| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Breast Imaging |
1 From the Departments of Radiology (H.M.Z., A.E.v.V.), Medical Statistics (J.H.), and Pathology (M.J.v.d.V.), Leiden University Medical Center, Albinusdreef 2, 2333 ZA, Leiden, the Netherlands; and the Department of Radiology, Medical Center Haaglanden-Westeinde, The Hague, the Netherlands (E.G.C.). Received August 11, 1998; revision requested September 25; final revision received February 16, 1999; accepted June 8. Address reprint requests to H.M.Z.
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: In a 2-year prospective study, 4,811 mammograms were classified according to level of suspicion of malignancy. Targeted US was performed to analyze (a) circumscribed lesions, possibly cysts; (b) palpable lesions visible at mammography; (c) palpable lesions not visible at mammography; and (d) nonpalpable lesions visible at mammography. After US was performed in 1,103 cases (23%), cases were reclassified for level of suspicion.
RESULTS: In 338 cases, breast cancer was diagnosed. The sensitivity of mammography for all 4,811 cases was 83%; the specificity was 97%. After US, the combined sensitivity increased to 91%, with a specificity of 98%. The increase was significant (P < .001). The increase in sensitivity was highest among women younger than 50 years. The positive predictive value for mammography was high (72%), which reflects a high threshold for biopsy; this may have augmented the yield of US.
CONCLUSION: The use of US as an adjunct to mammography resulted in an increase in diagnostic accuracy. Its contribution to the diagnosis of breast cancer in this study was 7.4%.
Index terms: Breast neoplasms, 00.30 • Breast neoplasms, diagnosis, 00.11, 00.1298, 00.30 • Breast neoplasms, US, 00.1298 • Receiver operating characteristic (ROC) curve
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Studies performed to evaluate US as a screening modality have failed to establish its efficacy, and it has been concluded that US should not be used as a screening tool (2,4). In the Netherlands, a national biennial screening program is available for asymptomatic women aged 50–69 years. The examination is performed in mobile mammography units, and US does not play a role in this examination. However, in a population consisting of patients with a large variety of complaints and indications, targeted US is frequently used as an adjunct to mammography.
The purpose of the present study was to determine the diagnostic yield of US when used as an adjunct to mammography for the diagnosis of breast cancer in an outpatient population.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
In total, 4,728 consecutive patients who underwent mammography entered the study prospectively; 3,546 (75%) patients were from the university medical center, and 1,182 (25%) were from the affiliated hospital. Patients who underwent mammography twice during the study period were included only once, unless they presented with another symptom or lesion. Patients who presented with a second or third lesion were included two or three times, respectively. Consequently, 4,649 patients entered the study once, 75 patients were included in the study twice, and four patients were included in the study three times. In total, 4,811 mammographic examinations in 4,728 patients form the basis of the study.
The mean age of the patients was 51 years (age range, 30–94 years). Patients younger than 30 years were excluded because mammography was not performed systematically in this age group; often, the examination consisted of US alone. Also excluded were patients with clinically advanced symptoms of breast cancer and those with diffuse mastitis, abscess formation, or silicone gel implant complications.
In total, 950 (20%) patients had previously been treated for breast carcinoma—447 patients with mastectomy and 503 with conservative surgery.
The indications for mammography were (a) the presence of a palpable mass, (b) an abnormality of skin or nipple, (c) pain or other complaints without a clear cause, (d) fear of cancer or hormone-replacement therapy in asymptomatic patients, (e) high-risk family history or history of radiation therapy for previous breast cancer, (f) evaluation of a lesion detected on the screening mammogram, (g) follow-up mammography because of a previously detected lesion defined as probably benign, and (h) the presence of metastases of an unknown primary.
Mammography was performed with dedicated equipment (Bennett Contour, Oldelft Benelux, Delft, the Netherlands; and Mammomat 3000, Siemens, Erlangen, Germany). Standard craniocaudal and mediolateral oblique views were routinely obtained. When the glandular tissue was incompletely demonstrated, an additional axillary tail view or an extended craniocaudal view was obtained. To differentiate superposition or asymmetry from a mass or architectural distortion, collimated compression views were obtained. Magnification views were obtained for better delineation of a mass or microcalcifications. In cases of microcalcifications, a horizontal beam view also was obtained. If the mammographic findings were not conclusive with respect to a palpable lesion, a radiopaque marker was placed on the site of the palpable lesion for better identification.
The mammograms were assessed by one of seven radiologists (including H.M.Z., E.G.C.) who were experienced because mammography was part of their daily task. The radiologists in the affiliated hospital had originally been trained at the university hospital; therefore, the procedures were comparable in the two institutions. No double reading was performed.
In the patient's case record, the parenchymal density of the glandular tissue was noted. It was considered of high density if lesions could easily be overlooked, medium density if lesions could possibly be overlooked, and low density if the glandular tissue was sufficiently spread out and intermingled with variable amounts of fat. Furthermore, clinically important morphologic findings were described according to the guidelines of the Dutch National Breast Imaging Consensus with adaptations from the literature (5–8): (a) the presence of a clinically important lesion; (b) its size and location; (c) its shape, margins, and density; (d) changes in comparison with previous mammograms; (e) the presence and characteristics of microcalcifications; and (f) the presence of features that are associated with malignancy, such as skin thickening, nipple retraction, and dense axillary nodes.
Prospective classification into one of five categories that represented five levels of suspicion of malignancy followed. Level 1 was defined as benign with no abnormalities seen. Level 2 was defined as probably benign. This indicated that a lesion or cluster of microcalcifications was present, with a high probability of being benign. A follow-up study at 6 months was recommended. Level 3 was defined as equivocal and indicated that a lesion or microcalcifications were present with indeterminate characteristics that necessitated histopathologic examination. Level 4 was defined as probably malignant. This indicated that a lesion or microcalcifications aroused suspicion of malignancy. Level 5 was defined as malignant and indicated that a highly suspicious lesion was present in combination with the associated features of malignancy, as described in point f in the previous paragraph.
The decision to perform US in addition to mammography was made by the radiologist who supervised the mammographic examination. US was performed by the same radiologist immediately after mammography to analyze circumscribed lesions that might be cysts, palpable lesions visible on mammograms, palpable lesions not visible on mammograms, or nonpalpable lesions visible on mammograms.
The latter category included nonpalpable masses and focal asymmetric areas that were difficult to distinguish from a mass. Asymmetric distribution of the glandular tissue in itself was not considered an indication for US. Also, microcalcifications were not considered an indication for US.
The US examination was performed with a high-frequency transducer (10-MHz sector scanner with built-in water path [Aloka model 650 CL; Biomedic, Almere, the Netherlands] or a 7.5-MHz linear-array transducer [Aloka 280, Biomedic]).
Of the 4,811 mammographic examinations, 1,103 (23%) were followed by US. The mean age of the patients who underwent US was 50 years (age range, 30–93 years). In each patient's case record, it was noted whether the lesion could be identified at US and described according to guidelines from the literature (4,6,9,10). A simple cyst was considered benign; a cystic lesion with debrislike contents was considered probably benign. A complex cyst with internal echoes and septa was classified as equivocal; a thickened wall and the presence of an intracystic mass were classified as probably malignant. A solid lesion with smooth or lobulated margins that was sharply defined, with homogeneous hypoechoic contents and a horizontal orientation was assumed to be a fibroadenoma and was classified as probably benign. Solid, hypoechoic lesions with irregular margins and an indeterminate or horizontal orientation but without a definite probability of being malignant or benign were classified as equivocal.
Hypoechoic lesions with irregular and poorly defined margins and with shadowing and vertical orientation were classified as probably malignant. Lesions highly suspected to be malignant with infiltration into the surrounding fatty tissue or other associated features of malignancy were classified as malignant.
An additional note was made if the glandular tissue showed areas of increased echogenicity intermingled with small cystic lesions and hypoechoic branching ductal structures concomitant with fibrocystic disease or if ductal ectasia was present.
Finally, reclassification was performed according to one of five levels of suspicion. The classification was now based on the results of the US examination in addition to and in accordance with the findings already obtained at mammography.
The definitive diagnosis of a lesion was obtained as follows.
1. The definitive diagnosis was obtained at histopathologic examination in cases for which a surgical procedure, core-needle biopsy, or fine-needle aspiration biopsy was performed. The results of fine-needle aspiration biopsy were accepted for the definitive diagnosis only if they revealed a specific benign or malignant diagnosis; an aspirate consisting of normal breast epithelial cells was considered insufficient for the definitive diagnosis. For patients who had undergone more than one procedure, the result of the most invasive procedure was used.
2. The diagnosis of a simple cyst was considered correct if US showed a simple cyst or if the presence of a cyst was proved with US-guided aspiration.
3. If the level of suspicion was benign or probably benign, a normal follow-up mammogram obtained after at least 12 months was considered adequate for establishing the definitive, benign diagnosis.
4. The Dutch National Archives of Pathology (PALGA) contains the results of all cytologic or histologic examinations performed in the Netherlands. The data on patients without known follow-up were verified with PALGA in July 1997, so that patients with a malignant lesion diagnosed or treated in other hospitals in the Netherlands could be traced. The mammograms of all other patients were considered true-negative; therefore, the definitive diagnosis was classified as benign.
By using the definitive diagnosis and the five level-of-suspicion categories, receiver operating characteristic (ROC) curves were drawn for (a) the results of all 4,811 mammographic examinations, (b) the combined results of all 4,811 mammographic and 1,103 US examinations, (c) the results of mammography in only the patients who underwent the 1,103 US examinations, and (d) the combined results of mammography and US for only the patients who underwent the 1,103 US examinations.
The area under the ROC curve (Az) was calculated by using the trapezium rule. The two Az values for the 4,811 examinations and those for the 1,103 examinations were compared by using the Wilcoxon signed rank test (11).
The positive predictive value (PPV) of the five levels of suspicion at mammography (the number of cancers per level of suspicion) was calculated separately. The sensitivity, the specificity, and the PPVs and negative predictive values (NPVs) were calculated for the prospective classification after mammography and for the prospective classification after mammography and US, whereby the categories "benign" and "probably benign" were considered to indicate benignity and "equivocal," "probably malignant," and "malignant" were considered to indicate malignancy. Sensitivity and specificity were studied as a function of the different age groups by decade. The subgroups with false-negative and false-positive results were analyzed in more detail.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Breast Cancer
A surgical procedure revealed breast cancer in 331 cases and metastasis of a thyroid carcinoma in one case.
The diagnosis of breast cancer in six cases was established by means of core-needle biopsy; fine-needle aspiration biopsy revealed the diagnosis of breast cancer in another 10 cases and metastasis of an ovarian carcinoma in one case. In 15 of these 17 cases, surgery was not performed because of age, poor general condition, or the presence of metastases. In one case, surgical treatment was performed abroad; in one case, the patient refused surgery.
Review of PALGA revealed breast cancer in five patients who were treated in another hospital in the Netherlands.
Therefore, of the 4,811 cases, 354 were malignant lesions: 312 invasive carcinomas, 40 in situ carcinomas, and two intramammary metastases. These metastases are henceforth referred to as "breast cancer" in this article. In 72 cases, the mammograms were considered to indicate a lesion that was benign or probably benign at the time of the study, but a breast cancer was diagnosed at further evaluation, at follow-up mammography, or at clinical examination.
Mammograms from these cases, together with all previous and new mammograms, were reviewed retrospectively by three radiologists (including H.M.Z., E.G.C.) in consensus after the diagnosis was obtained. A mammogram was considered true-negative if the cancer was nonpalpable and not visible, except for the cancers that manifested within 1 year of the study. According to this definition, 16 mammograms were considered to be true-negative. The other 56 mammograms were considered to be false-negative. So, for 338 cases (7%), the definitive diagnosis was breast cancer. Figure 1 shows the distribution of these 338 breast cancers as a function of patient age and stage. The PPV for the benign lesions was 1% (27 of 3,880 cases); that for the probably benign lesions, 5% (29 of 538 cases); that for the equivocal lesions, 35% (42 of 119 cases); that for the probably malignant lesions, 79% (127 of 160 cases); and that for the malignant lesions, 99% (113 of 114 cases).
|
US depicted a simple cyst in 91 cases. By means of US-guided aspiration, a cyst was diagnosed in another 224 cases.
By means of histopathologic examination, a fibroadenoma was diagnosed in 91 cases, and a nonspecific benign lesion (mastitis, galactocele, fibrocystic changes, and scar tissue) was diagnosed in 182 cases.
A subsequent mammogram obtained after an interval of at least 12 months with normal or benign findings was noted for 1,558 cases.
Review of the PALGA did not reveal either the diagnosis or the treatment of breast cancer for the other 2,327 cases.
Mammography and US
In Table 1, the indications for mammography and the resultant diagnoses of breast cancer are shown for all 4,811 cases. In Table 2, the indications for US and the resultant diagnoses of breast cancer are shown for the 1,103 cases in which US was performed. In Table 3, the definitive diagnoses of breast cancer versus benign disease are given for the 4,811 cases in which mammography was performed and the 1,103 cases in which US was performed. The 18 intraductal breast cancers consisted of three intracystic carcinomas, six palpable solid lesions, and nine nonpalpable solid lesions. In 11 cases, they were accompanied by microcalcifications. The distribution of the density of the glandular tissue as a function of age is shown in Figure 2. The percentage of cases per age group in which mammography or mammography and US were performed is shown in Figure 3.
|
|
|
|
|
|
|
|
|
Subgroup 1.—In 19 cases (mean patient age, 57 years; age range, 37–83 years), the lesion was not prospectively recognized on the mammogram; consequently, no US was performed. At review, the mammograms obtained in these cases were considered false-negative.
Subgroup 2.—In 12 cases (mean patient age, 56 years; age range, 44–82 years), targeted US was performed, but the US diagnosis was also false-negative.
The mammogram was considered normal in three cases; the lesion was recognized and considered probably benign in the other nine cases.
In eight cases, the lesion was identified by means of US but was interpreted incorrectly as fibrocystic disease or ductal ectasia. In one of these cases, the patient presented with bloody nipple discharge. Mammographic findings were normal, which was also determined in retrospect. US of the retroareolar area yielded a diagnosis of ductal ectasia with a possible papilloma; however, at surgery, ductal carcinoma in situ was diagnosed (Fig 6).
|
|
|
Subgroup 3.—In 25 cases (mean patient age, 51 years; age range, 32–80 years), targeted US was performed and the US diagnosis was true-positive.
For 19 cases, the indication for mammography was a palpable mass. In one case, the patient presented with an abnormality of the nipple. In five cases, patients were asymptomatic but underwent mammography for various other indications.
In 12 cases, the cancer was not delineated or recognized on the mammogram; in 13 cases, the cancer was visible on the mammogram and was described as a probably benign lesion.
US was performed because of a circumscribed lesion, possibly a cyst, in eight cases; a palpable lesion visible on the mammogram in six cases; a palpable lesion not visible on the mammogram in nine cases (Fig 7); and a nonpalpable lesion visible on the mammogram in two cases (Fig 8).
|
|
|
|
|
False-Positive Results
For 112 of 4,811 cases, the mammographic results were false-positive, which resulted in a PPV for mammography of 72% (Table 5). For 22 cases, there was no indication for US. For 45 cases, the level of suspicion after mammography was "equivocal" or "probably malignant," which was correctly converted into "benign" or "probably benign" because of US findings. Diagnoses in this subgroup were 18 cysts, one fibroadenoma, and eight nonspecific benign lesions. In the other 18 cases, only fibrocystic disease or normal glandular tissue was depicted at US; the follow-up study did not show signs of malignancy. In another 45 cases, the level of suspicion on both the mammogram and the US scan was "equivocal," "probably malignant," or "malignant." At histopathologic examination, diagnoses in this subgroup were three complicated cysts, eight atypical fibroadenomas, and 34 cases with nonspecific benign lesions.
For 23 cases, the level of suspicion on the mammogram was "benign" or "probably benign," but US resulted in a level of suspicion of "equivocal" in 17 cases and "probably malignant" in six cases. This subgroup was predominantly in symptomatic patients. Histopathologic examination revealed three complicated cysts filled with hemorrhagic or purulent debris, nine fibroadenomas with an atypical appearance, and 11 nonspecific benign lesions. US never resulted in conversion from a true-positive mammographic result to a false-negative result.
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Accuracy of Mammography in the Study
The contribution of US cannot be assessed without considering the accuracy of mammography because of the complementary character of these examinations. The sensitivity, specificity, and PPVs and NPVs of mammography depend on several factors, the most important of which is the composition of the study population and the prevalence of breast cancer in this population. Also, the number of symptomatic patients, the age distribution, and the tumor stage, in combination with the percentage of intraductal carcinomas, the duration of follow-up, and the definition of the true-negative mammogram, play a role (12).
The relatively high numbers of stage T1 breast cancer (13) in the groups of patients aged 50–59 years and 60–69 years may be due to the national biennial screening program (Fig 1). In the Netherlands, the nationwide biennial screening program is funded by the government. The compliance rate is high: 69%–82% (14). The screening units have to comply with the requirements of the National Evaluation Team; one of the requirements is to keep the false-positive rate low and minimize the number of unnecessary biopsies. As a consequence, the threshold for referral is high and contains only lesions with a high likelihood of malignancy. The percentage of referred cases that were cancer, 50% (72 of 144 cases) (Table 1), and the PPV of mammography, 72% (Table 5), are in accordance with the expectations of the National Evaluation Team (14). Three of the participating radiologists are engaged in the screening and were trained at the National Expert and Training Center, Nijmegen, the Netherlands; the expertise of this center also applies to the mammographic work-up of an outpatient population, as described in this article.
Malignant lesions are frequently evident in retrospect, even on more than one previous study (15). Therefore, we reviewed all mammograms that were considered benign or probably benign at the time of the study in which a breast cancer was diagnosed at later stages for accurate assessment of the false-negative mammograms.
Our follow-up covers a mean of 30 months (range, 17–41 months); with the help of the PALGA, we were able to detect five malignancies diagnosed and treated elsewhere in the Netherlands that otherwise would have been lost to follow-up. It is inevitable that a meticulous review of false-negative mammograms and an extended follow-up results in a decrease in sensitivity and does so to an even greater extent among younger women (16).
Most population-based studies have been obtained from breast screening centers, that is, in asymptomatic women, and in these programs, double reading helps increase cancer detection (17,18). Most diagnostic studies are based on patient series or compilations of abnormalities, and the results are frequently determined retrospectively (19). In studies (20,21) in symptomatic women, sensitivity varies between 76% and 93%. Our study population consisted of 4,728 consecutive patients 30 years or older with a wide variety of indications; both symptomatic and asymptomatic patients were included and are thus not fully comparable. But, if the length of our follow-up and the composition of our study population are taken into account, the sensitivity, specificity, and PPVs and NPVs for our study population are within acceptable limits.
US Examinations
As in most breast imaging centers, we performed US as an adjunct to mammography with full knowledge of the clinical and mammographic findings. The categories of indications as applied in our study population proved to be feasible. We refrained from screening US in agreement with the views of Jackson (3) and Hall (22), and we did not perform color Doppler US. Although Doppler techniques can support the diagnosis of malignancy (23), neither hospital had such a machine readily available inside the breast imaging section. However, US was used frequently to guide fine-needle aspiration biopsy, core-needle biopsy, and localization procedures.
We did not restrict our conclusions to differentiation between cystic and solid lesions but prospectively classified the findings according to five levels of suspicion. However, while our system was not as refined as the system of Stavros et al (1), we did not restrict ourselves to breast nodules; indeterminate findings were also included, with the exception of microcalcifications without associated mass, because they are so rarely demonstrable by using US (24).
Subsequent US was performed in 23% of all cases. The distribution of the indications among the different age groups was fairly uniform, although high-density breast tissue was found predominantly in the younger age groups, as shown in Figures 2 and 3. On the other hand, a majority of all mammograms were considered to show low or medium density, which did not hamper assessment.
There are studies (16,25,26) with findings that have shown that more cancers will be missed at mammography among younger women; findings of other studies (21,27) do not demonstrate a trend in diagnostic accuracy across the age spectrum. Our study findings confirm the influence of age on the sensitivity of mammography. For the groups of patients aged 30–39 years and 40–49 years, the sensitivity of mammography was 68% and 78%, respectively. US resulted in an increase in sensitivity, so that these values were ultimately comparable to the sensitivity and specificity for older patients (Table 5).
US depicted 25 breast cancers in addition to those seen at mammography. For 12 patients, the lesion was not recognized on the mammograms; even at review, the cancer was occult on eight mammograms. In the other 13 cases, US improved breast cancer detection with more accurate identification and characterization of the lesion than at mammography.
It has to be acknowledged that the accuracy of mammography has its repercussion on the role of US. From Table 5, it is clear that the higher the sensitivity of mammography the smaller the benefit of US. The PPV for probably benign lesions at mammography in this study was 5%, which is higher than that in other studies (5,28). It reflects a higher threshold for biopsy than is generally pursued in the United States. This less aggressive attitude leaves room for US to adjust the level of suspicion and increase sensitivity but at the same time keep false-positive rates low.
Evaluation of patients with a false-negative US result revealed that the size of the breast cancer is not a restriction but that interpretation of US scans is difficult when it spreads without forming a discrete mass. Causes of error are unawareness of locally increased acoustic attenuation, for instance, because of lack of overview of the breast anatomy and the interpretation of hypoechoic, small, and diffusely scattered malignant nodules as ductal ectasia (1). Invasive lobular carcinoma is known to resemble breast tissue (29). It was diagnosed at US in one case (Fig 9) but was missed in two cases.
|
|
|
In addition to the controversy about whether US should be used to help differentiate a benign from a malignant lesion, the fear of overuse and high false-positive rates is often mentioned in the literature, while a negative US result does not provide reliable reassurance that an actual mass does not exist (4,22). All participating radiologists were aware of these caveats, and our policy was not to perform US of an entire breast. US was restricted to the area of the indication. As a consequence, we did not diagnose unsuspected mammographically occult cancers, as described by others (31,32).
In 112 cases, the mammographic result was false-positive. This finding was converted to benign or probably benign after US in 45 cases. There was never a case in which incorrect conversion took place, because whenever the mammographic findings could not be correlated with the US findings, the mammographic results prevailed.
In 23 cases with true-negative mammograms, US resulted in a false-positive finding because of the presence of a complex cyst, an atypical manifestation of a fibroadenoma, or a US image that aroused concern because of the inhomogeneous aspect of the glandular tissue. If the size of our cohort is taken into account, this number is not alarming. With respect to the diagnosis of simple cysts, the study findings confirmed the positive reputation of US (30,33): No false-negative results were encountered.
In this study, US as an adjunct to mammography yielded a statistically significant improvement in breast cancer detection. With respect to the total cohort of patients, the improvement may not be very dramatic, but for the subgroup of patients in whom US was performed, improvement was considerable, especially among the younger patients, for whom the sensitivity of mammography was low. The extent of the role of US was also influenced by the mammographic work-up: Mammographic classification based on a relatively high threshold for biopsy allowed US more room for increasing sensitivity. If further refinements in US diagnosis, especially with respect to the US characteristics of diffusely growing cancers, can be achieved, a further increase in the contribution of US to the diagnosis of breast cancer can be expected.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: Az = area under the ROC curve NPV = negative predictive value PALGA = Dutch National Archives of Pathology PPV = positive predictive value ROC = receiver operating characteristic
Author contributions: Guarantor of integrity of entire study, H.M.Z.; study concepts and design, H.M.Z., E.G.C., J.H., A.E.v.V.; definition of intellectual content, H.M.Z., E.G.C.; literature research, H.M.Z.; clinical studies, H.M.Z., E.G.C.; data analysis, J.H.; statistical analysis, J.H.; manuscript preparation and editing, H.M.Z., E.C.G., J.H., M.J.v.d.V.; manuscript review, A.E.v.V.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  E.-K. Kim, K. H. Ko, K. K. Oh, J. Y. Kwak, J. K. You, M. J. Kim, and B.-W. Park Clinical Application of the BI-RADS Final Assessment to Breast Sonography in Conjunction with Mammography Am. J. Roentgenol., May 1, 2008; 190(5): 1209 - 1215. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. J. Shin, H. H. Kim, S. M. Kim, G. Y. Kwon, G. Gong, and O. K. Cho Screening-Detected and Symptomatic Ductal Carcinoma in Situ: Differences in the Sonographic and Pathologic Features Am. J. Roentgenol., February 1, 2008; 190(2): 516 - 525. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Forsberg, C. W. Piccoli, D. A. Merton, J. J. Palazzo, and A. L. Hall Breast Lesions: Imaging with Contrast-enhanced Subharmonic US Initial Experience Radiology, September 1, 2007; 244(3): 718 - 726. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. Jesneck, J. Y. Lo, and J. A. Baker Breast Mass Lesions: Computer-aided Diagnosis Models with Mammographic and Sonographic Descriptors Radiology, August 1, 2007; 244(2): 390 - 398. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. B. Mainiero, A. Goldkamp, E. Lazarus, L. Livingston, S. L. Koelliker, B. Schepps, and W. W. Mayo-Smith Characterization of Breast Masses With Sonography: Can Biopsy of Some Solid Masses Be Deferred? J. Ultrasound Med., February 1, 2005; 24(2): 161 - 167. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Ozdemir, K. Kilic, H. Ozdemir, C. Yucel, S. Andac, and M. Colak Contrast-Enhanced Power Doppler Sonography in Breast Lesions: Effect on Differential Diagnosis After Mammography and Gray Scale Sonography J. Ultrasound Med., February 1, 2004; 23(2): 183 - 195. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. B. Kopans Sonography Should Not Be Used for Breast Cancer Screening Until Its Efficacy Has Been Proven Scientifically Am. J. Roentgenol., February 1, 2004; 182(2): 489 - 491. [Full Text] [PDF]
|
|
|
|
|
|
W. T. Yang and G. M. K. Tse Sonographic, Mammographic, and Histopathologic Correlation of Symptomatic Ductal Carcinoma In Situ Am. J. Roentgenol., January 1, 2004; 182(1): 101 - 110. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Leconte, C. Feger, C. Galant, M. Berliere, B. V. Berg, W. D'Hoore, and B. Maldague Mammography and Subsequent Whole-Breast Sonography of Nonpalpable Breast Cancers: The Importance of Radiologic Breast Density Am. J. Roentgenol., June 1, 2003; 180(6): 1675 - 1679. [Abstract] [Full Text] [PDF]
|
|
|
