Breast Cancer Screening Performance Research Articles

Parameter optimization of a computer-aided diagnosis system for detection of masses on digitized mammograms.

Reading mammograms is a difficult task and for this reason any development that may improve the performance in breast cancer screening is of great importance. We proposed optimized computer aided diagnosis (CAD) system, equipped with reliability estimate module, for mass detection on digitized mammograms. Proposed CAD system consists of four major steps: preprocessing, segmentation, feature extraction and classification. We propose a simple regression function as a threshold function for extraction of potential masses. By running optimization procedure we estimate parameters of the preprocessing and segmentation steps thus ensuring maximum mass detection sensitivity. In addition to the classification, where we tested seven different classifiers, the CAD system is equipped with reliability estimate module. By performing segmentation 91.3% of masses were correctly segmented with 4.14 false positives per image (FPpi). This result is improved in the classification phase where, among the seven tested classifiers, multilayer perceptron neural network achieved the best result including 77.4% sensitivity and 0.49 FPpi. By using the proposed regression function and parameter optimization we were able to improve segmentation results comparing to the literature. In addition, we showed that CAD system has high potential for being equipped with reliability estimate module.

Is clinical breast examination effective in Japan? Consideration from the age-specific performance of breast cancer screening combining mammography with clinical breast examination.

There is controversy about the value of clinical breast examination (CBE) in breast cancer screening programs that include mammography. In Fukui Prefecture, a screening combining mammography with CBE was employed on 62,447 women from 2004 to 2009. We examined the sensitivity and specificity of mammography alone, and mammography and CBE together for each age group (40-49, 50-59, 60-69, and 70-79). 167 breast cancers and 49 false-negative cancers were detected during 5 years. For the combined screening, the sensitivities were 73.1, 74.1, 78.3, and 86.5 %, and the specificities were 83.8, 87.5, 89.8, and 90.9 % in the groups of 40-49, 50-59, 60-69, and 70-79 years, respectively. In the mammography-specific analysis, sensitivity decreased to 69.8 % (-3.3 %), 66.7 % (-7.7 %), 77.3 % (-1.0 %), and 83.8 % (-2.7 %) in the groups of 40-49, 50-59, 60-69, and 70-79 years, respectively. There were greater reductions in the groups of 40-49 and 50-59 years than in those of 60-69 and 70-79 years, but there was no statistically significant decrease. Specificity generally increased with increasing age and there was a significant improvement in specificity among all age groups, except that of 70-79 years. Our findings suggest that there is a trade-off between sensitivity and specificity associated with CBE added to mammography. This tendency is greater in those 40-50 years of age than in those 60-70 years of age. We consider that CBE may be omitted from breast cancer screening among women aged 60 and 70 years. Furthermore, another modality to complement mammography screening in younger Japanese women is expected.

Technical and clinical breast cancer screening performance indicators for computed radiography versus direct digital radiography

To compare technical and clinical screening performance parameters between computed radiography (CR) and direct digital radiography (DR) systems. The number of women screened with CR was 73,008 and with DR 116,945. Technical and patient dose survey data of 25 CR and 37 DR systems were available. Technical performance was expressed by threshold thickness values at the mean glandular dose (MGD) level of routine practice. Clinical indicators included recall rate (RR), cancer detection rate (CDR), percentage of ductal carcinoma in situ (DCIS), percentage of cancers with T-scores smaller than 1 cm and positive predictive value (PPV). Contrast threshold values for the 0.1-mm gold disk were 1.44 μm (SD 0.13 μm) for CR and 1.20 μm (SD 0.13 μm for DR). MGD was 2.16 mGy (SD 0.36 mGy) and 1.35 mGy (SD 0.32 mGy) for CR and DR respectively. We obtained for CR, respectively DR, the following results: RR in the first round of 5.48 % versus 5.61 %; RR in subsequent rounds of 2.52 % versus 2.65 %; CDR of 0.52 % versus 0.53 %; DCIS of 0.08 % versus 0.11 %; a rate of cancers with T-scores smaller than 1 cm of 0.11 % versus 0.11 %; PPV of 18.45 % versus 18.64 %; none of them was significantly different. Our screening indicators are reassuring for the use of CR and DR, with CR operating at 60 % higher MGD. • Breast cancer screening can employ both computed (CR) and direct digital radiography (DR). • Screening performance parameters for CR and DR technology are not significantly different. • Screening parameters are in accordance with European Guidelines. • Radiation doses employed for CR are generally 60 % greater than for DR.

Adherence to recommended breast cancer screening in Iranian turkmen women: the role of knowledge and beliefs.

The aim of the current study was to investigate breast cancer screening performance among Iranian Turkmen women along with their knowledge and beliefs. A cross-sectional study was carried out in June to December 2011. Through clustered sampling method, 1080 Iranian Turkmen women completed the questionnaire including breast cancer screening adherence, knowledge, fatalism beliefs, and perceived threat using Champions Health Belief Model Scale (CHBMS).The mean age of the participants was 43.04 (SD = 11.80) years. Compliance rate in a regular basis based on national guidelines was 13.1%, 2.5%, and 0.9% for SBE, CBE, and mammography, respectively. A mere 4% have been provided adequately with information about breast cancer. Having knowledge was the best predictor of breast cancer screening adherence along with high educated husbands for SBE performing. Susceptibility and fatalism were low and were influenced by participants' educational level and age. In conclusion, Iranian Turkmen women had insufficient knowledge, low perceived susceptibility, high fatalistic belief, and very poor adherence to breast cancer screening. There is a need for providing breast cancer education programs among the Iranian Turkmen women to increase their adherence rate.

Mammographic Image Denoising and Enhancement Using the Anscombe Transformation, Adaptive Wiener Filtering, and the Modulation Transfer Function

A new restoration methodology is proposed to enhance mammographic images through the improvement of contrast features and the simultaneous suppression of noise. Denoising is performed in the first step using the Anscombe transformation to convert the signal-dependent quantum noise into an approximately signal-independent Gaussian additive noise. In the Anscombe domain, noise is filtered through an adaptive Wiener filter, whose parameters are obtained by considering local image statistics. In the second step, a filter based on the modulation transfer function of the imaging system in the whole radiation field is applied for image enhancement. This methodology can be used as a preprocessing module for computer-aided detection (CAD) systems to improve the performance of breast cancer screening. A preliminary assessment of the restoration algorithm was performed using synthetic images with different levels of quantum noise. Afterward, we evaluated the effect of the preprocessing on the performance of a previously developed CAD system for clustered microcalcification detection in mammographic images. The results from the synthetic images showed an increase of up to 11.5 dB (p = 0.002) in the peak signal-to-noise ratio. Moreover, the mean structural similarity index increased up to 8.3 % (p < 0.001). Regarding CAD performance, the results suggested that the preprocessing increased the detectability of microcalcifications in mammographic images without increasing the false-positive rates. Receiver operating characteristic analysis revealed an average increase of 14.1 % (p = 0.01) in overall CAD performance when restored image sets were used.

E4. Current issues in breast cancer screening

E4. Current issues in breast cancer screening

Automated Whole-Breast Ultrasound: Advancing the Performance of Breast Cancer Screening

Screening has been successful in the early detection of some cancers, including cervical, colon, and breast. However, the success in mortality reduction by screening mammography has been limited in women with mammographically dense tissue. Magnetic resonance imaging has been used with success in the screening of high-risk women, but it is expensive. Also, its use in a screening program requires a contrast medium that has not been tested in humans or animals for long-term safety in humans or animals for repeated biannual or annual injections. Ultrasound also has the potential to be an ideal screening tool because it is relatively inexpensive and requires no injected contrast or ionizing radiation. However, the relatively poor conspicuity of some cancers by hand scanning and the considerable radiologist time necessary limit its use. Automated whole-breast ultrasound (AWBU) allows the radiologist to read the images quickly, at a convenient time, while being free from doing the scan. Two-dimensional AWBU uses a cine loop of axial images, with <1 mm image spacing, which are read on a high-resolution monitor to improve the conspicuity of small cancers. A blinded study of this system combined with screening mammography showed that adding AWBU both doubles overall cancer detection and triples the 1 cm-or-less invasive cancers found in dense-breasted women. As expected, mammography had a significantly greater detection rate of ductal carcinoma in situ than AWBU. As yet no clinical studies of other AWBU systems have been published.

Incremental impact of breast cancer SNP panel on risk classification and screening.

1522 Background: Accurate breast cancer risk prediction could improve the performance of breast cancer screening. However, current models are limited, particularly among minorities. Multiple single nucleotide polymorphisms (SNPs) associated with breast cancer have been validated in multiple studies. While each SNP confers a relatively small increase in risk, a multiplicative allelic risk model generates relative risk estimates ranging from 0.45 to 3.77, which appear independent of other risk factors. Methods: Using a prospective cohort study, standard risk information and buccal swab specimen are being collected at the point of screening mammography. A 12 SNP panel was performed by deCODE Genetics. Minor allele frequencies were compared to reported frequencies and the total genetic risk ratio was calculated using reported risk ratios assuming independence. 5 year and lifetime risks incorporating SNPs were calculated by multiplying estimated Gail risk by genotype risk ratio. Concordance between the Gail model and the Gail+SNP model in identifying high risk women (5 year risk > 1.7, lifetime risk > 20) was measured using the kappa statistic. Results: SNP results were available for 601 women (39% AA, 61% W). Mean genetic risk ratios were 1.07 in W and 1.31 in AA women. High risk allele frequencies for W women were the same as previously reported for 8 SNPs and higher than previously reported for 4 SNPs. For AA women, high risk allele frequencies were higher in 8 SNPs, lower in 3 SNPs, and the same for 1 SNP. There was relatively low agreement between the Gail model and Gail +SNP model (kappa of 0.53 for 5-year risk and 0.435 for lifetime risk). Addition of SNP information had the greatest effect among AA women, with 12% of AA women identified as 5 year high risk by Gail alone, but 35% by Gail +SNPs, and 0.4% of AA women identified as lifetime high risk by Gail alone but 7% by Gail + SNPs. Conclusions: The addition of SNP information to traditional breast cancer risk prediction models reclassifies the high risk status of some women undergoing screening mammography, particularly among AA women. Further research is needed to determine the clinical validity and utility of this reclassification in breast cancer risk prediction, screening and prevention.

Trends in Quality of Care and Barriers to Improvement in the Indian Health Service

Although Native Americans experience substantial disparities in health outcomes, little information is available regarding healthcare delivery for this population. To analyze trends in ambulatory quality of care and physician reports of barriers to quality improvement within the Indian Health Service (IHS). Longitudinal analysis of clinical performance from 2002 to 2006 within the IHS, and a physician survey in 2007. Adult patients cared for within the IHS and 740 federally employed physicians within the IHS. Clinical performance for 12 measures of ambulatory care within the IHS; as well as physician reports of ability to access needed health services and use of quality improvement strategies. We examined the correlation between physician reports of access to mammography and clinical performance of breast cancer screening. A similar correlation was analyzed for diabetic retinopathy screening. Clinical performance significantly improved for 10 of the 12 measures from 2002 to 2006, including adult immunizations, cholesterol testing, and measures of blood pressure and cholesterol control for diabetes and cardiovascular disease. Breast cancer screening rates decreased (44% to 40%, p = 0.002), while screening rates for diabetic retinopathy remained constant (51%). Fewer than half of responding primary care physicians reported adequate access to high-quality specialists (29%), non-emergency hospital admission (37%), high-quality imaging services (32%), and high-quality outpatient mental health services (16%). Breast cancer screening rates were higher at sites with higher rates of physicians reporting routine access to mammography compared to sites with lower rates of physicians reporting such access (46% vs. 35%, ρ = 0.27, p = 0.04). Most physicians reported using patient registries and decision support tools to improve patient care. Quality of care has improved within the IHS for many services, however performance in specific areas may be limited by access to essential resources.

Mammography Use and Its Demographic Correlates Among Women in South Korea

This study aimed to investigate the mammography rate and its demographic correlates among Korean women in order to provide basic understanding of factors related to the performance of mammography as a method of breast cancer screening. A descriptive secondary analysis was conducted using a national data set from the Korean National Health and Nutrition Examination Survey (KNHNES III) in South Korea. A total number of 2,602 women over 40 years of age were chosen for the analysis. Main research variables included the use of mammography, monthly income, residential area, age, marriage, education, insurance, smoking habits, drinking habits, job type, current health status, and other cancer screening results. Only 30.59% of women adhered to the national guidelines. Monthly income, age, education, insurance, and smoking habits had statistically significant effects on breast cancer screening performance. The demographic correlates of the performance of mammography highlight the fact that not all Korean women, at this stage, are in a socioeconomic or societal position to undergo biannual mammography screening. Targeted and tailored nursing intervention should be implemented in middle- and senior-aged women who lack the ability to access medical resources in South Korea. This will make it possible for marginalized women to utilize cancer screening tests and in turn promote their health.

Receiver Operating Characteristic Analysis: A Proper Measurement for Performance in Breast Cancer Screening?

Receiver Operating Characteristic Analysis: A Proper Measurement for Performance in Breast Cancer Screening?

Interdisciplinary Breast Care: Essential Information for the Treatment Team

In an effort to provide women with efficient detection, diagnosis, and treatment of breast cancer, increasing numbers of institutions are adopting a comprehensive, interdisciplinary approach to breast care. This trend, along with technological advances in breast imaging, has resulted in an ever-expanding role for radiologists as clinicians and consultants with direct patient interaction. Therefore, clinical breast radiologists have emerged as pivotal in the leadership of and support for interdisciplinary breast care centers; understanding the importance of these centers is optimal to patient outcome and medical cost containment. This article delineates the essential elements to be included in a comprehensive diagnostic evaluation and, in the report of that evaluation, elements that further aid treating physicians in management of patients with breast cancer. In addition, 3 years of breast cancer screening performance data from a comprehensive breast care center are presented and compared with national benchmarks.

An Evaluation of Standardized Patients in Improving Clinical Breast Examinations for Military Women

There is a need to improve the quality of clinical breast examination (CBE) and breast cancer screening for women. The purpose of this study was to determine whether instructions from a standardized patient to military healthcare providers would increase the quality of CBE and breast cancer screening for military women. The study used a 2-group pretest and posttest experimental design with random assignment by study site. Before and after the intervention, the providers completed a 13-item survey to assess their current breast cancer screening practices and the standardized patient used an investigator-developed checklist to assess the providers' breast cancer screening performance. The survey of breast cancer screening practice scores and the interview and CBE performance scores were analyzed using analysis of covariance with the pretest scores as covariates. Results showed the experimental group made significantly more improvement than did the control group in their total scores on the observational checklist of interview and CBE skills (F = 19.18, P < .001, observed power = 0.99). In conclusion, this method of continuing education was effective with military healthcare providers.

Effect of mammographic breast density on breast cancer screening performance: a study in Nijmegen, The Netherlands.

STUDY OBJECTIVE: To study the implications of breast density on mammographic screening performance. DESIGN: Screening outcomes of women with dense breast patterns were compared with those of women with lucent...

Long term breast cancer screening in Nijmegen, The Netherlands: the nine rounds from 1975-92.

To assess the performance of breast cancer screening in different age categories over two decades. Important determinants of reduced breast cancer mortality such as attendance, mammography performance, cancer detection, and disease stage were recorded. Nijmegen, The Netherlands, 1975-92. Since 1975 more than 40,000 women aged 35 years and older have been invited biennially for breast screening in a population based project in Nijmegen. Rates of attendance, referral, detection, and disease stage were calculated, as well as the specificity of screening mammography and the predictive value of referral and biopsy. From round 3 onwards, the attendance rate of women younger than 50 years stabilised at 70%, in women of 50-69 years it was 62%, and in women aged 70 and over it was 22%. In these three age categories, the referral rates of a positive screening mammography per 1000 screened women were 4.9, 6.2, and 11.8, respectively. Specificity rates were between 99% and 100%. Current predictive values of referral were high: in the specific age categories 39%, 59%, and 68% of the referred women had cancer. Detection rates remained fairly stable over the rounds 4-9, at 1.9, 3.6, and 8.0 cancers per 1000 screened women. In the two year period between screening the numbers of interval cancers per 1000 screened women were 2.2, 2.2, and 2.9, for the three age categories respectively. With regard to invasive cancers detected during screening, the percentage of small tumours (< or = 20 mm on the mammogram) was 84% in each age category. For women younger than 50 years, the proportion of intraductal carcinoma in all the cancers detected at screening was 40%, while it was 15% in the other age categories. Throughout the nine rounds, the screening outcomes were found to be adequate, particularly considering the high specificity rate and the predictive value of referral without the interference of a low detection rate. Although the occurrence of interval cancers seemed high, it was similar to other screening programmes. Despite a relatively low referral rate, the ratios of screen detected versus interval cancer cases were favourable. Well organised screening programmes can achieve good mammography results without too many false positives. It is important that women continue to participate in a screening programme because cancer can still be detected even after several successive negative screening examinations.

PP-4-1 Influence of mammographic patterns on breast cancer screening performance

PP-4-1 Influence of mammographic patterns on breast cancer screening performance

Breast Cancer Screening: Effect of Physician Specialty, Practice Setting, Year of Medical School Graduation, and Sex

We surveyed physicians of different specialties in a large metropolitan area to determine how their characteristics affected their performance and beliefs about breast cancer screening. Of 664 general internists, obstetrician-gynecologists, and cardiologists surveyed, we received 298 responses (45%). We found significant differences in reported performance of breast cancer screening and physicians' beliefs about mammography screening among practicing obstetrician-gynecologists, internists, and cardiopulmonary specialists. Cardiopulmonary specialists performed the fewest breast examinations and screening mammograms and were most likely to believe annual mammography screening unnecessary even for women in their 50s. We observed no difference between physicians graduating before 1960 and those graduating afterward and no differences according to physician sex. We found similar screening practices and beliefs in the three types of practice settings examined: community-based, private practices, a large health maintenance organization (HMO), and academic medical centers. Obstetrician-gynecologists and internists differed only in the frequency with which they performed breast examinations. Physicians graduating before 1960 in these two groups reported somewhat poorer performance and knowledge of breast cancer screening than those graduating more recently. A majority of all respondents disagreed with American Cancer Society guidelines for mammography screening. Physicians of all specialties reported performing far more breast examinations than screening mammograms on women of all ages, even for those 50-59 years of age. We conclude that all physicians need to improve their screening rates. However, intervention programs should first target those physicians with the greatest deficiencies in breast cancer screening performance and knowledge; these include medical specialists and older physicians in primary care specialties.

National quality assessment of film reading performance in breast cancer screening

National quality assessment of film reading performance in breast cancer screening

Analysis of the diagnostic performance in breast cancer screening by relative operating characteristics.

In medical usage, relative operating characteristics (ROC) refer to the symptoms, signs, and laboratory values used for making a diagnosis. Diagnosis analysis by ROC measures diagnosis performance independently of the diagnostic criteria used, thus revealing the arbitrary character of diagnosis. In order to carry out ROC analysis, the diagnosis has to be graded in levels of certainty of the presence of disease (or suspicion of a lesion). When both the graded diagnosis decisions and the ultimate proof of presence or absence of disease are available, ROC analysis can easily be carried out. Analysis of 50,102 successive examinations of 19,400 women in the Utrecht breast cancer screening projects presents a remarkably good fit of the observed data with the calculated values. In the medical context even small deviations of the actual sensitivity and specificity values from the ROC curve values must be considered as signs of differences in the diagnosis procedure and should lead to further analysis of the diagnostic criteria and the examined population. Consequently the separate effect of diagnosis impairment by the mammographic risk patterns (Wolfe) on diagnostic accuracy could be isolated. The effect of age on the detection capability of breast cancer by mammography appeared to operate in narrow age groups.