FRAME: separating sampling variation from representational cause in medical imaging fairness

Subgroup performance differences are the standard evidence for fairness bias in medical imaging, and the usual response removes the demographic information that a model encodes. Here we introduce Fair-model Reference And Mechanism Evaluation (FRAME), a two-step framework for auditing such a claim. The first step derives a fair-model reference, the distribution of the difference under exact fairness at the observed subgroup sizes. In the second step, we test the remainder with two operators in representation space. One operator cannot change a within-group ranking by construction. Across 702,206 images and 36 encoders, the reference accounts for a median 41% of the reported race difference and 22% of the age difference. Injecting demographic decodability leaves the remainder unchanged, while entangling the group with the disease direction raises the race difference from 0.077 to 0.118. No intervention we tested changes the remainder more than a change of random seed does. Those interventions reduce a difference at the operating point and leave the within-group ranking difference at a median of 0.000. Applied to 89 differences in 9 published studies across 6 medical imaging modalities, the reference accounts for a median 25% of a rate difference and 70% of a difference in the area under the receiver operating characteristic curve. Image-text pretraining instead raises worst-group performance by about 0.05. Applying FRAME before choosing an intervention could distinguish differences that need a mechanistic explanation from differences compatible with sampling variation at the current cohort sizes.

A Statistical Framework for Data-Driven Discovery of Differential Performance in Clinical Risk Prediction Models

Predictive models employing artificial intelligence (AI) and machine learning (ML) are increasingly being used for decision support in healthcare settings. These models may exhibit differential performance across population subgroups defined by race, age, sex, and other factors and cause disparate clinical impacts, leading to intensive recent study of what has been termed "model fairness". While many methods have been proposed to assess risk prediction model fairness, these techniques generally require that the end user pre-specify the groups across which fairness is to be evaluated. In real-world settings, however, important model performance disparities may arise in unknown subgroups defined by multiple intersecting characteristics. To address this problem, we propose the unfairness tree (utree), a data-driven recursive partitioning framework for identifying subgroups with differential model performance. In simulations, the utree exhibits nominal empirical type I error rates and good ability to detect, quantify, and characterize performance discrepancies defined by higher-order variable interactions. In six mortality risk models fit to the GUSTO-I acute myocardial infarction trial dataset, utrees identified subgroup-specific performance patterns, with age, sex, blood pressure, and Killip class consistently associated with differential model performance.

From Unsupervised Subgroups to Hypothetical State-Intervention Policies: An Evaluation of Selected Subgrouping Methods in Observational Health Data

Conventional subgroup analyses can yield unstable and difficult-to-interpret conclusions, especially in observational biomedical data where each individual is observed under only one exposure state, true individual treatment effects are unavailable, and causal structure is uncertain. We investigate whether subgroups constructed from pretreatment characteristics, without using exposure, outcome, or estimated treatment-effect information, can serve as interpretable units for budget-constrained policy prioritization. We propose a framework combining causal-discovery-informed covariate selection, discovery-evaluation sample splitting, inductive unsupervised clustering, uncertainty-aware subgroup selection, and held-out doubly robust policy evaluation. We compare K-means, hard, membership-weighted, and stochastic Fuzzy C-means, Bayesian Gaussian mixture models, and a supervised causal-forest-derived CATE-tree comparator. Policies are evaluated under a 70% budget for hypothetical obesity-to-non-obesity and elevated-to-lower-glucose state shifts in the PIMA Indians Diabetes dataset and for a lifetime-smoking-history contrast in NHANES. The highest estimated ungated utilities were 0.799 for the BMI policy using Bayesian GMM, 0.735 for the glucose policy using hard or membership-weighted FCM, and 0.775 for the smoking-history policy using K-means. All paired 95% confidence intervals for policy-risk differences included zero, and no comparison remained statistically significant after Holm adjustment. Bayesian pooling generally preserved ungated allocations, whereas Empirical Bernstein gating was more conservative. Policies with similar estimated utility could nevertheless prioritize different individuals. The findings should be interpreted as assumption-dependent decision-support evidence for hypothetical state contrasts rather than proof of intervention benefit.

The Resolution of Causal Heterogeneity

Causal subgroup analyses often report a small number of groups summarizing treatment effect heterogeneity, as if that number were a well-defined estimand. Outside genuinely latent class populations, however, a ``true'' subgroup count is model dependent rather than a population functional. We replace it with a new population estimand, the resolution profile, a functional of the causal feature law giving the fewest groups explaining a prescribed fraction of causal heterogeneity, defined for every population without latent structure. Inference is organized around one cross-fitted Bayesian-bootstrap posterior for a single structured moment process, its scores corrected with influence functions, so that paths, profiles, fixed-resolution summaries, and subgroup effects follow by composition. A uniform conditional Bernstein--von Mises theorem over a loss class containing the nonsmooth quantization losses shows this posterior merges with the efficient Gaussian limit under stated nuisance-rate and margin conditions. Subgroup-number uncertainty is not model selection but threshold nonregularity, the profile being an integer-valued threshold of a continuous path, discontinuous in the law at each knot. At these knots no single-valued selector is locally uniformly consistent over root-$n$ neighborhoods, and the set-valued report obtained by inverting a simultaneous band retains locally uniform validity over exactly the same perturbations. Simulations support the approximations, and an analysis of the MineThatData e-mail experiment illustrates the resolution-indexed report, in which two to three groups summarize the visit response while finer structure falls below a noise-floor diagnostic.

Interpretable Machine Learning for Air Pollution and Respiratory Health Prediction: A Socioeconomic Subgroup Analysis

Air pollution and climate-related stressors are increasingly important concerns for respiratory health, especially in settings with unequal environmental exposure and healthcare capacity. This study evaluates an interpretable machine learning framework for predicting respiratory disease rates and air-quality status using structured country-level weekly data. Two supervised learning tasks were considered: regression of respiratory disease rate per 100,000 population and binary classification of air-quality status. Nine regression models and nine classification models were compared using nested cross-validation. Model interpretation was conducted using SHAP values, and subgroup analysis was performed across income levels and geographic regions. The results showed that PM2.5 concentration was the dominant predictor of respiratory disease rate, with linear and regularized linear models achieving the strongest regression performance. For air-quality classification, models achieved high balanced accuracy when PM2.5 was included, but performance decreased substantially when PM2.5 was removed, indicating strong dependence on pollutant-related information. SHAP analysis showed that, without PM2.5, socioeconomic and meteorological variables such as GDP per capita, precipitation, and healthcare access became more influential. Subgroup analysis showed similar aggregate regression error across income groups, but PM2.5 contributed more strongly to predictions in lower-middle-income countries. These results show that model accuracy alone is not sufficient for climate-health prediction. Interpretable models can help identify dominant pollution-related signals, test whether results depend on key pollutant variables, and show whether prediction patterns differ across socioeconomic groups.

Robust Subgroup Analysis for Heterogeneous Censored Data

Subgroup analysis is important in practice because real-world data typically come from heterogeneous populations, where meaningful patterns can differ substantially across subpopulations. Correctly identifying these subgroups can improve prediction accuracy, prevent biased or misleading conclusions, and support more effective, targeted decision-making. While most existing subgroup analysis methods are developed for complete data, in this paper we propose a novel and robust approach for censored data under heterogeneous accelerated failure time (AFT) models. Specifically, we combine inverse probability weighting, M-estimation, and concave pairwise fusion penalization to simultaneously identify subgroups and estimate covariate effects for heterogeneous censored data, without requiring prior knowledge of individual subgroup memberships. We further develop an efficient RISA-ADMM algorithm to implement the method and establish its convergence. Furthermore, we derive the theoretical properties of the proposed estimators under mild regularity conditions. Extensive simulations and an application to the German credit dataset demonstrate the robustness and effectiveness of our approach.

Beyond Metadata: CAPRA for Hidden Subgroup Analysis under Missing Metadata in Medical Imaging

Medical imaging models are often deployed without the demographic, acquisition, and quality metadata needed for subgroup auditing. Once those metadata disappear, clinically critical failure modes can be masked by strong aggregate performance, and many robust-learning methods lose the group structure they rely on. We present CAPRA, a calibrated proxy-axis framework for hidden subgroup analysis under missing metadata. CAPRA predicts image-derived semantic axes, calibrates axis posteriors on a small metadata-labeled split via patient-level cross-fitting, and organizes those posteriors into a calibrated subgroup interface that supports both deployment-time failure analysis and downstream robust learning without requiring subgroup labels at deployment. Across fundus, dermoscopy, and chest radiography, CAPRA reveals disparity patterns missed by metadata-only slicing, remains informative under dataset shift, and produces subgroup partitions that align more closely with explicit failure axes than image-only or latent-slice baselines. The same interface can also be reused by downstream robust learners, although those gains are domain-dependent. Overall, CAPRA turns hidden subgroup analysis under missing metadata into a calibrated, interpretable, and reusable subgroup interface for deployment-time analysis and robust transfer.

Who Gets Missed in the Tail? Thresholded Subgroup Underdiagnosis in Long-Tailed Chest X-ray Classification

In chest X-ray (CXR) classification, acceptable ranking performance can still leave rare-positive patients below threshold, especially within subgroups. We study this pre-deployment fairness problem as an audit question: after a long-tailed multi-label CXR model is converted from scores into decisions, who is missed? Across VinDr-CXR and MIMIC-CXR/CXR-LT, we use a diagnostic ladder to separate class-level long-tail losses, subgroup-aware weighting, group robustness, and threshold selection. On VinDr-CXR, group-tail weighting followed by tail-aware thresholding reduces tail FNR from 0.665 to 0.269, sex worst-group FNR from 0.705 to 0.157, and age worst-group FNR from 0.822 to 0.133, while macro-mAP increases from 0.611 to 0.635. On MIMIC-CXR/CXR-LT, the same score-to-threshold comparison reduces tail FNR from 0.866 to 0.741 and lowers worst-group FNR across sex, age, race, and insurance; residual missed-positive rates nevertheless remain high. Paired bootstrap contrasts on VinDr support the thresholded FNR reductions, and GroupDRO reference runs indicate that aggregate group robustness alone does not remove rare subgroup misses in this setting. The study supports a narrow audit claim: rare-label fairness in CXR depends jointly on the finding, subgroup, and operating threshold, not on label frequency or ranking metrics alone.

BenchX: Benchmarking AI Models for Cancer Detection and Localization with Demographic and Protocol Biases

Artificial intelligence (AI) has achieved remarkable success in medical imaging, but it is widely recognized that these models often perform inconsistently across real-world clinical settings. Such inconsistencies occur when patient demographics and imaging protocols vary, for example, in detecting small tumors, analyzing scans from different contrast phases, or evaluating patients of different ages or sexes. To quantify these inconsistencies, we develop a large-scale, open benchmark of 85,355 CT scans that systematically evaluates 12 tumor-detection AI models across tumor size, location, patient subgroup, and imaging protocol. We leverage large language models (LLMs) to extract and organize subgroup information from clinical data, which makes the analysis both scalable and reproducible. Our benchmark reveals that current state-of-the-art AI models, optimized for average accuracy, perform poorly in rare or underrepresented subgroups, such as young, female African Americans. However, collecting sufficient annotated data for these rare cases is often impractical. The benchmark provides a foundation for building more reliable and robust AI models for tumor detection and highlighting the need for rigorous, subgroup-level evaluation in medical imaging and computer vision. Datasets, code