Background: Virtual patients (VPs) are computer screen–based simulations of patient-clinician encounters. VP use is limited by cost and low scalability. Objective: We aimed to show that VPs powered by large language models (LLMs) can generate authentic dialogues, accurately represent patient preferences, and provide personalized feedback on clinical performance. We also explored using LLMs to rate the quality of dialogues and feedback. Methods: We conducted an intrinsic evaluation study rating 60 VP-clinician conversations. We used carefully engineered prompts to direct OpenAI’s generative pretrained transformer (GPT) to emulate a patient and provide feedback. Using 2 outpatient medicine topics (chronic cough diagnosis and diabetes management), each with permutations representing different patient preferences, we created 60 conversations (dialogues plus feedback): 48 with a human clinician and 12 “self-chat” dialogues with GPT role-playing both the VP and clinician. Primary outcomes were dialogue authenticity and feedback quality, rated using novel instruments for which we conducted a validation study collecting evidence of content, internal structure (reproducibility), relations with other variables, and response process. Each conversation was rated by 3 physicians and by GPT. Secondary outcomes included user experience, bias, patient preferences represented in the dialogues, and conversation features that influenced authenticity. Results: The average cost per conversation was US $0.51 for GPT-4.0-Turbo and US $0.02 for GPT-3.5-Turbo. Mean (SD) conversation ratings, maximum 6, were overall dialogue authenticity 4.7 (0.7), overall user experience 4.9 (0.7), and average feedback quality 4.7 (0.6). For dialogues created using GPT-4.0-Turbo, physician ratings of patient preferences aligned with intended preferences in 20 to 47 of 48 dialogues (42%-98%). Subgroup comparisons revealed higher ratings for dialogues using GPT-4.0-Turbo versus GPT-3.5-Turbo and for human-generated versus self-chat dialogues. Feedback ratings were similar for human-generated versus GPT-generated ratings, whereas authenticity ratings were lower. We did not perceive bias in any conversation. Dialogue features that detracted from authenticity included that GPT was verbose or used atypical vocabulary (93/180, 51.7% of conversations), was overly agreeable (n=56, 31%), repeated the question as part of the response (n=47, 26%), was easily convinced by clinician suggestions (n=35, 19%), or was not disaffected by poor clinician performance (n=32, 18%). For feedback, detractors included excessively positive feedback (n=42, 23%), failure to mention important weaknesses or strengths (n=41, 23%), or factual inaccuracies (n=39, 22%). Regarding validation of dialogue and feedback scores, items were meticulously developed (content evidence), and we confirmed expected relations with other variables (higher ratings for advanced LLMs and human-generated dialogues). Reproducibility was suboptimal, due largely to variation in LLM performance rather than rater idiosyncrasies. Conclusions: LLM-powered VPs can simulate patient-clinician dialogues, demonstrably represent patient preferences, and provide personalized performance feedback. This approach is scalable, globally accessible, and inexpensive. LLM-generated ratings of feedback quality are similar to human ratings.

Obesity and elevated breast density are common risk factors for breast cancer, and their effects may vary by estrogen receptor (ER) subtype. However, their joint effects on ER subtype-specific risk are unknown. Understanding this relationship could enhance risk stratification for screening and prevention. Thus, we assessed the association between breast density and ER subtype according to body mass index (BMI) and menopausal status. We conducted a case-control study nested within two mammography screening cohorts, the Mayo Mammography Health Study and the San Francisco Bay Area Breast Cancer SPORE/San Francisco Mammography Registry. Our pooled analysis contained 1538 ER-positive and 285 ER-negative invasive breast cancer cases and 4720 controls matched on age, menopausal status at time of mammogram, and year of mammogram. Percent density was measured on digitized film mammograms using computer-assisted techniques. We used polytomous logistic regression to evaluate the association between percent density and ER subtype by BMI subgroup (normal/underweight, < 25 kg/m2 versus overweight/obese, ≥ 25 kg/m2). We used Wald chi-squared tests to assess for interactions between percent density and BMI. Our analysis was stratified by menopausal status and hormone therapy usage at the time of index mammogram. Percent density was associated with increased risk of overall breast cancer regardless of menopausal status or BMI. However, when analyzing breast cancer across ER subtype, we found a statistically significant (p = 0.008) interaction between percent density and BMI in premenopausal women only. Specifically, elevated percent density was associated with a higher risk of ER-negative than ER-positive cancer in overweight/obese premenopausal women [OR per standard deviation increment 2.17 (95% CI 1.50–3.16) vs 1.33 (95% CI 1.11–1.61) respectively, Pheterogeneity = 0.01]. In postmenopausal women, elevated percent density was associated with similar risk of ER-positive and ER-negative cancers, and no substantive differences were seen after accounting for BMI or hormone therapy usage. The combination of overweight/obesity and elevated breast density in premenopausal women is associated with a higher risk of ER-negative compared with ER-positive cancer. Eighteen percent of premenopausal women in the USA have elevated BMI and breast density and may benefit from lifestyle modifications involving weight loss and exercise.

Accurately identifying women with dense breasts (Breast Imaging Reporting and Data System [BI-RADS] heterogeneously or extremely dense) who are at high breast cancer risk will facilitate discussions of supplemental imaging and primary prevention. We examined the independent contribution of dense breast volume and BI-RADS breast density to predict invasive breast cancer and whether dense breast volume combined with Breast Cancer Surveillance Consortium (BCSC) risk model factors (age, race/ethnicity, family history of breast cancer, history of breast biopsy, and BI-RADS breast density) improves identifying women with dense breasts at high breast cancer risk. We conducted a case-control study of 1720 women with invasive cancer and 3686 control subjects. We calculated ORs and 95% CIs for the effect of BI-RADS breast density and Volpara™ automated dense breast volume on invasive cancer risk, adjusting for other BCSC risk model factors plus body mass index (BMI), and we compared C-statistics between models. We calculated BCSC 5-year breast cancer risk, incorporating the adjusted ORs associated with dense breast volume. Compared with women with BI-RADS scattered fibroglandular densities and second-quartile dense breast volume, women with BI-RADS extremely dense breasts and third- or fourth-quartile dense breast volume (75% of women with extremely dense breasts) had high breast cancer risk (OR 2.87, 95% CI 1.84–4.47, and OR 2.56, 95% CI 1.87–3.52, respectively), whereas women with extremely dense breasts and first- or second-quartile dense breast volume were not at significantly increased breast cancer risk (OR 1.53, 95% CI 0.75–3.09, and OR 1.50, 95% CI 0.82–2.73, respectively). Adding continuous dense breast volume to a model with BCSC risk model factors and BMI increased discriminatory accuracy compared with a model with only BCSC risk model factors (C-statistic 0.639, 95% CI 0.623–0.654, vs. C-statistic 0.614, 95% CI 0.598–0.630, respectively; P < 0.001). Women with dense breasts and fourth-quartile dense breast volume had a BCSC 5-year risk of 2.5%, whereas women with dense breasts and first-quartile dense breast volume had a 5-year risk ≤ 1.8%. Risk models with automated dense breast volume combined with BI-RADS breast density may better identify women with dense breasts at high breast cancer risk than risk models with either measure alone.

Several studies have shown that mammographic texture features are associated with breast cancer risk independent of the contribution of breast density. Thus, texture features may provide novel information for risk stratification. We examined the association of a set of established texture features with breast cancer risk by tumor type and estrogen receptor (ER) status, accounting for breast density. This study combines five case–control studies including 1171 breast cancer cases and 1659 controls matched for age, date of mammogram, and study. Mammographic breast density and 46 breast texture features, including first- and second-order features, Fourier transform, and fractal dimension analysis, were evaluated from digitized film-screen mammograms. Logistic regression models evaluated each normalized feature with breast cancer after adjustment for age, body mass index, first-degree family history, percent density, and study. Of the mammographic features analyzed, fractal dimension and second-order statistics features were significantly associated (p < 0.05) with breast cancer. Fractal dimensions for the thresholds equal to 10% and 15% (FD_TH_10 and FD_TH_15) were associated with an increased risk of breast cancer while thresholds from 60% to 85% (FD_TH_60 to FD_TH_85) were associated with a decreased risk. Increasing the FD_TH_75 and Energy feature values were associated with a decreased risk of breast cancer while increasing Entropy was associated with a increased risk of breast cancer. For example, 1 standard deviation increase of FD_TH_75 was associated with a 13% reduced risk of breast cancer (odds ratio = 0.87, 95% confidence interval 0.79–0.95). Overall, the direction of associations between features and ductal carcinoma in situ (DCIS) and invasive cancer, and estrogen receptor positive and negative cancer were similar. Mammographic features derived from film-screen mammograms are associated with breast cancer risk independent of percent mammographic density. Some texture features also demonstrated associations for specific tumor types. For future work, we plan to assess risk prediction combining mammographic density and features assessed on digital images.

Over 40% of women undergoing breast screening have mammographically dense breasts. Elevated mammographic breast density (MBD) is an established breast cancer risk factor and is known to mask tumors within the dense tissue. However, the association of MBD with high risk benign breast disease (BBD) is unknown. We analyzed data for 3400 women diagnosed with pathologically confirmed BBD in the Mayo Clinic BBD cohort from 1985–2001, with a clinical MBD measure (either parenchymal pattern (PP) or Breast Imaging Reporting and Data Systems (BI-RADS) density) and expert pathology review. Risk factor information was collected from medical records and questionnaires. MBD was dichotomized as dense (PP classification P2 or DY, or BI-RADS classification c or d) or non-dense (PP classification N1 or P1, or BI-RADS classification a or b). Associations of clinical and histologic characteristics with MBD were examined using logistic regression analysis to estimate odds ratios (ORs) with 95% confidence intervals (CIs). Of 3400 women in the study, 2163 (64%) had dense breasts. Adjusting for age and body mass index (BMI), there were positive associations of dense breasts with use of hormone therapy (HT), lack of lobular involution, presence of atypical lobular hyperplasia (ALH), histologic fibrosis, columnar cell hyperplasia/flat epithelia atypia (CCH/FEA), sclerosing adenosis (SA), cyst, usual ductal hyperplasia, and calcifications. In fully adjusted multivariate models, HT (1.3, 95% CI 1.1–1.5), ALH (1.5, 95% CI 1.0–2.2), lack of lobular involution (OR 1.6, 95% CI 1.2–2.1, compared to complete involution), fibrosis (OR 2.2, 95% CI 1.9–2.6) and CCH/FEA (OR 1.3, 95% CI 1.0–1.6) remained significantly associated with high MBD. Our findings support an association between high risk BBD and high MBD, suggesting that risks associated with the latter may act early in breast carcinogenesis.