Roentgenol 2003;180:333–341.[Abstract/Free Full Text]
  5. Liberman L, Morris EA, Dershaw DD, Abramson AF, Tan LK. MR imaging of the ipsilateral breast in women with percutaneously proven breast cancer. AJR Am J Roentgenol 2003;180:901–910.[Abstract/Free Full Text]
  6. Lee SG, Orel SG, Woo IJ, et al. MR imaging screening of the contralateral breast in patients with newly diagnosed breast cancer: preliminary results. Radiology 2003;226:773–778.[Abstract/Free Full Text]
  7. Morakkabati N, Leutner CC, Schmiedel A, Schild HH, Kuhl CK. Breast MR imaging during or soon after radiation therapy. Radiology 2003;229:893–901.[Abstract/Free Full Text]
  8. Morris EA, Liberman L, Ballon DJ, et al. MRI of occult breast carcinoma in a high-risk population. AJR Am J Roentgenol 2003;181:619–626.[Abstract/Free Full Text]
  9. Munot K, Dall B, Achuthan R, Parkin G, Lane S, Horgan K. Role of magnetic resonance imaging in the diagnosis and single-stage surgical resection of invasive lobular carcinoma of the breast. Br J Surg 2002;89:1296–1301.[CrossRef][Medline]
  10. Orel SG, Schnall MD, Powell M, et al. Staging of suspected breast cancer: effect of MR imaging and MR-guided biopsy. Radiology 1995;196:115–122.[Abstract/Free Full Text]
  11. Kuhl CK, Schmutzler RK, Leutner CC, et al. Breast MR imaging screening in 192 women proved or suspected to be carriers of a breast cancer susceptibility gene: preliminary results. Radiology 2000;215:267–279.[Abstract/Free Full Text]
  12. Stoutjesdijk MJ, Boetes C, Jager GJ, et al. Magnetic resonance imaging and mammography in women with a hereditary risk of breast cancer. J Natl Cancer Inst 2001;93:1095–1102.[Abstract/Free Full Text]
  13. Tilanus-Linthorst MM, Obdeijn IM, Bartels KC, de Koning HJ, Oudkerk M. First experiences in screening women at high risk for breast cancer with MR imaging. Breast Cancer Res Treat 2000;63:53–60.[CrossRef][Medline]
  14. Warner E, Plewes DB, Hill KA, et al. Surveillance of BRCA1 and BRCA2 mutation carriers with magnetic resonance imaging, ultrasound, mammography, and clinical breast examination. JAMA 2004;292:2735–2742.[Abstract/Free Full Text]
  15. Kriege M, Brekelmans CT, Boetes C, et al. Magnetic Resonance Imaging Screening Study Group: efficacy of MRI and mammography for breast-cancer screening in women with a familial or genetic predisposition. N Engl J Med 2004;351:427–437.[Abstract/Free Full Text]
  16. Van den Brink JS, Watanabe Y, Kuhl CK, et al. Implications of SENSE MR in routine clinical practice. Eur J Radiol 2003;46(1):3–27.[CrossRef][Medline]
  17. Huber ME, Kozerke S, Pruessmann KP, Smink J, Boesiger P. Sensitivity-encoded coronary MRA at 3T. Magn Reson Med 2004;52(2):221–227.[CrossRef][Medline]
  18. 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]
  19. Lin W, An H, Chen Y, et al. Practical consideration for 3T imaging. Magn Reson Imaging Clin N Am 2003;11:615–639.[CrossRef][Medline]
  20. Pattany PM. 3T MR imaging: the pros and cons. AJNR Am J Neuroradiol 2004;25:1455–1456.[Free Full Text]
  21. 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]
  22. 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]
  23. Saekho S, Boada FE, Noll DC, Stenger VA. Small tip angle three-dimensional tailored radiofrequency slab-select pulse for reduced B1 inhomogeneity at 3 T. Magn Reson Med 2005;53:479–484.[CrossRef][Medline]
  24. Agoston AT, Daniel BL, Herfkens RJ, et al. Intensity-modulated parametric mapping for simultaneous display of rapid dynamic and high-spatial-resolution breast MR imaging data. RadioGraphics 2001;21:217–226.[Abstract/Free Full Text]
  25. Baum F, Fischer U, Vosshenrich R, Grabbe E. Classification of hypervascularized lesions in CE MR imaging of the breast. Eur Radiol 2002;12:1087–1092.[CrossRef][Medline]
  26. Leong CS, Daniel BL, Herfkens RJ, et al. Characterization of breast lesion morphology with delayed 3DSSMT: an adjunct to dynamic breast MRI. J Magn Reson Imaging 2000;11:87–96.[CrossRef][Medline]
  27. Liu PF, Debatin JF, Caduff RF, Kacl G, Garzoli E, Krestin GP. Improved diagnostic accuracy in dynamic contrast enhanced MRI of the breast by combined quantitative and qualitative analysis. Br J Radiol 1998;71(845):501–509.[Abstract]
  28. Nunes LW, Schnall MD, Siegelman ES, et al. Diagnostic performance characteristics of architectural features revealed by high spatial–resolution MR imaging of the breast. AJR Am J Roentgenol 1997;169:409–415.[Abstract/Free Full Text]
  29. Schnall MD, Ikeda DM. Lesion diagnosis working group report. J Magn Reson Imaging 1999;10(6):982–990.[CrossRef][Medline]
  30. Kinkel K, Helbich TH, Esserman LJ, et al. Dynamic high-spatial-resolution MR imaging of suspicious breast lesions: diagnostic criteria and interobserver variability. AJR Am J Roentgenol 2000;175(1):35–43.[Abstract/Free Full Text]
  31. Ikeda DM, Hylton NM, Kinkel K, et al. Development, standardization, and testing of a lexicon for reporting contrast-enhanced breast magnetic resonance imaging studies. J Magn Reson Imaging 2001;13:889–895.[CrossRef][Medline]
  32. Kuhl CK, Mielcarek P, Klaschik S, et al. Are signal time course data useful for differential diagnosis of enhancing lesions in dynamic breast MR imaging? Radiology 1999;211:101–110.
  33. 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]
  34. 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]
  35. Manka C, Traber F, Gieseke J, Schild HH, Kuhl CK. Three-dimensional dynamic susceptibility-weighted perfusion MR imaging at 3.0T: feasibility and contrast agent dose. Radiology 2005;234:869–877.[Abstract/Free Full Text]
  36. Schar M, Kozerke S, Fischer SE, Boesiger P. Cardiac SSFP imaging at 3 Tesla. Magn Reson Med 2004;51:799–806.[CrossRef][Medline]
  37. Sosna J, Pedrosa I, Dewolf WC, Mahallati H, Lenkinski RE, Rofsky NM. MR imaging of the prostate at 3 Tesla: comparison of an external phased-array coil to imaging with an endorectal coil at 1.5 Tesla. Acad Radiol 2004;11:857–862.[CrossRef][Medline]
  38. Morakkabati-Spitz N, Gieseke J, Kuhl C, et al. 3.0-T high-field magnetic resonance imaging of the female pelvis: preliminary experiences. Eur Radiol 2005;15(4):639–644.[CrossRef][Medline]
  39. Sommer T, Hackenbroch M, Hofer U, et al. Coronary MR angiography at 3.0 T versus that at 1.5 T: initial results in patients suspected of having coronary artery disease. Radiology 2005;234:718–725.[Abstract/Free Full Text]
  40. 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(3):1014–1019.[Abstract/Free Full Text]
  41. Tofts PS, Berkowitz B, Schnall MD. Quantitative analysis of dynamic Gd-DTPA enhancement in breast tumors using a permeability model. Magn Reson Med 1995;33(4):564–568.[Medline]
  42. Huang W, Fisher PR, Dulaimy K, Tudorica LA, O'Hea B, Button TM. Detection of breast malignancy: diagnostic MR protocol for improved specificity. Radiology 2004;232:585–591.[Abstract/Free Full Text]
  43. Meisamy S, Bolan PJ, Baker EH, et al. Neoadjuvant chemotherapy of locally advanced breast cancer: predicting response with in vivo (1)H MR spectroscopy—a pilot study at 4 T. Radiology 2004;233:424–431.[Abstract/Free Full Text]
  44. Stanwell P, Gluch L, Clark D, et al. Specificity of choline metabolites for in vivo diagnosis of breast cancer using 1H MRS at 1.5 T. Eur Radiol 2005;15(5):1037–1043.[CrossRef][Medline]
  45. Guo Y, Cai YQ, Cai ZL, et al. Differentiation of clinically benign and malignant breast lesions using diffusion-weighted imaging. J Magn Reson Imaging 2002;16:172–178.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 Br. J. Radiol.Home page
A M BLAMIRE
The technology of MRI -- the next 10 years?
Br. J. Radiol., August 1, 2008; 81(968): 601 - 617.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
C. E. Mountford, P. Stanwell, S. Ramadan, C. K. Kuhl, H. Kooijman, J. Gieseke, and H. H. Schild
Breast MR Imaging at 3.0 T
Radiology, July 1, 2008; 248(1): 319 - 320.