ATMTN: a telemammography network architecture

One of the goals of the National Cancer Institute (NCI) to reach more than 80% of eligible women in mammography screening by the year 2000 yet remains as a challenge. In fact, a recent medical report reveals that while other types of cancer are experiencing negative growth, breast cancer has been the only one with a positive growth rate over the last few years. This is primarily due to the fact that 1) examination process is a complex and lengthy one and 2) it is not available to the majority of women who live in remote sites. Currently for mammography screening, women have to go to doctors or cancer centers/hospitals annually while high-risk patients may have to visit more often. One way to resolve these problems is by the use of advanced networking technologies and signal processing algorithms. On one hand, software modules can help detect, with high precision, true negatives (TN), while marking true positives (TP) for further investigation. Unavoidably, in this process some false negatives (FN) will be generated that are potentially life threatening; however, inclusion of the detection software improves the TP detection and, hence, reduces FNs drastically. Since TNs are the majority of examinations on a randomly selected population, this first step reduces the load on radiologists by a tremendous amount. On the other hand, high-speed networking equipment can accelerate the required clinic-lab connection and make detection, segmentation, and image enhancement algorithms readily available to the radiologists. This will bring the breast cancer care, caregiver, and the facilities to the patients and expand diagnostics and treatment to the remote sites. This research describes asynchronous transfer mode telemammography network (ATMTN) architecture for real-time, online screening, detection and diagnosis of breast cancer. ATMTN is a unique high-speed network integrated with automatic robust computer-assisted diagnosis-detection/digital signal processing (CAD/DSP) methods for mass detection, region of interest (ROI) compression algorithms using Digital Imaging and Communications in Medicine (DICOM) 3.0 medical image standard. While ATMTN has the advantage of higher penetration for cancer screening, it provides the diagnosis with higher efficiency, better accuracy and potentially lower cost. This paper presents the development of the infrastructure and algorithm design for ATMTN-based telemammography. The research goals involved: 1) networking stations for telemammography to demonstrate, evaluate, and validate technologies and methods for delivering mammography screening services via high-speed (155 MB/s) links, performing real-time network-transmitted, high-resolution mammograms for immediate diagnosis as a "second opinion" strategy; 2) development of object-oriented compression methods for storage, retrieval and transmission of mammograms; 3) inclusion and optimization of detection algorithms for identification of normal images in different resolutions to increase the speed and effectiveness of telemammography as a "second opinion" strategy; 4) resolving the compatibility issues between images from different equipment (DICOM standards); and 5) optimization of an integrated ATMTN with adaptive CAD/DSP methods that are robust for large image databases and input sources.

Should breast screening programmes limit their detection of ductal carcinoma in situ?

OBJECTIVE

Previous research has shown that the detection of ductal carcinoma in situ (DCIS) aids the detection of small invasive cancers at mammographic screening. A correlation may therefore exist between a screening unit's DCIS detection rate and their small invasive cancer detection rate. We have therefore investigated the effect of DCIS detection rate on the detection of small (<15 mm) invasive cancers in the 95 units of the U.K. NHS Breast Screening Programme (NHSBSP).

MATERIALS AND METHODS

DCIS detection rates were examined against large (> or =15 mm) and small (<15 mm) invasive cancer detection rates in women aged 50-64 years at prevalent and incident screens over a 3-year period.

RESULTS

After adjusting for background incidence, screening units with the highest DCIS detection of > or =1.3/1000 detected over 20% more small invasive cancers than units with DCIS detection rates within the NHSBSP guidelines of 0.5-1/1000 (P<0.001). Sixty percent of units had DCIS detection rates above the guidelines. There was no correlation between DCIS detection and > or =15 mm invasive cancer detection. The results suggest that over the range of DCIS rates studied, that for every two extra DCIS cancers detected, an additional small invasive cancer (<15 mm) is detected that may otherwise not have been. The results therefore provide supporting evidence that the detection of DCIS aids the detection of small invasive cancers.

CONCLUSION

Units with DCIS detection above the NHSBSP guidelines have significantly better small invasive cancer detection rates. The existence of an upper limit for DCIS detection within the NHSBSP may be preventing the detection of small invasive cancers, because units are not recalling some small clusters of calcification in order to keep DCIS detection rates down. The upper limit may therefore be inappropriate.

An improved method of region grouping for microcalcification detection in digital mammograms

A very important issue, namely region grouping, in computer-assisted diagnostic detection of microcalcification clusters (MCC) in digital mammograms is addressed in this work. In the diagnosis of breast cancer, MCC, instead of single and isolated microcalcifications, are considered clinically significant. Grouping individual regions segmented from digital mammograms, therefore, should be a component in an automatic MCC detection system. Actually this component may concern several system modules, such as segmentation, feature extraction, performance estimation aiming at both algorithm optimization and consistent evaluation and ultimately computerized malignancy estimation of calcified lesions. The previous work in the literature used a kernel-based method for region grouping. We proposed a distance-based and dense-to-sparse grouping method. The grouping result should be independent of the size, shape and orientation of real clusters. The application, namely cluster-oriented analysis including an adaptive segmentation method and cluster level feature extraction scheme, is discussed. A preliminary study was performed on a set of 30 full mammograms at 60 microm resolution, containing 40 MCC. The introduction of the cluster level feature extraction and a simple rule-based method reduces false positives from 7.1 to 2.4 per image at the sensitivity of 92.5%. This grouping method provides a solid basis for effective feature extraction-analysis and candidate cluster classification.

The effect of spectra on calibration and measurement with mammographic ionization chambers

Mammographic imaging uses x-ray tubes with molybdenum, rhodium, or tungsten anodes with the produced bremsstrahlung filtered by thin sheets of molybdenum, rhodium, or aluminum. The National Institute of Standards and Technology, the Accredited Dosimetry Calibration Laboratories, and several manufacturers offer calibrations of mammography ionization chambers with reference x-ray beams with different radiation qualities in the range 23-40 kVp. The energy response of ten commercially available chambers was determined for these reference radiation qualities using the Attix variable-length free-air chamber. The evaluated chambers are designed with thin entrance windows of varying thickness and composition. The chambers show variation in their air kerma response as a function of beam radiation quality. This response with beam radiation quality may affect the measurement of clinical beam half value layer (HVL) and the determination of the mean glandular dose. The combined effect of the chamber's energy dependence and HVL measurement affects the mean glandular dose calculation resulting in differences ranging from -1.8% to +2.5%.

Will screening mammography in the East do more harm than good?

OBJECTIVES

We sought to systematically review the evidence for population-based mammography as applied to a Chinese population.

METHODS

Primary reports for meta-analysis were identified by a search of MEDLINE and the Cochrane Library. Outcome measures included breast cancer-related mortality, the number needed to be screened to prevent 1 death, and the positive predictive value of mammography.

RESULTS

Pooled relative risk for breast cancer-related death in the screened group was 0.80 (95% confidence interval = 0.71, 0.90). Applied to Hong Kong, this figure translates into a number needed to screen of 1 302 healthy women screened annually for 10 years to prevent 1 death.

CONCLUSIONS

Evidence is insufficient to justify population-based breast cancer screening by mammography for women in Hong Kong and other Asian populations with low breast cancer prevalence.

Efficacy of 3D-MR mammography for breast conserving surgery after neoadjuvant chemotherapy

BACKGROUND

One of the main roles of neoadjuvant chemotherapy for breast cancer is to shrink large tumors to increase patient eligibility for breast conserving surgery. Three dimensional MR Mammography (3D-MRM) can detect tumor extension more accurately compared with mammography and Ultrasonography (US). Therefore, the shrinkage pattern observed on 3D-MRM was analyzed with regard to several pathological factors.

METHODS

A total of 27 breast cancer cases were examined by 3D-MRM before and after neoadjuvant chemotherapy. The volume reduction and shrinkage patterns were assessed and compared with the pathological diagnosis.

RESULTS

There were two shrinkage patterns. Twelve of 25 evaluable breast cancers (48%) showed a concentric shrinkage pattern while 13 cases (52%) showed a dendritic shrinkage pattern. The cases with concentric shrinkage were good candidates for breast conserving surgery, But tumors showing dendritic shrinkage often had positive margins necessitating mastectomy. Pathologically, tumors with a papillotubular pattern, Estrogen receptor (ER) positivity, low nuclear grade and c-erbB 2 negativity tended to show dendritic shrinkage.

CONCLUSIONS

3D-MRM is a useful modality for evaluating whether breast conserving surgery can be safely done in the neoadjuvant setting.

[Experiences with phantom measurements in different mammographic systems]

PURPOSE

Determination of image quality between conventional film screen system, digital phosphor storage plate mammography (high resolution) and digital mammography.

MATERIALS AND METHODS

Mammograms of the Wisconsin Mammographic Random Phantom, Model 152 A (Radiation Measurements Inc., Wisconsin) were acquired using a conventional film-screen system, a digital storage phosphor plate system and a digital system.

RESULTS

Of 225 possible details, 191/a 38.2, 193/a 38.6, and 202/a 40.4 details were detected with conventional film-screen system, digital phosphor storage plate mammography and digital mammography, respectively. There was no significant difference (p < 0,058). The entrance surface air kerma was 9.64 mGy, 7.60 mGy and 7.02 mGy, respectively.

CONCLUSIONS

Based on these results, conventional film-screen system can be replaced with both digital phosphor storage plate mammography and digital mammography, to be confirmed with further clinical trials.

Screening mammograms by community radiologists: variability in false-positive rates

BACKGROUND

Previous studies have shown that the agreement among radiologists interpreting a test set of mammograms is relatively low. However, data available from real-world settings are sparse. We studied mammographic examination interpretations by radiologists practicing in a community setting and evaluated whether the variability in false-positive rates could be explained by patient, radiologist, and/or testing characteristics.

METHODS

We used medical records on randomly selected women aged 40-69 years who had had at least one screening mammographic examination in a community setting between January 1, 1985, and June 30, 1993. Twenty-four radiologists interpreted 8734 screening mammograms from 2169 women. Hierarchical logistic regression models were used to examine the impact of patient, radiologist, and testing characteristics. All statistical tests were two-sided.

RESULTS

Radiologists varied widely in mammographic examination interpretations, with a mass noted in 0%-7.9%, calcification in 0%-21.3%, and fibrocystic changes in 1.6%-27.8% of mammograms read. False-positive rates ranged from 2.6% to 15.9%. Younger and more recently trained radiologists had higher false-positive rates. Adjustment for patient, radiologist, and testing characteristics narrowed the range of false-positive rates to 3.5%-7.9%. If a woman went to two randomly selected radiologists, her odds, after adjustment, of having a false-positive reading would be 1.5 times greater for the radiologist at higher risk of a false-positive reading, compared with the radiologist at lowest risk (95% highest posterior density interval [similar to a confidence interval] = 1.17 to 2.08).

CONCLUSION

Community radiologists varied widely in their false-positive rates in screening mammograms; this variability range was reduced by half, but not eliminated, after statistical adjustment for patient, radiologist, and testing characteristics. These characteristics need to be considered when evaluating false-positive rates in community mammographic examination screening.

Development of a protocol for evaluation of mammographic surveillance services in women under 50 with a family history of breast cancer

BACKGROUND

Preliminary retrospective data suggest it is possible to identify impalpable breast cancer in women presenting with a family history of breast cancer under the age of 50, by using regular mammography. In consequence, this service is offered in a number of centres in the UK. The effectiveness of such a service, however, has not been fully evaluated.

METHODS

We propose to perform such an evaluation in a cohort of 20000 women under the age of 50 with a significant family history of breast cancer, given regular mammographic surveillance over 5 years. Comparison of surgical and pathological data with completed and ongoing population screening trials using analysis techniques of varying complexity will be performed to obtain an accurate prediction of future breast-cancer mortality reduction. The formal aims are: i) to estimate the difference in breast-cancer mortality in women under the age of 50 with a significant family history of breast cancer having regular mammography, compared with those not being screened; ii) to estimate the cost-effectiveness of regular mammography in this group of women, compared with no screening. The increase in health service resource use attributable to such a policy will be compared with no screening, and costed. Incremental cost-effectiveness ratios of implementing the standardised mammography strategy compared with no screening will be presented in terms of the additional cost per cancer detected, per life saved and per life-year saved.

The Canadian National Breast Screening Study-1: breast cancer mortality after 11 to 16 years of follow-up. A randomized screening trial of mammography in women age 40 to 49 years

BACKGROUND

The efficacy of breast cancer screening in women age 40 to 49 years remains controversial.

OBJECTIVE

To compare breast cancer mortality in 40- to 49-year-old women who received either 1) screening with annual mammography, breast physical examination, and instruction on breast self-examination on 4 or 5 occasions or 2) community care after a single breast physical examination and instruction on breast self-examination.

DESIGN

Individually randomized, controlled trial.

SETTING

15 Canadian centers.

PARTICIPANTS

50 430 volunteers age 40 to 49 years, recruited from January 1980 to March 1985, who were not pregnant, had no previous breast cancer diagnosis, and had not had mammography in the preceding 12 months.

INTERVENTIONS

Breast physical examination and instruction on breast self-examination preceded random assignment of 25 214 women to receive mammography and annual mammography, breast physical examination, and breast self-examination and 25 216 women to receive usual community care with annual follow-up.

MEASUREMENTS

Verified breast cancer incidence and cohort mortality through 31 December 1993 and deaths from breast cancer through 30 June 1996.

RESULTS

The 105 breast cancer deaths in the mammography group and 108 breast cancer deaths in the usual care group yielded a cumulative rate ratio, adjusted for mammography done outside the study, of 1.06 (95% CI, 0.80 to 1.40). A total of 592 cases of invasive breast cancer and 71 cases of in situ breast cancer were diagnosed by 31 December 1993 in the mammography group compared with 552 and 29 cases, respectively, in the usual care group. The expected proportions of nonpalpable and small invasive tumors were detected on mammography.

CONCLUSION

After 11 to 16 years of follow-up, four or five annual screenings with mammography, breast physical examination, and breast self-examination had not reduced breast cancer mortality compared with usual community care after a single breast physical examination and instruction on breast self-examination. The study data show that true effects of 20% or greater are unlikely.

Interobserver agreement and performance score comparison in quality control using a breast phantom: screen-film mammography vs computed radiography

Our objective was to evaluate interobserver agreement and to compare the performance score in quality control of screen-film mammography and computed radiography (CR) using a breast phantom. Eleven radiologists interpreted a breast phantom image (CIRS model X) by four viewing methods: (a) original screen-film; (b) soft-copy reading of the digitized film image; (c) hard-copy reading of CR using an imaging plate; and (d) soft-copy reading of CR. For the soft-copy reading, a 17-in. CRT monitor (1024x1536x8 bits) was used. The phantom image was evaluated using a scoring system outlined in the instruction manual, and observers judged each object using a three-point rating scale: (a) clearly seen; (b) barely seen; and (c) not seen. For statistical analysis, the kappa statistic was employed. For "mass" depiction, interobserver agreement using CR was significantly lower than when using screen-film ( p<0.05). There was no significant difference in the kappa value for detecting "microcalcification"; however, the performance score of "microcalcification" on CR hard-copy was significantly lower than on the other three viewing methods ( p<0.05). Viewing methods (film or CR, soft-copy or hard-copy) could affect how the phantom image is judged. Paying special attention to viewing conditions is recommended for quality control of CR mammograms.

Mammographic x-ray unit kilovoltage test tool based on k-edge absorption effect

A simple tool to determine the peak kilovoltage (kVp) of a mammographic x-ray unit has been designed. Tool design is based on comparing the effect of k-edge discontinuity of the attenuation coefficient for a series of element filters. Compatibility with the mammography accreditation phantom (MAP) to obtain a single quality control film is a second design objective. When the attenuation of a series of sequential elements is studied simultaneously, differences in the absorption characteristics due to the k-edge discontinuities are more evident. Specifically, when the incident photon energy is higher than the k-edge energy of a number of the elements and lower than the remainder, an inflection may be seen in the resulting attenuation data. The maximum energy of the incident photon spectra may be determined based on this inflection point for a series of element filters. Monte Carlo photon transport analysis was used to estimate the photon transmission probabilities for each of the sequential k-edge filter elements. The photon transmission corresponds directly to optical density recorded on mammographic x-ray film. To observe the inflection, the element filters chosen must have k-edge energies that span a range greater than the expected range of the end point energies to be determined. For the design, incident x-ray spectra ranging from 25 to 40 kVp were assumed to be from a molybdenum target. Over this range, the k-edge energy changes by approximately 1.5 keV between sequential elements. For this design 21 elements spanning an energy range from 20 to 50 keV were chosen. Optimum filter element thicknesses were calculated to maximize attenuation differences at the k-edge while maintaining optical densities between 0.10 and 3.00. Calculated relative transmission data show that the kVp could be determined to within +/-1 kV. To obtain experimental data, a phantom was constructed containing 21 different elements placed in an acrylic holder. MAP images were used to determine appropriate exposure techniques for a series of end point energies from 25 to 35 kVp. The average difference between the kVp determination and the calibrated dial setting was 0.8 and 1.0 kV for a Senographe 600 T and a Senographe DMR, respectively. Since the k-edge absorption energies of the filter materials are well known, independent calibration or a series of calibration curves is not required.

Performance parameters for screening and diagnostic mammography: specialist and general radiologists

PURPOSE

To evaluate performance parameters for radiologists in a practice of breast imaging specialists and general diagnostic radiologists who interpret a large series of consecutive screening and diagnostic mammographic studies.

MATERIALS AND METHODS

Data (ie, patient age; family history of breast cancer; availability of previous mammograms for comparison; and abnormal interpretation, cancer detection, and stage 0-I cancer detection rates) were derived from review of mammographic studies obtained from January 1997 through August 2001. The breast imaging specialists have substantially more initial training in mammography and at least six times more continuing education in mammography, and they interpret 10 times more mammographic studies per year than the general radiologists. Differences between specialist and general radiologist performances at both screening and diagnostic examinations were assessed for significance by using Student t and chi(2) tests.

RESULTS

The study involved 47,798 screening and 13,286 diagnostic mammographic examinations. Abnormal interpretation rates for screening mammography (ie, recall rate) were 4.9% for specialists and 7.1% for generalists (P <.001); and for diagnostic mammography (ie, recommended biopsy rate), 15.8% and 9.9%, respectively (P <.001). Cancer detection rates at screening mammography were 6.0 cancer cases per 1,000 examinations for specialists and 3.4 per 1,000 for generalists (P =.007); and at diagnostic mammography, 59.0 per 1,000 and 36.6 per 1,000, respectively (P <.001). Stage 0-I cancer detection rates at screening mammography were 5.3 cancer cases per 1,000 examinations for specialists and 3.0 per 1,000 for generalists (P =.012); and at diagnostic mammography, 43.9 per 1,000 and 27.0 per 1,000, respectively (P <.001).

CONCLUSION

Specialist radiologists detect more cancers and more early-stage cancers, recommend more biopsies, and have lower recall rates than general radiologists.

Mammography processor quality assurance

After completing this article on mammographic processing, readers will: Understand the relationship between mammography and breast cancer survival. Define quality assurance and quality control. Discuss the history of the Mammography Quality Standards Act. List the important considerations in processor setup. Name the tests in the medical physicist survey. Know how to perform daily processor quality control. Understand the importance of continuous quality improvement.

Does training in the Breast Imaging Reporting and Data System (BI-RADS) improve biopsy recommendations or feature analysis agreement with experienced breast imagers at mammography?

PURPOSE

To determine whether training in the Breast Imaging Reporting and Data System (BI-RADS) improves observer performance and agreement with the consensus of experienced breast imagers with regard to mammographic feature analysis and final assessment.

MATERIALS AND METHODS

A test set of mammograms was developed, with 54 proven lesions consisting of 28 masses (nine [32%] malignancies) and 26 microcalcifications (10 [38%] malignancies). Three experienced breast imagers reviewed cases independently and by means of consensus. Twenty-three practicing mammogram-interpreting physicians reviewed mammograms before and after a day's lectures on BI-RADS. Observer performance before and after training was measured by means of agreement (kappa) with consensus description and assessments, rate of biopsy of malignant and benign lesions, and areas under receiver operating characteristic (ROC) curves. Performance was also measured for 11 participants 2-3 months after training.

RESULTS

Improved agreement with consensus feature analysis was found for mass margins and/or asymmetries, with a pretraining generalized kappa value of 0.36 and a posttraining generalized kappa value of 0.41. Similar improvement was seen for description of calcification morphology (pretraining kappa value of 0.36 improving to 0.44 after training). No improvement was seen in describing calcification distribution. Final assessments were more consistent after training, with a pretraining kappa value of 0.31, as compared with 0.45 after training. The mean biopsy rate for malignant lesions improved from 73% (range, 53%-89%) before training to 88% (range, 74%-100%) after training, with minimal increase in mean biopsy rate of benign lesions (43% [range, 26%-60%] before to 51% [range, 31%-63%] after training), and no net change in area under the ROC curve, as compared with histopathologic findings. For the subset of participants with delayed follow-up, no significant decline in posttraining results was seen.

CONCLUSION

BI-RADS training resulted in improved agreement with the consensus of experienced breast imagers for feature analysis and final assessment. It is important that trainees showed improved rates of recommending biopsy for malignant lesions. This effect was maintained over 2-3 months.

[Stereotactic 11-gauge directional vacuum-assisted breast biopsy: experience with 249 patients]

PURPOSE

To assess the value of percutaneous vacuum-assisted core biopsy to improve the diagnosis of non palpable mammographic abnormalities.

MATERIALS AND METHODS

A total of 252 core biopsies using an 11G Mammotome((R))were performed in 249 patients. Stereotactic localization was performed in the prone position on a dedicated digital Fischer table.

RESULTS

Fifty-one, or 25%, of 200 clusters of microcalcifications corresponded to carcinomas: 126 benign lesions, 23 atypical hyperplasia and LCIS, 31 DCIS, 15 invasive ductal carcinomas, and 4 false negative biopsies. In these 4 last cases, surgery was performed because radiographs of the core biopsy showed no microcalcifications; carcinoma was confirmed at histology of the surgical specimen. Using the BI-RADS system, 7 lesions were category 3, 175 lesions were category 4, and 18 lesions were category 5. From a total of 52 masses, 31 were benign lesions, 2 were borderline lesions, and 19 were invasive carcinomas. From these, 5 lesions were category 3, 31 were category 4, and 16 were category 5. Diagnostic surgical biopsy was avoided in 161 cases (63%), in 152 cases for benign lesions including 151 lesions classified as category 4 lesions and in 9 cases for multifocal or recurrent malignant lesions.

CONCLUSION

When technical pitfalls are avoided and when presence of microcalcifications in the core biopsy sample is verified, vacuum assisted core biopsy with Mammotome((R)) 11G provides accurate diagnosis of non-palpable mammographic abnormalities.

Disparities in adherence to recommended followup on screening mammography: interaction of sociodemographic factors

OBJECTIVE

The objective of this study was to examine disparities in adherence to screening mammography and, specifically, to investigate whether race/ethnicity, education, age, health insurance, and family history of breast cancer (FHBC), as unique factors and in interactions, influence adherence to recommended follow up on screening mammography.

DESIGN

The study involved retrieval and analyses of data collected by the Colorado Mammography Project (CMAP) for 167,232 diverse (82.8% White, 3.4% Black, 11% Hispanic, 1.6% Asian, 0.6% Native American, and 0.6% "other") screening participants during the 1990-1997 study period.

METHODS

Subjects' first mammograms captured by CMAP were tracked in the database to identify women who received follow-up recommendations, women who adhered within 12 months and those that did not. Analyses included comparisons of adherence rates among women with various sociodemographic characteristics.

RESULTS

Of the 17,358 women who received follow-up recommendations, 80.7% adhered. Overall, non-White women in each of the racial/ethnic groups were less likely to adhere to recommendations than were White women (P<.05). Also less likely to adhere were the younger, less educated, uninsured/underinsured, and women who reported not having FHBC.

CONCLUSION

Race/ethnicity appeared to interact with age, education, health insurance, and FHBC to influence the probability of adherence, suggesting the need to explore further cultural, psychosocial, and situational factors.

Detection method and breast carcinoma histology

BACKGROUND

The association between method of detection and breast carcinoma histopathology has not been assessed adequately in a population-based setting.

METHODS

Among women who were included in a population-based, case-control study of breast cancer, patients who were newly diagnosed with invasive breast carcinoma were identified from Wisconsin's statewide tumor registry. Only women age > or = 50 years were analyzed, because screening by mammography was not recommended before age 50 years at the time of the study. The breast tumors among these women (n = 2341 tumors) included the following histopathologies: lobular carcinoma (n = 206 tumors); ductal carcinoma, not otherwise specified (n = 1920 tumors); papillary carcinoma (n = 15 tumors); medullary carcinoma (n = 36 tumors); mucinous adenocarcinoma (n = 56 tumors); tubular adenocarcinoma (n = 41 tumors); invasive comedocarcinoma (n = 24 tumors); scirrhous adenocarcinoma (n = 15 tumors); and mixed ductal/lobular carcinoma (n = 28 tumors).

RESULTS

Overall, women reported that 41% of tumors were detected by mammography, 48% of tumors were self detected, and 11% of tumors were detected by clinical breast examination (CBE). Detection by mammography was significantly more likely for women who had tubular carcinoma (83%; P < 0.001) and invasive comedocarcinoma (67%; P = 0.23) compared with women who had ductal carcinoma (40%). Mammography was significantly less likely to detect medullary carcinoma (17%) than ductal carcinoma (40%; P = 0.01). Lobular carcinoma was the only histopathology that, compared with ductal carcinoma, was detected significantly more often by CBE than by self detection. Mammography detected lobular carcinoma (42%) as frequently as ductal carcinoma (40%). However, the use of postmenopausal hormones may have modified these detection patterns: Among current users, mammography discovered a greater percentage of ductal carcinomas (51%) and fewer lobular carcinomas (36%) than nonusers.

CONCLUSIONS

Among women age > or = 50 years, breast cancer detection by mammography, self detection, and CBE varied according to tumor histopathology.

Use of the American College of Radiology BI-RADS guidelines by community radiologists: concordance of assessments and recommendations assigned to screening mammograms

OBJECTIVE

This study evaluated the use of the American College of Radiology Breast Imaging Reporting and Data System (BI-RADS) by community radiologists by determining the concordance of assessment categories and recommendations assigned to screening mammograms.

MATERIALS AND METHODS

The study comprised the interpretations of 82,620 consecutive screening mammograms by 18 radiologists between January 1, 1995, and December 31, 1998. For all mammograms, assessment categories and recommendations were compared to determine whether they were in accordance with BI-RADS guidelines. Overall patterns of discordance were analyzed, and comparisons of discordant patterns by assessment category, patient age, breast density, and year of examination were made.

RESULTS

The overall discordance between BI-RADS assessments and recommendations was low (3%). The assessment with the highest discordance was "probably benign finding" (category 3), at 53.5%. Mammograms obtained in 1998 were almost half as likely to have assessment-recommendation discordance compared with those obtained in 1995 (2.4% vs 4.5%, respectively; odds ratio = 0.52; p < 0.001). Mammograms of women with dense breast tissue were 30% more likely to have lesions assigned discordant assessments and recommendations compared with those of women with fatty tissue (3.4% vs 2.7%, respectively; odds ratio = 1.3; p < 0.001). No differences in the patterns of discordance were found between mammograms of women younger than 50 years and those of women 50 years old and older (p = 0.10).

CONCLUSION

There has been improvement in the accurate application of BI-RADS since its introduction. However, variation in the pairing of BI-RADS assessments and recommendations persists. Continued efforts to educate radiologists about the use of BI-RADS and to clarify BI-RADS terms would promote maximum consistency in this use of this reporting method.

Survival from primary breast cancer after routine clinical use of mammography

Clinical trials indicate that mammography provides a substantial breast cancer survival benefit; however, there is a need to demonstrate that this benefit extends to clinical practice and to determine the extent that current reductions in mortality are attributable to regular screening or adjuvant systemic therapy. Mammography was used routinely at our institution across a broad age range, in an era when most patients received no adjuvant systemic therapy. We examined breast cancer survival for a cohort of 678 stage I-III primary invasive breast cancer patients accrued from 1971 to 1990, and followed to 1996; 18% received adjuvant hormonal therapy and 15% received adjuvant chemotherapy. There were 61 women less than 40 years old; 136, 40-49 years; 341, 50-69 years; 140, > or =70 years. Factors available for multivariate investigations were age (years), tumor size (cm), nodal status (N-, Nx, N+), ER (fmol/mg protein), PgR (fmol/mg protein), adjuvant radiotherapy (no, yes), adjuvant hormonal therapy (no, yes), and adjuvant chemotherapy (no, yes). Forward stepwise multivariate regression with log-normal survival analysis was used to examine the effects of these factors on disease-specific survival. Ten-year survival by tumor size was adjusted for the effects of other significant factors. For women less than 40 years of age, 10-year survival at the T1a, T1b, T1c, and T2 cut-points for tumor size is, respectively, 0.77, 0.74, 0.67, 0.44; for 40-49 years it is 0.92, 0.90, 0.85, 0.62; for 50-69 years it is 0.81, 0.79, 0.75, 0.62; for > or =70 years it is 0.84, 0.81, 0.73, 0.44. With routine use of clinical mammography and up to 26 years of follow-up, we found breast cancer survival to be significantly better (p< or = 0.05) for all women with smaller tumors and that survival indicated a change in natural disease history with early detection. The Canadian National Breast Screening Study (NBSS) controls had significantly smaller tumors (p < 0.001) than our patients, which may indicate access to mammography outside of the NBSS that reduced the apparent survival benefit for clinical trial mammography.