| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Breast Imaging |
1 From the MR Center, Medical Division (K.A.K., J.R., J.V., O.H.), Departments of Pathology (H.B.S.), Oncology (S.L.), and Surgery (H.E.F.), Trondheim University Hospital, N-7006, Trondheim, Norway. Received June 29, 1999; revision requested August 11; final revision received December 13; accepted January 11, 2000. Supported by a grant from the Norwegian Cancer Society. Address correspondence to K.A.K. (e-mail: kjell.kvistad@mr.ntnu.no).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: One hundred thirty patients with breast tumors underwent MR imaging with dynamic contrast-enhanced T1-weighted imaging of the entire breast, which was followed immediately with single-section, T2*-weighted imaging of the tumor.
RESULTS: With T2*-weighted perfusion imaging, 57 of 72 carcinomas but only four of 58 benign lesions had a signal intensity loss of 20% or more during the first pass, for a sensitivity of 79% and a specificity of 93%. With dynamic contrast-enhanced T1-weighted imaging, 64 carcinomas and 19 benign lesions showed a signal intensity increase of 90% or more in the first image obtained after the administration of contrast material, for a sensitivity of 89% and a specificity of 67%.
CONCLUSION: T2*-weighted first-pass perfusion imaging can help differentiate between benign and malignant breast lesions with a high level of specificity. The combination of T1-weighted and T2*-weighted imaging is feasible in a single patient examination and may improve breast MR imaging.
Index terms: Breast neoplasms, diagnosis, 00.311, 00.3115, 00.312, 00.319, 00.321, 00.324, 00.327, 00.329 • Breast neoplasms, MR, 00.121412, 00.12143, 00.12149 • Magnetic resonance (MR), perfusion study, 00.121419, 00.12144
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
In the results of several large studies, sensitivities of 83%–96% for the detection of breast carcinomas were reported (3–6). However, the specificity with which contrast material enhancement can be used to predict malignancy remains variable and ranges from 37% to 89% (5–11). The limited specificity of contrast-enhanced MR imaging observed in several studies (8,12,13) can be attributed to the fact that several benign breast lesions also can show strong contrast agent enhancement. Several strategies have evolved to improve the specificity of breast MR imaging: One approach has been to emphasize the morphology of the lesions (14–17); other strategies are pharmacokinetic analysis of the time-signal intensity curves (18,19) and very rapid single-section dynamic MR imaging, in which breast coverage is traded against high temporal resolution (20).
In recent years, yet another method, one that uses dynamic T2*-weighted first-pass perfusion imaging, has been proposed. The results of studies of small and selected groups of patients have indicated that the use of this imaging sequence can increase the differentiation between benign and malignant breast lesions (21,22). This conclusion is supported by the results of an animal study (23), in which the use of 14 previously proposed T1-weighted dynamic contrast-enhanced imaging sequences failed to show a difference between experimentally induced fibroadenomas and breast carcinomas in rats, while T2*-weighted first-pass perfusion imaging allowed differentiation between the benign and malignant tumors. To our knowledge, the diagnostic value of T2*-weighted first-pass perfusion imaging in breast lesions has not been evaluated previously in a large patient study.
The purpose of the present study was to evaluate if a combination protocol of T1-weighted dynamic contrast-enhanced imaging of the entire breast followed by T2*-weighted rapid single-section imaging is feasible in a large patient study and to determine if T2*-weighted first-pass perfusion imaging of the suspicious breast lesions can be used to provide useful diagnostic information in addition to that obtained with T1-weighted dynamic contrast-enhanced imaging.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Mammography was performed in 124 patients; 86 had masses, 12 had masses and calcifications, and 26 had no abnormalities. US was performed in 90 patients and demonstrated masses in 71 patients, masses and cystic components in four patients, and no abnormalities in 15 patients.
Patients with microcalcifications detected at mammography but with no other abnormalities and patients with lesions diagnosed at US as simple cysts were not eligible for the study. Patients who, because of old age or poor physical condition, were thought to be unable to lie in a prone position during the examination were not included in the study. The patients’ ages and menstrual statuses were registered, but no attempts were made to perform imaging in premenopausal women during certain weeks of their menstrual cycle or to interrupt hormone replacement therapy in postmenopausal women.
The final diagnosis was made with histopathologic examination of surgically excised specimens in 96 patients and was made in 31 patients with a combination of fine-needle aspiration cytologic analysis, mammography, and breast US, with a mean follow-up of 18 months (follow-up range, 11–23 months). In the remaining three patients, mammography and US but no biopsy or fine-needle aspiration cytologic analysis were performed. At the time this article was written, these patients had been observed for more than 18 months, and the lesions were classified as benign.
In patients in whom histopathologic examination showed malignancy, the assessment of the tumor size and the grade of the tumor were registered. Approximately 80% of all patients with breast cancer admitted to Trondheim University Hospital in the study period were examined; thus, the study results are supposed to be representative of the hospital’s entire population of patients with breast cancer. Since the presence of a solid breast tumor was emphasized as necessary for study inclusion, patients who underwent fine-needle aspiration cytologic analysis because of diffuse breast abnormalities or because of cystic lesions detected at mammography or at clinical examination were not referred for MR imaging and were not included in the study. The most frequent reasons for exclusion from the study were claustrophobia, poor physical condition, or lack of available time slots for MR imaging.
Seven patients had locally advanced breast carcinomas (T3 or T4 tumors) and received neoadjuvant chemotherapy prior to surgery. These patients were examined with MR imaging before chemotherapy was initiated and at several times during treatment. The imaging results from the first examination were used in this study.
MR Imaging
All images were acquired at 1.5 T Edge EPI II; Picker, Cleveland, Ohio) by using a commercially available, dedicated, receive-only double breast coil (Picker). The patients were examined while in a prone position, and the breasts were cushioned gently to reduce patient motion.
Dynamic contrast-enhanced T1-weighted images in the sagittal plane were obtained by using a three-dimensional radio-frequency spoiled gradient-echo sequence with a repetition time of 9.0 msec, an echo time of 3.8 msec (9.0/3.8), a flip angle of 30°, one signal acquired, a field of view of 250 mm, and an acquisition matrix of 128 x 256. The three-dimensional volume covered the whole breast and the axilla and was obtained with 44 partitions, which corresponded to an effective section thickness of 4 mm. The three-dimensional sequence was repeated continuously nine times, with a temporal resolution of 57 seconds; during the last 10 seconds of the acquisition of the third set of images, a bolus injection of 0.1 mmol/kg of body weight gadodiamide (Omniscan; Nycomed Amersham, Oslo, Norway) was administered through a needle in the antecubital vein and was followed with a 20-mL flushing bolus of isotonic saline solution. The total injection time was 10 seconds.
Immediately after image acquisition, the dynamic contrast-enhanced T1-weighted images (native and subtraction) were reviewed by a radiologist (K.A.K.) with experience in breast MR imaging and with knowledge of the patients’ clinical and mammographic findings. In all patients in whom a distinct lesion that corresponded to the clinical or mammographic findings was detected on the dynamic contrast-enhanced T1-weighted images, a single section in the most contrast–enhancing part of the tumor was chosen for T2*-weighted first-pass perfusion imaging. If the lesion was detected on the T1-weighted images obtained before contrast material enhancement but showed no subsequent enhancement, the single section was positioned in the center of the lesion. In four patients in whom no lesion was detected on the T1-weighted images before or after contrast material injection, the single section was positioned to correspond as closely as possible to the clinical or mammographic finding.
For T2*-weighted first-pass perfusion imaging, a gradient-echo sequence (54/35, 10° flip angle, one signal acquired, 250-mm field of view, 92 x 256 acquisition matrix, 5-mm section thickness, in the sagittal plane, with a resolution of 4.8 seconds) was used. The sequence was repeated 40 times. After the first 10 repetitions, a rapid (<3 seconds) injection of 0.1 mmol/kg gadodiamide was administered immediately and was followed with a 20-mL flushing bolus of isotonic saline. The time between the contrast agent injection for T1-weighted imaging and the contrast agent injection for T2*-weighted imaging was approximately 15 minutes. The time for the entire imaging study was approximately 25 minutes.
Image Analysis
The images were analyzed independently by two radiologists (J.V. and J.R.) who were not involved in patient inclusion or in image acquisition. The radiologists knew that the patients had breast lesions but did not know the histopathologic findings or patient outcomes. Since only a single section was available with T2*-weighted perfusion imaging, it was not possible to blind the readers with respect to the location of the lesion. In cases of discrepancy between the two readers, a consensus was reached. On the dynamic contrast-enhanced T1-weighted images, a region of interest was positioned in the most contrast-enhancing part of the lesion, and time–signal intensity curves were obtained. For the selection of the most contrast-enhancing part of the tumor for the positioning of the single section in T2*-weighted imaging, regions of interest were placed in the lesion in different sections by the radiologist (K.A.K.) immediately after the acquisition of the dynamic T1-weighted images. These regions of interest were documented on the hard copies and were made available for the independent readers (J.V. and J.R.), who also drew their own regions of interest and chose for further analysis the region of interest that showed the strongest and most rapid signal intensity increase in the lesion.
The time–signal intensity curve for each lesion was analyzed in two ways: In the first analysis, in accordance with commonly used diagnostic criteria (2,10,11), the signal intensity increase within the first image acquisition (57 seconds) was calculated as a percentage of the precontrast signal intensity. The second analysis consisted of a subjective evaluation of the time–signal intensity curve, in which each curve was classified in accordance with the evaluation system shown in Figure 1. This evaluation system is learned easily and, in a study by Daniel et al (24), achieved a very good diagnostic performance in differentiating between breast cancers and benign abnormalities.
|
On the T2*-weighted first-pass perfusion images, a region of interest was positioned in the part of the tumor that enhanced the most on the T1-weighted images, and time–signal intensity curves were obtained. The maximum signal intensity loss within the first 30 seconds after bolus injection was calculated as a percentage of the baseline signal intensity values from the 10 precontrast images.
Statistical Analysis
The Student t test was performed for the statistical comparison of the signal intensity increases in the group of carcinomas and in the group of benign lesions during the first postcontrast image acquisition with T1-weighted imaging. Because of unequal variances between the carcinomas and the benign lesions, the nonparametric Mann-Whitney test was performed for the statistical comparison of the maximum signal intensity losses for benign versus malignant lesions during the first 30 seconds after contrast material injection in T2*-weighted first-pass perfusion imaging. The Pearson correlation coefficient (r) was calculated in the carcinomas to determine the correlation between the signal intensity increase in T1-weighted imaging and the signal intensity loss in T2*-weighted imaging. Statistical analysis was performed by using SPSS software version 8.0 (SPSS, Chicago, Ill).
The following parameters were calculated by using these formulas: Sensitivity = TP x 100/(TP + FN); specificity = TN x 100/(TN + FP); positive predictive value = TP x 100/(TP + FP); and negative predictive value = TN x 100/(TN + FN), where TP is true-positive findings, TN is true-negative findings, FP is false-positive findings, and FN is false-negative findings.
Receiver operating characteristic (ROC) analysis was performed to assess the diagnostic performances of the signal intensity increase on the T1-weighted dynamic images and of the signal intensity loss on the T2*-weighted first-pass perfusion images (25). The area beneath the fitted binormal ROC curve (Az) was used as a measure of diagnostic efficacy. The Az values were calculated and were compared by using ROCKIT software program version 9.0B (C.E. Metz, University of Chicago, Ill). A P value of .05 was considered to indicate a significant difference.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
In the remaining 58 patients, aged 17–79 years (mean age, 47 years), malignancy was ruled out. Thirty-one of these patients were premenopausal and 27 were postmenopausal. Eleven of the postmenopausal women received hormone replacement therapy. Surgical excision biopsy was performed in 25 patients. Histologic examination showed fibroadenoma in 14 patients, fibrocystic disease in two, sclerosing adenosis in two, hemangioma in one, epithelial hyperplasia in one, and fibrosis or scar in five. In 30 patients, the final diagnosis was determined with fine-needle aspiration cytologic analysis, clinical examination, mammography, and US (mean follow-up, 18 months). Fine-needle aspiration cytologic analysis showed fibroadenoma in eight patients, fibrocystic disease in six, fatty tissue necrosis in three, papilloma in two, lipoma in two, epithelial hyperplasia in one, and normal ductal epithelium in eight.
In 126 patients, dynamic contrast-enhanced T1-weighted images showed a lesion that corresponded to the lesion location at clinical examination and/or at mammography. In the four patients in whom no lesion could be detected, the final diagnosis was fatty tissue necrosis in two, lipoma in one, and fibrosis in one. In these patients, the positions of the regions of interest were uncertain, but, to not exclude these patients, a region of interest was positioned in the tissue that corresponded approximately to the palpable or mammographically detected lesions on both the T1-weighted and T2*-weighted images.
The analysis of the dynamic contrast-enhanced T1-weighted images showed a mean signal intensity increase during the acquisition of the first postcontrast images of 179% ± 87 SD (95% CI: 158%, 199%) in the carcinomas and of 83% ± 74 (95% CI: 63%, 103%) in the benign lesions. Although this difference was highly significant (P < .001), a considerable overlap between the signal intensity increase in the carcinomas and that in the benign lesions was found (Fig 2a).
|
|
|
In the subjective evaluation of the time–signal intensity curves with the system shown in Figure 1, 63 of the carcinomas were classified as having type III, IV, or V curves, while 46 of the benign lesions were classified as having type I or II curves. The nine carcinomas with time–signal intensity curves classified as type I or II had a mean size of 1.5 cm. If the malignancy criteria were restricted to lesions with curve type IV or V, 51 of the carcinomas and six of the benign lesions would fall into this category. Curve type V was observed in 37 carcinomas and in only two benign lesions.
Sixty of 72 lesions with spiculated borders and 53 of 65 lesions with peripheral contrast material enhancement were malignant.
The analysis of the T2*-weighted first-pass perfusion images showed a mean maximum signal intensity loss in carcinomas of 31% ± 15 during the first 30 seconds after contrast material injection (95% CI: 28%, 35%) (Fig 2b). On average, the signal intensity loss occurred 15–20 seconds after the bolus contrast agent injection (Fig 3). The mean signal intensity loss in the benign lesions was 9% ± 7 (95% CI: 7%, 11%) (Fig 4). The difference in signal intensity decrease between the carcinomas and the benign lesions was highly significant (P < .001), and the overlap between the signal intensity decrease in the carcinomas and that in the benign lesions on the T2*-weighted images was less pronounced than the overlap in signal intensity increase on the T1-weighted images. If a signal intensity decrease of 20% or more was chosen as a threshold for malignancy, 57 of 72 carcinomas and 54 of 58 benign lesions were classified correctly, which resulted in a specificity of 93% and a sensitivity of 79%.
|
|
|
|
|
In carcinomas, there was a strong correlation between the signal intensity increase during postcontrast imaging with the T1-weighted sequence and the signal intensity loss after contrast material injection with the T2*-weighted sequence (r = 0.48, P < .001) (Fig 5); five of the carcinomas were outside the chosen threshold values for malignancy with both imaging sequences.
|
ROC analysis (Fig 6) showed that the signal intensity decrease during the first 30 seconds after contrast material injection on the T2*-weighted first-pass perfusion images had a significantly greater diagnostic value than the signal intensity increase on the first postcontrast dynamic T1-weighted image (Az = 0.909 vs 0.841, respectively; P = .03).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Investigators in studies (31,32) of the microvascular architecture in breast carcinomas have demonstrated an increased number of capillaries and a greater mean capillary diameter compared with those in normal tissue. Increased vascularity and capillary diameter increase the fractional volume of the intravascular space in carcinomas compared with other tissues and provide a biologic explanation for the high susceptibility effects observed at T2*-weighted imaging of breast carcinomas (21–23).
To obtain a standard contrast agent concentration in the arterial blood is difficult, as it depends on several factors, among them cardiac output (30); in the present study, the contrast agent dose was correlated with the patient’s body weight to approximate a standard arterial blood concentration. In the present study, it was not possible to determine the regional blood volume or the mean transit time in the first-pass perfusion sequence, since gadodiamide leaks rapidly through the capillary endothelium, and an estimated 50% of the injected contrast agent enters the extracellular space after the first pass through the capillary network (33). This leakage counteracts the desired signal intensity losses because of the diminution of the concentration gradient between the vessels and the surrounding tissues and because of the T1 shortening effects of the paramagnetic contrast agent in the extracellular space (30). Despite these issues, a strong signal intensity loss after contrast agent injection on the T2*-weighted perfusion images in the carcinomas was observed, even though residual contrast agent from the contrast-enhanced T1-weighted imaging performed an average of 15 minutes earlier was present in the lesions. This confirms the experience of investigators in two previously published studies with smaller series of patients (21,22) that T2*-weighted first-pass perfusion imaging can be performed with good results immediately after contrast-enhanced T1-weighted imaging. This is important, since the coverage of T2*-weighted imaging with current techniques is limited to one or a few sections, and breast tumors frequently are difficult to differentiate from the surrounding parenchyma on MR images without the use of paramagnetic contrast agents.
While breast carcinomas demonstrated a signal intensity loss after contrast agent injection in T2*-weighted imaging, this phenomenon was absent or small in most benign breast lesions. The mean signal intensity loss in benign breast lesions was 9% compared with the baseline value and in most cases was not different from the noise level. A more pronounced loss of 20% or higher, indicating high vascularization, was found only in three fibroadenomas and in one papilloma, which resulted in a specificity of 93% at this threshold (Table). Since the patients in whom these four lesions were found were premenopausal or were receiving hormone replacement therapy, all of the benign lesions in the postmenopausal women who were not receiving exogenous hormones were classified correctly.
Compared with the series of patients undergoing biopsy in the study by Azavedo et al (34), a lower proportion of patients with benign lesions was included in this study, with a probable bias toward lesions with a high likelihood for malignancy. This inclusion bias may have influenced the specificity of both T2*-weighted perfusion imaging and T1-weighted contrast-enhanced imaging.
Another issue that might have influenced the specificity of MR imaging was the decision to not interrupt hormone replacement therapy or to restrict the examinations of the premenopausal women to menstrual cycle weeks 2 and 3, in which the vascularization and the contrast material enhancement in the normal breast parenchyma are lowest (26). The mean age was lower and the proportion of premenopausal women was higher in patients with benign lesions, and if the menstrual cycle also affects contrast material enhancement and vascularization in benign breast lesions, the possibility of more false-positive cases may arise for both T1-weighted and T2*-weighted imaging. Although the mean follow-up period of 18 months in the patients with benign lesions probably was not long enough to completely exclude the possibility that any of these breast masses eventually could prove to be malignant, the combination of fine-needle aspiration cytologic analysis, mammography, and US is very reliable in experienced hands (35). The unlikely event that any of these lesions that were classified as benign were malignant probably would not change the main conclusions in this relatively large study.
The sensitivity of T2*-weighted first-pass perfusion imaging was 79% with the chosen malignancy threshold. While the specificity of first-pass T2*-weighted perfusion imaging was highest in postmenopausal patients, the sensitivity was higher in premenopausal patients, with a signal intensity loss of 20% or more in 18 of 21 patients. Two of four mucinous carcinomas, both tubular carcinomas, and two of five carcinomas in situ showed a first-pass perfusion effect of less than 20%. These histologic subtypes previously also were reported as false-negative findings with dynamic contrast-enhanced T1-weighted imaging (3,36,37), probably because of low tumor vascularization, which also explains the low first-pass perfusion effect. Other issues that could have influenced the measured tumor first-pass perfusion effects and that could have reduced the sensitivity of the method are artifacts due to susceptibility between fatty tissue, breast parenchyma, and tumor tissue; a low spatial resolution; or patient movement between dynamic T1-weighted and T2*-weighted imaging.
However, the results from the ROC analysis (Fig 6) confirm that, although the sensitivity of dynamic contrast-enhanced T1-weighted imaging was higher than that of T2*-weighted first-pass perfusion imaging, the diagnostic value of the latter was significantly higher. Improved sensitivity of T2*-weighted first-pass perfusion imaging can possibly be reached by increasing the measured T2* effects. One way of doing this is to increase the gadolinium-based contrast agent dose. The recommended gadolinium-based contrast agent dose in T2*-weighted first-pass perfusion imaging of the brain is 0.2 mmol/kg (31), but since the patients in the present study received 0.1 mmol/kg gadodiamide for dynamic T1-weighted imaging, we chose to limit the gadodiamide dose to 0.1 mmol/kg during T2*-weighted perfusion imaging.
Another way of increasing T2* effects is by prolonging the echo time in the gradient-echo sequence, but in a previous study by Kuhl et al (21), severe image degradation from motion and susceptibility artifacts when the echo time was increased to 40 msec was reported. In the near future, the use of true intravascular contrast agents that currently are in development (38) probably could minimize leakage from the capillaries during the first pass, and semiquantitative measurements such as regional blood volume and mean transit time may be extracted from the time–signal intensity curves in T2*-weighted first-pass perfusion imaging. It is also possible that T2*-weighted first-pass perfusion imaging, in addition to use in differentiating benign and malignant breast tumors, may have a future role in the evaluation of the effects of antiangiogenic drugs that are currently in clinical trials (39).
In dynamic contrast-enhanced T1-weighted imaging, a lesion corresponding to the clinical or mammographic findings was detected in 126 patients. The four patients in whom no lesion could be detected on the MR images all had benign lesions, which confirmed that the absence of an MR imaging–visible abnormality has a high negative predictive value (14). All but one of the 72 carcinomas in this study showed contrast material enhancement with T1-weighted imaging, and all carcinomas were detectable with MR imaging. A sensitivity of 89% and a specificity of 67% were achieved with a malignancy threshold of a 90% signal intensity increase during the acquisition of the first postcontrast image; this is comparable to findings in previous large studies (5,8,10), in which specificity values of 37%–89% were reported, and with findings in studies (3,5,11) in which strong and rapid contrast material enhancement was detected in 81%–95% of breast carcinomas.
A simple classification system for the subjective assessment of the entire time–signal intensity curve was introduced recently in a study by Daniel et al (24), with results comparable to those achieved with sophisticated pharmacokinetic analysis (Fig 1). When the results from the present study were evaluated according to this system, 109 of 130 lesions were classified correctly, and a specificity of 79% was achieved (Table). These results indicate that a significant improvement in specificity can be gained by evaluating the shape of the time–signal intensity curve compared with using a threshold value for the signal intensity increase during the acquisition of the first postcontrast image.
The presence of peripheral contrast material enhancement and spiculated lesion margins on the dynamic T1-weighted images correlated well with malignancy in the present study (Table); 53 and 60 of the carcinomas, respectively, had these features. Fischer et al (10) found only 22 of 42 cancers with peripheral contrast material enhancement, and Nunes et al (14) found only 10 of 32 cancers with spiculation, but the evaluation of these features is subjective, and the high number of carcinomas with these features in the present study may be explained by differences in the application of these criteria.
In conclusion, T2*-weighted first-pass perfusion imaging performed immediately after dynamic contrast-enhanced T1-weighted imaging can be used to differentiate between benign and malignant breast tumors with a higher degree of certainty than any of the applied analysis methods based on T1-weighted imaging. The high specificity of T2*-weighted first-pass perfusion imaging probably reflects differences in vascularization, with a higher vessel number and with a larger vessel diameter in carcinomas. The sensitivity of T2*-weighted first-pass perfusion imaging was not optimal and could probably be increased with technical improvements in the method. The combined use of T1-weighted dynamic contrast-enhanced imaging and T2*-weighted first-pass perfusion imaging in a single session has the potential to improve the clinical utility of MR imaging of the breast.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, K.A.K., O.H.; study concepts and design, K.A.K., O.H.; definition of intellectual content, K.A.K., O.H.; literature research, K.A.K.; clinical studies, K.A.K., S.L., H.E.F., H.B.S.; data acquisition, K.A.K.; data analysis, K.A.K., J.R., J.V.; statistical analysis, K.A.K.; manuscript preparation, K.A.K.; manuscript editing, K.A.K., O.H.; manuscript review, all authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Raza, M. Vallejo, S. A. Chikarmane, and R. L. Birdwell Pure Ductal Carcinoma in Situ: A Range of MRI Features Am. J. Roentgenol., September 1, 2008; 191(3): 689 - 699. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. K. Kuhl, F. Traber, J. Gieseke, W. Drahanowsky, N. Morakkabati-Spitz, W. Willinek, M. von Falkenhausen, C. Manka, and H. H. Schild Whole-Body High-Field-Strength (3.0-T) MR Imaging in Clinical Practice * Part II. Technical Considerations and Clinical Applications Radiology, April 1, 2008; 247(1): 16 - 35. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Medved, G. M. Newstead, H. Abe, M. A. Zamora, O. I. Olopade, and G. S. Karczmar High Spectral and Spatial Resolution MRI of Breast Lesions: Preliminary Clinical Experience Am. J. Roentgenol., January 1, 2006; 186(1): 30 - 37. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Sardanelli, A. Iozzelli, A. Fausto, A. Carriero, and M. A. Kirchin Gadobenate Dimeglumine-enhanced MR Imaging Breast Vascular Maps: Association between Invasive Cancer and Ipsilateral Increased Vascularity Radiology, June 1, 2005; 235(3): 791 - 797. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Huang, P. R. Fisher, K. Dulaimy, L. A. Tudorica, B. O'Hea, and T. M. Button Detection of Breast Malignancy: Diagnostic MR Protocol for Improved Specificity Radiology, August 1, 2004; 232(2): 585 - 591. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. F. Schaefer, J. Vollmar, F. Schick, R. Vonthein, M. D. Seemann, H. Aebert, R. Dierkesmann, G. Friedel, and C. D. Claussen Solitary Pulmonary Nodules: Dynamic Contrast-enhanced MR Imaging--Perfusion Differences in Malignant and Benign Lesions Radiology, August 1, 2004; 232(2): 544 - 553. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-P. Delille, P. J. Slanetz, E. D. Yeh, D. B. Kopans, and L. Garrido Breast Cancer: Regional Blood Flow and Blood Volume Measured with Magnetic Susceptibility-based MR Imaging—Initial Results Radiology, March 21, 2002; (2002) 2232010428. [Abstract] [Full Text]
|
|
|
|
|
|
J.-P. Delille, P. J. Slanetz, E. D. Yeh, D. B. Kopans, and L. Garrido Breast Cancer: Regional Blood Flow and Blood Volume Measured with Magnetic Susceptibility-based MR Imaging—Initial Results Radiology, May 1, 2002; 223(2): 558 - 565. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
