Search

PURPOSE: The aim of this study was to describe the current state of science regarding independent external validation of artificial intelligence (AI) technologies for screening mammography. METHODS: A systematic review was performed across five databases (Embase, PubMed, IEEE Explore, Engineer Village, and arXiv) through December 10, 2020. Studies that used screening examinations from real-world settings to externally validate AI algorithms for mammographic cancer detection were included. The main outcome was diagnostic accuracy, defined by area under the receiver operating characteristic curve (AUC). Performance was also compared between radiologists and either stand-alone AI or combined radiologist and AI interpretation. Study quality was assessed using the Quality Assessment of Diagnostic Accuracy Studies 2 tool. RESULTS: After data extraction, 13 studies met the inclusion criteria (148,361 total patients). Most studies (77% [n = 10]) evaluated commercially available AI algorithms. Studies included retrospective reader studies (46% [n = 6]), retrospective simulation studies (38% [n = 5]), or both (15% [n = 2]). Across 5 studies comparing stand-alone AI with radiologists, 60% (n = 3) demonstrated improved accuracy with AI (AUC improvement range, 0.02-0.13). All 5 studies comparing combined radiologist and AI interpretation with radiologists alone demonstrated improved accuracy with AI (AUC improvement range, 0.028-0.115). Most studies had risk for bias or applicability concerns for patient selection (69% [n = 9]) and the reference standard (69% [n = 9]). Only two studies obtained ground-truth cancer outcomes through regional cancer registry linkage. CONCLUSIONS: To date, external validation efforts for AI screening mammographic technologies suggest small potential diagnostic accuracy improvements but have been retrospective in nature and suffer from risk for bias and applicability concerns.

BackgroundMost evidence for interpregnancy interval (IPI) and adverse birth outcomes come from studies that are prone to incomplete control for confounders that vary between women. Comparing pregnancies to the same women can address this issue.MethodsWe conducted an international longitudinal cohort study of 5,521,211 births to 3,849,193 women from Australia (1980-2016), Finland (1987-2017), Norway (1980-2016) and the United States (California) (1991-2012). IPI was calculated based on the time difference between two dates-the date of birth of the first pregnancy and the date of conception of the next (index) pregnancy. We estimated associations between IPI and preterm birth (PTB), spontaneous PTB, and small-for-gestational age births (SGA) using logistic regression (between-women analyses). We also used conditional logistic regression comparing IPIs and birth outcomes in the same women (within-women analyses). Random effects meta-analysis was used to calculate pooled adjusted odds ratios (aOR).ResultsCompared to an IPI of 18-23 months, there was insufficient evidence for an association between IPI 24 month IPIs.ConclusionsWe found consistently elevated odds of adverse birth outcomes following long IPIs. IPI shorter than 6 months were associated with elevated risk of spontaneous PTB, but there was insufficient evidence for increased risk of other adverse birth outcomes. Current recommendations of waiting at least 24 months to conceive after a previous pregnancy, may be unnecessarily long in high-income countries.

IntroductionShort interpregnancy interval (IPI) has been linked to adverse pregnancy outcomes. WHO recommends waiting at least 2 years after a live birth and 6 months after miscarriage or induced termination before conception of another pregnancy. The evidence underpinning these recommendations largely relies on data from low/middle-income countries. Furthermore, recent epidemiological investigations have suggested that these studies may overestimate the effects of IPI due to residual confounding. Future investigations of IPI effects in high-income countries drawing from large, population-based data sources are needed to inform IPI recommendations. We aim to assess the impact of IPIs on maternal and child health outcomes in high-income countries.Methods and analysisThis international longitudinal retrospective cohort study will include more than 18 million pregnancies, making it the largest study to investigate IPI in high-income countries. Population-based data from Australia, Finland, Norway and USA will be used. Birth records in each country will be used to identify consecutive pregnancies. Exact dates of birth and clinical best estimates of gestational length will be used to estimate IPI. Administrative birth and health data sources with >99% coverage in each country will be used to identify maternal sociodemographics, pregnancy complications, details of labour and delivery, birth and child health information. We will use matched and unmatched regression models to investigate the impact of IPI on maternal and infant outcomes, and conduct meta-analysis to pool results across countries.Ethics and disseminationEthics boards at participating sites approved this research (approval was not required in Finland). Findings will be published in peer-reviewed journals and presented at international conferences, and will inform recommendations for optimal IPI in high-income countries. Findings will provide important information for women and families planning future pregnancies and for clinicians providing prenatal care and giving guidance on family planning.

Background Controversy exists regarding the optimal negative margin width for ductal carcinoma in situ (DCIS) treated with breast-conserving surgery and whole-breast irradiation (WBRT). Methods A multidisciplinary consensus panel used a meta-analysis of margin width and ipsilateral breast tumor recurrence (IBTR) from a systematic review of 20 studies including 7883 patients and other published literature as the evidence base for consensus. Results Negative margins halve the risk of IBTR compared with positive margins defined as ink on DCIS. A 2 mm margin minimizes the risk of IBTR compared with smaller negative margins. More widely clear margins do not significantly decrease IBTR compared with 2 mm margins. Negative margins less than 2 mm alone are not an indication for mastectomy, and factors known to impact rates of IBTR should be considered in determining the need for re-excision. Conclusion The use of a 2 mm margin as the standard for an adequate margin in DCIS treated with WBRT is associated with low rates of IBTR and has the potential to decrease re-excision rates, improve cosmetic outcome, and decrease health care costs. Clinical judgment should be used in determining the need for further surgery in patients with negative margins < 2 mm.

Artificial intelligence (AI) algorithms have been retrospectively evaluated as replacement for one radiologist in screening mammography double-reading; however, methods for resolving discordance between radiologists and AI in the absence of ‘real-world’ arbitration may underestimate cancer detection rate (CDR) and recall. In 108,970 consecutive screens from a population screening program (BreastScreen WA, Western Australia), 20,120 were radiologist/AI discordant without real-world arbitration. Recall probabilities were randomly assigned for these screens in 1000 simulations. Recall thresholds for screen-detected and interval cancers (sensitivity) and no cancer (false-positive proportion, FPP) were varied to calculate mean CDR and recall rate for the entire cohort. Assuming 100% sensitivity, the maximum CDR was 7.30 per 1000 screens. To achieve >95% probability that the mean CDR exceeded the screening program CDR (6.97 per 1000), interval cancer sensitivities ≥63% (at 100% screen-detected sensitivity) and ≥91% (at 80% screen-detected sensitivity) were required. Mean recall rate was relatively constant across sensitivity assumptions, but varied by FPP. FPP > 6.5% resulted in recall rates that exceeded the program estimate (3.38%). CDR improvements depend on a majority of interval cancers being detected in radiologist/AI discordant screens. Such improvements are likely to increase recall, requiring careful monitoring where AI is deployed for screen-reading. [ABSTRACT FROM AUTHOR]