Search

This diagnostic accuracy study evaluates whether artificial intelligence can overcome human mammography interpretation limits with a rigorous, unbiased evaluation of machine learning algorithms., Key Points Question How do deep learning algorithms perform compared with radiologists in screening mammography interpretation? Findings In this diagnostic accuracy study using 144 231 screening mammograms from 85 580 women from the United States and 166 578 screening mammograms from 68 008 women from Sweden, no single artificial intelligence algorithm outperformed US community radiologist benchmarks; including clinical data and prior mammograms did not improve artificial intelligence performance. However, combining best-performing artificial intelligence algorithms with single-radiologist assessment demonstrated increased specificity. Meaning Integrating artificial intelligence to mammography interpretation in single-radiologist settings could yield significant performance improvements, with the potential to reduce health care system expenditures and address resource scarcity experienced in population-based screening programs., Importance Mammography screening currently relies on subjective human interpretation. Artificial intelligence (AI) advances could be used to increase mammography screening accuracy by reducing missed cancers and false positives. Objective To evaluate whether AI can overcome human mammography interpretation limitations with a rigorous, unbiased evaluation of machine learning algorithms. Design, Setting, and Participants In this diagnostic accuracy study conducted between September 2016 and November 2017, an international, crowdsourced challenge was hosted to foster AI algorithm development focused on interpreting screening mammography. More than 1100 participants comprising 126 teams from 44 countries participated. Analysis began November 18, 2016. Main Outcomes and Measurements Algorithms used images alone (challenge 1) or combined images, previous examinations (if available), and clinical and demographic risk factor data (challenge 2) and output a score that translated to cancer yes/no within 12 months. Algorithm accuracy for breast cancer detection was evaluated using area under the curve and algorithm specificity compared with radiologists’ specificity with radiologists’ sensitivity set at 85.9% (United States) and 83.9% (Sweden). An ensemble method aggregating top-performing AI algorithms and radiologists’ recall assessment was developed and evaluated. Results Overall, 144 231 screening mammograms from 85 580 US women (952 cancer positive ≤12 months from screening) were used for algorithm training and validation. A second independent validation cohort included 166 578 examinations from 68 008 Swedish women (780 cancer positive). The top-performing algorithm achieved an area under the curve of 0.858 (United States) and 0.903 (Sweden) and 66.2% (United States) and 81.2% (Sweden) specificity at the radiologists’ sensitivity, lower than community-practice radiologists’ specificity of 90.5% (United States) and 98.5% (Sweden). Combining top-performing algorithms and US radiologist assessments resulted in a higher area under the curve of 0.942 and achieved a significantly improved specificity (92.0%) at the same sensitivity. Conclusions and Relevance While no single AI algorithm outperformed radiologists, an ensemble of AI algorithms combined with radiologist assessment in a single-reader screening environment improved overall accuracy. This study underscores the potential of using machine learning methods for enhancing mammography screening interpretation.

This cohort study used data from the Breast Cancer Surveillance Consortium to examine associations between incidence of invasive breast cancer and breast density, body mass index, and age among women aged 65 years or older., Key Points Question Is breast density associated with invasive breast cancer among women aged 65 years or older, and does body mass index (BMI) modify any associations between breast density and breast cancer risk? Findings In this cohort study of 221 714 screening mammograms among 193 787 US women aged 65 years or older, heterogeneous or extreme breast density vs scattered fibroglandular density was associated with an increased risk of invasive breast cancer among women aged 65 to 74 years and among those aged 75 years or older with a BMI of 25 or higher. No interaction between BMI and breast density was found in either age group. Meaning The findings suggest that breast density is associated with an increased risk of invasive breast cancer among women aged 65 years or older., Importance Breast density is associated with breast cancer risk in women aged 40 to 65 years, but there is limited evidence of its association with risk of breast cancer among women aged 65 years or older. Objective To compare the association between breast density and risk of invasive breast cancer among women aged 65 to 74 years vs women aged 75 years or older and to evaluate whether the association is modified by body mass index (BMI). Design, Setting, and Participants This prospective cohort study used data from the Breast Cancer Surveillance Consortium from January 1, 1996, to December 31, 2012, for US women aged 65 years or older who underwent screening mammography. Data were analyzed from January 1, 2018, to December 31, 2020. Exposures Breast Imaging Reporting and Data System breast density category, age, and BMI. Main Outcomes and Measures The 5-year cumulative incidence of invasive breast cancer by level of breast density (almost entirely fat, scattered fibroglandular densities, or heterogeneous or extreme density) and age (65-74 vs ≥75 years) was calculated using weighted means. Cox proportional hazards models were fit to estimate the association of breast density with invasive breast cancer risk. The likelihood ratio test was used to test the interaction between BMI and breast density. Results A total of 221 714 screening mammograms from 193 787 women were included in the study; a total of 38% of the study population was aged 75 years or older. Of the mammograms, most were from women aged 65 to 74 years (64.6%) and non-Hispanic White individuals (81.4%). The 5-year cumulative incidence of invasive breast cancer increased in association with increasing breast density among women aged 65 to 74 years (almost entirely fatty breasts: 11.3 per 1000 women [95% CI, 10.4-12.5 per 1000 women]; scattered fibroglandular densities: 17.2 per 1000 women [95% CI, 16.1-17.9 per 1000 women]; extremely or heterogeneously dense breasts: 23.7 per 1000 women [95% CI, 22.4-25.3 per 1000 women]) and among those aged 75 years or older (fatty breasts: 13.5 per 1000 women [95% CI, 11.6-15.5]; scattered fibroglandular densities: 18.4 per 1000 women [95% CI, 17.0-19.5 per 1000 women]; extremely or heterogeneously dense breasts: 22.5 per 1000 women [95% CI, 20.2-24.2 per 1000 women]). Extreme or heterogeneous breast density was associated with increased risk of breast cancer compared with scattered fibroglandular breast density in both age categories (65-74 years: hazard ratio [HR], 1.39 [95% CI, 1.28-1.50]; ≥75 years: HR, 1.23 [95% CI, 1.10-1.37]). Women with almost entirely fatty breasts had a decrease of approximately 30% (range, 27%-34%) in the risk of invasive breast cancer compared with women with scattered fibroglandular breast density (65-74 years: HR, 0.66 [95% CI, 0.58-0.75]; ≥75 years: HR, 0.73; 95% CI, 0.62-0.86). Associations between breast density and breast cancer risk were not significantly modified by BMI (for age 65-74 years: likelihood ratio test, 2.67; df, 2; P = .26; for age ≥75 years, 2.06; df, 2; P = .36). Conclusions and Relevance The findings suggest that breast density is associated with increased risk of invasive breast cancer among women aged 65 years or older. Breast density and life expectancy should be considered together when discussing the potential benefits vs harms of continued screening mammography in this population.

This cross-sectional study evaluates the out-of-pocket costs of diagnostic breast imaging services incurred by commercially insured women who underwent additional imaging evaluation and procedures after screening mammography.

This cohort study investigates the rates of breast cancer screening before and during the COVID-19 pandemic among women in Washington State.