Radiomic Phenotypes of Mammographic Parenchymal Complexity: Toward Augmenting Breast Density in Breast Cancer Risk Assessment.

Purpose To identify phenotypes of mammographic parenchymal complexity by using radiomic features and to evaluate their associations with breast density and other breast cancer risk factors. Materials and Methods Computerized image analysis was used to quantify breast density and extract parenchymal texture features in a cross-sectional sample of women screened with digital mammography from September 1, 2012, to February 28, 2013 (n = 2029; age range, 35-75 years; mean age, 55.9 years). Unsupervised clustering was applied to identify and reproduce phenotypes of parenchymal complexity in separate training (n = 1339) and test sets (n = 690). Differences across phenotypes by age, body mass index, breast density, and estimated breast cancer risk were assessed by using Fisher exact, χ2, and Kruskal-Wallis tests. Conditional logistic regression was used to evaluate preliminary associations between the detected phenotypes and breast cancer in an independent case-control sample (76 women diagnosed with breast cancer and 158 control participants) matched on age. Results Unsupervised clustering in the screening sample identified four phenotypes with increasing parenchymal complexity that were reproducible between training and test sets (P = .001). Breast density was not strongly correlated with phenotype category (R2 = 0.24 for linear trend). The low- to intermediate-complexity phenotype (prevalence, 390 of 2029 [19%]) had the lowest proportion of dense breasts (eight of 390 [2.1%]), whereas similar proportions were observed across other phenotypes (from 140 of 291 [48.1%] in the high-complexity phenotype to 275 of 511 [53.8%] in the low-complexity phenotype). In the independent case-control sample, phenotypes showed a significant association with breast cancer (P = .001), resulting in higher discriminatory capacity when added to a model with breast density and body mass index (area under the curve, 0.84 vs 0.80; P = .03 for comparison). Conclusion Radiomic phenotypes capture mammographic parenchymal complexity beyond conventional breast density measures and established breast cancer risk factors. © RSNA, 2018 Online supplemental material is available for this article. See also the editorial by Pinker in this issue.

Using Convolutional Neural Networks for Enhanced Capture of Breast Parenchymal Complexity Patterns Associated with Breast Cancer Risk.

RATIONALE AND OBJECTIVES: We evaluate utilizing convolutional neural networks (CNNs) to optimally fuse parenchymal complexity measurements generated by texture analysis into discriminative meta-features relevant for breast cancer risk prediction. MATERIALS AND METHODS: With Institutional Review Board approval and Health Insurance Portability and Accountability Act compliance, we retrospectively analyzed "For Processing" contralateral digital mammograms (GE Healthcare 2000D/DS) from 106 women with unilateral invasive breast cancer and 318 age-matched controls. We coupled established texture features (histogram, co-occurrence, run-length, structural), extracted using a previously validated lattice-based strategy, with a multichannel CNN into a hybrid framework in which a multitude of texture feature maps are reduced to meta-features predicting the case or control status. We evaluated the framework in a randomized split-sample setting, using the area under the curve (AUC) of the receiver operating characteristic (ROC) to assess case-control discriminatory capacity. We also compared the framework to CNNs directly fed with mammographic images, as well as to conventional texture analysis, where texture feature maps are summarized via simple statistical measures that are then used as inputs to a logistic regression model. RESULTS: Strong case-control discriminatory capacity was demonstrated on the basis of the meta-features generated by the hybrid framework (AUC = 0.90), outperforming both CNNs applied directly to raw image data (AUC = 0.63, P <.05) and conventional texture analysis (AUC = 0.79, P <.05). CONCLUSIONS: Our results suggest that informative interactions between patterns exist in texture feature maps derived from mammographic images, which can be extracted and summarized via a multichannel CNN architecture toward leveraging the associations of textural measurements to breast cancer risk.

Aspirin use is associated with lower mammographic density in a large screening cohort.

BACKGROUND: Observational and biologic studies suggest that aspirin is a promising prevention therapy for breast cancer. However, clinical trials to date have not corroborated this evidence, potentially due to study design. We evaluated the effect of aspirin on mammographic density (MD), an established modifiable risk factor for breast cancer. METHODS: Electronic medical records from the University of Pennsylvania were evaluated for women who underwent screening mammography, saw their primary care provider, and had a confirmed list of medications during 2012-2013. Logistic regression was performed to test for associations between clinically recorded MD and aspirin use, after adjusting for age, body mass index (BMI), and ethnicity. RESULTS: We identified 26,000 eligible women. Mean age was 57.3, mean BMI was 28.9 kg/m2, 41% were African American, and 19.7% reported current aspirin use. Aspirin users were significantly older and had higher BMI. There was an independent, inverse association between aspirin use and MD (P trend < 0.001). Women with extremely dense breasts were less likely to be aspirin users than women with scattered fibroglandular density (OR 0.73; 95% CI 0.57-0.93). This association was stronger for younger women (P = 0.0002) and for African Americans (P = 0.011). The likelihood of having dense breasts decreased with aspirin dose (P trend = 0.007), suggesting a dose response. CONCLUSIONS: We demonstrate an independent association between aspirin use and lower MD in a large, diverse screening cohort. This association was stronger for younger and African American women: two groups at greater risk for ER- breast cancer. These results contribute to the importance of investigating aspirin for breast cancer prevention.

Breast parenchymal patterns in processed versus raw digital mammograms: A large population study toward assessing differences in quantitative measures across image representations.

PURPOSE: With raw digital mammograms (DMs), which retain the relationship with x-ray attenuation of the breast tissue, not being routinely available, processed DMs are often the only viable means to acquire imaging measures. The authors investigate differences in quantitative measures of breast density and parenchymal texture, shown to have value in breast cancer risk assessment, between the two DM representations. METHODS: The authors report data from 8458 pairs of bilateral raw ("FOR PROCESSING") and processed ("FOR PRESENTATION") DMs acquired from 4278 women undergoing routine screening evaluation, collected with DM units from two different vendors. Breast dense tissue area and percent density (PD), as well as a range of quantitative descriptors of breast parenchymal texture (statistical, co-occurrence, run-length, and structural descriptors), were measured using previously validated, fully automated software. Feature measurements were compared using matched-pairs Wilcoxon signed-ranks test, correlation (r), and linear-mixed-effects (LME) models, where potential interactions with woman- and system-specific factors were also assessed. The authors also compared texture feature correlations with the established risk factors of the Gail lifetime risk score (rG) and breast PD (rPD), and evaluated the within woman intraclass feature correlation (ICC), a measure of bilateral breast-tissue symmetry, in raw versus processed images. RESULTS: All density measures and most of the texture features were strongly (r ≥ 0.6) or moderately (0.4 ≤ r < 0.6) correlated between raw and processed images. However, measurements were significantly different between the two imaging formats (Wilcoxon signed-ranks test, pw < 0.05). The association between measurements varied across features and vendors, and was substantially modified by woman- and system-specific image acquisition factors, such as age, BMI, and mAs/kVp, respectively. The strongest correlation, combined with minimal LME-model interactions, was observed for structural texture features. Overall, texture measures from either image representation were weakly associated with Gail lifetime risk (-0.2 ≤ rG ≤ 0.2), weakly to moderately associated with breast PD (-0.6 ≤ rPD ≤ 0.6), and had overall strong bilateral symmetry (ICC ≥ 0.6). CONCLUSIONS: Differences in measures from processed versus raw DM depend highly on the feature, the DM vendor, and image acquisition settings, where structural features appear to be more robust across the different DM settings. The reported findings may serve as a reference in the design of future large-scale studies on mammographic features and breast cancer risk assessment involving multiple DM representations.

Effectiveness of Digital Breast Tomosynthesis Compared With Digital Mammography: Outcomes Analysis From 3 Years of Breast Cancer Screening.

IMPORTANCE: Breast cancer screening with digital breast tomosynthesis (DBT) combined with digital mammography (DM) decreases false-positive examinations and increases cancer detection compared with screening with DM alone. However, the longitudinal performance of DBT screening is unknown. OBJECTIVES: To determine whether the improved outcomes observed after initial implementation of DBT screening are sustainable over time at a population level and to evaluate the effect of more than 1 DBT screening at the individual level. DESIGN, SETTING, AND PARTICIPANTS: Retrospective analysis of screening mammography metrics was performed for all patients presenting for screening mammography in an urban, academic breast center during 4 consecutive years (DM, year 0; DBT, years, 1-3). The study was conducted from September 1, 2010, to September 30, 2014 (excluding September 2011, which was the transition period from DM to DBT), for a total of 44 468 screening events attributable to a total of 23 958 unique women. Differences in screening outcomes between each DBT year and the DM year, as well as between groups of women with only 1, 2, or 3 DBT screenings, were assessed, and the odds of recall adjusted for age, race/ethnicity, breast density, and prior mammograms were estimated. Data analysis was performed between February 16 and October 26, 2015. EXPOSURE: Digital mammography screening supplemented with DBT. MAIN OUTCOMES AND MEASURES: Recall rates, cancer cases per recalled patients, and biopsy and interval cancer rates were determined. RESULTS: Screening outcome metrics were evaluated for a total of 44 468 examinations attributable to 23 958 unique women (mean [SD] age, 56.8 [11.0] years) over a 4-year period: year 0 cohort (DM0), 10 728 women; year 1 cohort (DBT1), 11 007; year 2 cohort (DBT2), 11 157; and year 3 cohort (DBT3), 11 576. Recall rates rose slightly for years 1 to 3 of DBT (88, 90, and 92 per 1000 screened, respectively) but remained significantly reduced compared with the DM0 rate of 104 per 1000 screened. Reported as odds ratios (95% CIs), the findings were DM vs DBT1, 0.83 (0.76-0.91, P < .001); DM vs DBT2, 0.85 (0.78-0.93, P < .001); and DM vs DBT3, 0.87 (0.80-0.95, P = .003). The cancer cases per recalled patients continued to rise from DM0 rate of 4.4% to 6.2% (P = .06), 6.5% (P = .03), and 6.7% (P = .02) for years 1 to 3 of DBT, respectively. Outcomes assessed for the most recent screening for individual women undergoing only 1, 2, or 3 DBT screenings during the study period demonstrated decreasing recall rates of 130, 78, and 59 per 1000 screened, respectively (P < .001). Interval cancer rates, determined using available follow-up data, decreased from 0.7 per 1000 women screened with the use of DM to 0.5 per 1000 screened with the use of DBT1. CONCLUSIONS AND RELEVANCE: Digital breast tomosynthesis screening outcomes are sustainable, with significant recall reduction, increasing cancer cases per recalled patients, and a decline in interval cancers.

Parenchymal texture analysis in digital mammography: robust texture feature identification and equivalence across devices.

An analytical framework is presented for evaluating the equivalence of parenchymal texture features across different full-field digital mammography (FFDM) systems using a physical breast phantom. Phantom images (FOR PROCESSING) are acquired from three FFDM systems using their automated exposure control setting. A panel of texture features, including gray-level histogram, co-occurrence, run length, and structural descriptors, are extracted. To identify features that are robust across imaging systems, a series of equivalence tests are performed on the feature distributions, in which the extent of their intersystem variation is compared to their intrasystem variation via the Hodges-Lehmann test statistic. Overall, histogram and structural features tend to be most robust across all systems, and certain features, such as edge enhancement, tend to be more robust to intergenerational differences between detectors of a single vendor than to intervendor differences. Texture features extracted from larger regions of interest (i.e., [Formula: see text]) and with a larger offset length (i.e., [Formula: see text]), when applicable, also appear to be more robust across imaging systems. This framework and observations from our experiments may benefit applications utilizing mammographic texture analysis on images acquired in multivendor settings, such as in multicenter studies of computer-aided detection and breast cancer risk assessment.