Fairness evaluation of clinical prediction models for binary outcomes often involves fairness metrics that compare the area under the receiver operating characteristic curve (AUC, or c-statistic) acro...
FairGlucose is a 300‑patient CGM cohort balanced across 12 demographic strata, providing 132,480 forecasting samples and 3,945 behavioral events. Benchmarking 33 models on 2‑hour glucose forecasting revealed that population‑level validation masks significant subgroup disparities, with error ratios ranging from 0.8 to 1.4 and T1D patients experiencing 6 mg/dL higher error than T2D. The study shows that these gaps persist across all models, align with clinically hard cases, and vary with input‑length sensitivity, underscoring the need for subgroup‑disaggregated reporting in digital health AI.
The paper examines how fairness conclusions in ICU mortality prediction using MIMIC-IV depend on the choice of metrics and the granularity of demographic analysis. It compares predictive-utility and subgroup-error metrics across various fairness interventions and introduces a lightweight adaptation strategy that balances ethnicity, gender, and insurance representation without conditioning on mortality outcomes. The study finds that different interventions can be evaluated differently across accuracy, sensitivity, and false-positive rate, and that marginal demographic summaries may hide heterogeneous error patterns within intersectional subgroups.
arXiv:2607. 07717v1 Announce Type: new Abstract: In chest X-ray (CXR) classification, acceptable ranking performance can still leave rare-positive patients below threshold, especially within subgroups.
arXiv:2611.08048v1 Announce Type: new Abstract: Performance metrics can yield different conclusions from the same predictions. This methodological tutorial explains the effects of class prevalence, d...
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?
arXiv:2607. 08953v1 Announce Type: new Abstract: Algorithmic fairness methods are increasingly used to identify and mitigate bias in machine learning models, yet most approaches are evaluated in isolation and along single demographic axes.
The paper introduces FRAME, a two‑step framework for auditing fairness claims in medical imaging. First, it derives a fair‑model reference distribution that captures the portion of subgroup performance differences attributable to sampling variation. Second, it tests the remaining difference using operators in representation space to assess whether demographic information or disease entanglement drives the bias. Across a large dataset of 702,206 images and 36 encoders, the reference explains a substantial median share of race and age differences, while interventions such as injecting demographic decodability or entangling disease direction have limited impact on the residual bias.
The paper investigates demographic bias in large language model (LLM) predictions of ICU mortality and introduces Case-Based Prompting (CAP), a training‑free method that adds clinically similar historical cases as contextual evidence. CAP improves predictive performance on the MIMIC‑IV dataset, raising AUROC from 0.806 to 0.873 and AUPRC from 0.497 to 0.694, while reducing subgroup disparities for sex and White‑Black comparisons. The study also shows that simply removing demographic information does not uniformly enhance fairness or accuracy, and that feature‑dependence patterns remain consistent across groups.
Reweighting, a common pre‑processing technique used to promote fairness by assigning different weights to data points, can unintentionally worsen fairness under certain data conditions. Controlled simulations varying group size and event prevalence show that reweighting may reduce underestimation for one group while increasing it for another, especially when the higher‑prevalence group is a minority. These effects occur across multiple classifiers and are not reflected in overall accuracy, making them hard to detect without detailed subgroup analysis.
arXiv:2607. 28608v1 Announce Type: new Abstract: Clinical risk models routinely achieve strong aggregate performance while producing materially different error rates across patient subgroups.
The paper investigates how uncertainty quantification fails in selective prediction systems when clinical data exhibit severe class imbalance, leading to systematic exclusion of minority patients. Through experiments on 56 synthetic scenarios and three neurodegenerative disease cohorts, the authors show that standard uncertainty measures become proxies for class membership rather than true ambiguity, and that combining class‑aware correction (CCR) with threshold‑aware metrics restores proper monotonicity and protects minority patients. The study proposes adopting threshold‑aligned uncertainty measures and class‑stratified evaluation to mitigate this bias and improve safety in clinical AI pipelines.