Chest X-rays (CXR) are essential for diagnosing a variety of conditions, but when used on new populations, model generalizability issues limit their efficacy. Generative AI, particularly denoising diffusion probabilistic models (DDPMs), offers a promising approach to generating synthetic images, enhancing dataset diversity. This study investigates the impact of synthetic data supplementation on the performance and generalizability of medical imaging research. The study employed DDPMs to create synthetic CXRs conditioned on demographic and pathological characteristics from the CheXpert dataset. These synthetic images were used to supplement training datasets for pathology classifiers, with the aim of improving their performance. The evaluation involved three datasets (CheXpert, MIMIC-CXR, and Emory Chest X-ray) and various experiments, including supplementing real data with synthetic data, training with purely synthetic data, and mixing synthetic data with external datasets. Performance was assessed using the area under the receiver operating curve (AUROC). Adding synthetic data to real datasets resulted in a notable increase in AUROC values (up to 0.02 in internal and external test sets with 1000% supplementation, p-value <0.01 in all instances). When classifiers were trained exclusively on synthetic data, they achieved performance levels comparable to those trained on real data with 200%-300% data supplementation. The combination of real and synthetic data from different sources demonstrated enhanced model generalizability, increasing model AUROC from 0.76 to 0.80 on the internal test set (p-value <0.01). Synthetic data supplementation significantly improves the performance and generalizability of pathology classifiers in medical imaging. Dr. Gichoya is a 2022 Robert Wood Johnson Foundation Harold Amos Medical Faculty Development Program and declares support from RSNA Health Disparities grant (#EIHD2204), Lacuna Fund (#67), Gordon and Betty Moore Foundation, NIH (NIBIB) MIDRC grant under contracts 75N92020C00008 and 75N92020C00021, and NHLBI Award Number R01HL167811.

Background and Aims: Women are underdiagnosed and undertreated for cardiovascular disease (CVD). Automatic quantification of breast arterial calcification (BAC) on screening mammography can identify women at risk for CVD. This study aimed to determine whether artificial intelligence–based automatic quantification of BAC from screening mammograms predicts CVD and mortality beyond PREVENT scores in a large, racially diverse, multi-institutional population. Methods: This retrospective cohort study included 123 762 women from two healthcare systems who had screening mammograms. Breast arterial calcification was quantified using a transformer-based neural network for segmentation. Breast arterial calcification severity was categorized as zero (0 mm2), mild (>0–10 mm2), moderate (>10–25 mm2), and severe (>25 mm2). Kaplan–Meier analysis, Cox proportional hazards, and Fine–Gray competing event models were used to examine the association between BAC and major adverse cardiovascular events (MACE). Results: Breast arterial calcification was detected in 16.1% (internal cohort) and 20.6% (external cohort) of women and provided significant prognostic value incremental to the PREVENT score. In PREVENT adjusted models, a clear dose–response was observed. Compared with zero BAC, mild [internal: hazard ratio (HR) 1.32, 95% confidence interval (CI) 1.10–1.59; external: HR 1.28, 95% CI 1.17–1.39], moderate (internal: HR 1.75, 95% CI 1.23–2.50; external: HR 1.79, 95% CI 1.55–2.06), and severe BAC (internal: HR 3.29, 95% CI 2.15–5.05; external: HR 2.80, 95% CI 2.36–3.32) were all prognostic for any MACE. Each 1 mm2 increase in BAC conferred an additional 2%–3% risk for MACE (P < .001). Conclusions: Automatically quantified BAC is an independent predictor of MACE and mortality, adding prognostic value to the PREVENT score. This approach may provide an opportunistic cardiovascular risk assessment during routine mammography screening without additional radiation exposure to guide earlier and more effective preventive care for women.

Large language model–generated labels can introduce systematic bias in artificial intelligence binary classification model performance evaluation, with disease prevalence critically determining whether sensitivity or specificity is most affected.

Background: The objective of this study is to understand chronic obstructive pulmonary disease (COPD) phenotypes and their progressions by quantifying heterogeneities of lung ventilation from the single photon emission computed tomography (SPECT) images and establishing associations with the quantitative computed tomography (qCT) imaging-based clusters and variables. Methods: Eight COPD patients completed a longitudinal study of three visits with intervals of about a year. CT scans of these subjects at residual volume, functional residual capacity, and total lung capacity were taken for all visits. The functional and structural qCT-based variables were derived, and the subjects were classified into the qCT-based clusters. In addition, the SPECT variables were derived to quantify the heterogeneity of lung ventilation. The correlations between the key qCT-based variables and SPECT-based variables were examined. Results: The SPECT-based coefficient of variation (CV _Total ), a measure of ventilation heterogeneity, showed strong correlations (|r| ≥ 0.7) with the qCT-based functional small airway disease percentage (fSAD% _Total ) and emphysematous tissue percentage (Emph% _Total ) in the total lung on cross-sectional data. As for the two-year changes, the SPECT-based maximum tracer concentration (TC _max ), a measure of hot spots, exhibited strong negative correlations with fSAD% _Total , Emph% _Total , average airway diameter in the left upper lobe, and airflow distribution in the middle and lower lobes. Conclusion: Small airway disease is highly associated with the heterogeneity of ventilation in COPD lungs. TC _max is a more sensitive functional biomarker for COPD progression than CV _Total . Besides fSAD% _Total and Emph% _Total , segmental airways narrowing and imbalanced ventilation between upper and lower lobes may contribute to the development of hot spots over time.

Around nine million people have been exposed to toxic humidifier disinfectants (HDs) in Korea. HD exposure may lead to HD-associated lung injuries (HDLI). However, many people who have claimed that they experienced HD exposure were not diagnosed with HDLI but still felt discomfort, possibly due to the unknown effects of HD. Therefore, this study examined HD-exposed subjects with normal-appearing lungs, as well as unexposed subjects, in clusters (subgroups) with distinct characteristics, classified by deep-learning-derived computed-tomography (CT)-based tissue pattern latent traits. Among the major clusters, cluster 0 (C0) and cluster 5 (C5) were dominated by HD-exposed and unexposed subjects, respectively. C0 was characterized by features attributable to lung inflammation or fibrosis in contrast with C5. The computational fluid and particle dynamics (CFPD) analysis suggested that the smaller airway sizes observed in the C0 subjects led to greater airway resistance and particle deposition in the airways. Accordingly, women appeared more vulnerable to HD-associated lung abnormalities than men.

Pulmonary embolism (PE) is a leading cause of cardiovascular mortality, yet the real-world performance of FDA-cleared AI detection models remains incompletely characterized. We retrospectively evaluated two FDA-cleared AI algorithms from a single commercial platform (Aidoc Medical BriefCase)—one for PE triage on dedicated CT pulmonary angiography (CTPA; n = 30,678) and one for incidental PE (iPE) detection on routine contrast-enhanced CTs (n = 37,191), across a 17-facility academic health system. Reference-standard labels were extracted from radiology reports using a validated LLM pipeline (97% accuracy, κ = 0.94). The PE model achieved 86.8% sensitivity and 99.1% specificity, with sensitivity declining from 99.3% for saddle emboli to 72.9% for subsegmental PE, and from 89.7% for acute to 65.3% for non-acute PE. The iPE model achieved 73.5% sensitivity and 99.8% specificity. Both models demonstrated lower sensitivity than FDA-clearance benchmarks while exceeding cleared specificity, with diminishing performance for peripheral and non-acute emboli mirroring known human reader limitations and underscoring the need for standardized post-market surveillance of AI-enabled medical devices.