Subgroup performance analysis of adaptation strategies for chest X-ray foundation models

• URL: http://arxiv.org/abs/2608.19078v1
• Date: Wed, 19 Aug 2026 16:25:44 GMT
• Title: Subgroup performance analysis of adaptation strategies for chest X-ray foundation models
• Abstract summary: We use MIMIC-CXR to evaluate eight pathologies across race, sex, and imaging-view subgroups on a prevalence-preserving, demographically balanced test set.<n>We find that attention pooling achieves the strongest overall discriminative performance and encodes attributes, particularly race, most strongly, but that improved overall performance does not consistently reduce subgroup disparities.
• Score: 12.306643760835229
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Foundation models are increasingly adapted for downstream medical imaging tasks, yet the influence of the chosen adaptation strategy on subgroup fairness remains poorly understood. We investigate how three parameter-efficient adaptation techniques, including linear heads on the raw CLS token, an MLP, and an attention-pooling module over multi-layer patch features, affect both pathology classification performance and subgroup disparities when applied to the frozen Rad-DINO chest X-ray encoder. Using MIMIC-CXR, we evaluate eight pathologies across race, sex, and imaging-view subgroups on a prevalence-preserving, demographically balanced test set, and additionally probe how strongly each adapter encodes protected attributes. We find that attention pooling achieves the strongest overall discriminative performance and encodes attributes, particularly race, most strongly, but that improved overall performance does not consistently reduce subgroup disparities. Notably, stronger attribute encoding did not correspond to larger disparities: early network layers encoded race most weakly yet produced the largest subgroup performance gaps. Exploring different attention-pooling layer combinations further revealed no consistent relationship between the layers pooled, attribute encoding strength, and subgroup fairness. Our results indicate that richer, more expressive representations can improve accuracy while leaving fairness implications task-dependent and unpredictable, which must be assessed directly and per-task rather than inferred from encoding strength or overall performance alone.

Related papers

• Evolution-Aware MSA Reasoning for Subsampling via Factor Graphs [55.715754117956514]
  Multiple Sequence Alignments provide protein language models with explicit evolutionary context.<n>MSA subsampling is unavoidable under limited token budgets.<n>We introduce AP-REASONER, which infers a fixed-budget MSA subset using factor-graph reasoning.
  arXiv  Detail & Related papers  (2026-07-24T13:55:21Z)
• PulmoSight-XAI: An Explainable Multi-View Attention Ensemble with Gradient Boosting Meta-Learning for Multi-Label Chest X-Ray Classification [1.2744523252873352]
  We propose an explainable hierarchical multi-view ensemble framework for the robust classification of 14 thoracic pathologies.<n>The framework employs view-specific training by independently modeling frontal and lateral radiographs.<n>The framework achieves state-of-the-art macro-average AUROC scores of 0.9319 for frontal and 0.9154 for lateral radiographs.
  arXiv  Detail & Related papers  (2026-07-05T19:50:53Z)
• Exploring Entropy-based Active Learning for Fair Brain Segmentation [9.205575630735263]
  Active learning has emerged as a crucial strategy for reducing prohibitive costs associated with medical image segmentation.<n>We introduce a fairness-aware active learning framework with a Weighted Entropy selection strategy that modulates uncertainty based on current group-specific performance estimates on the labeled set.<n>Experiments demonstrated that our method markedly reduces performance disparities between groups compared to random sampling and standard uncertainty sampling.
  arXiv  Detail & Related papers  (2026-05-03T04:21:48Z)
• Component-Adaptive and Lesion-Level Supervision for Improved Small Structure Segmentation in Brain MRI [0.0]
  We propose a unified objective function, termed CATMIL, that augments the base segmentation loss with two auxiliary supervision terms operating at different levels.<n>We evaluate the proposed objective on the MSLesSeg dataset using a consistent nnU-Net framework and 5-fold cross-validation.<n>Results show that CATMIL achieves the most balanced performance across segmentation accuracy, lesion detection, and error control.
  arXiv  Detail & Related papers  (2026-04-09T09:15:10Z)
• Reducing Class-Wise Performance Disparity via Margin Regularization [82.81746960548382]
  Deep neural networks often exhibit substantial disparities in class-wise accuracy, even when trained on class-balanced data.<n>We present Margin Regularization for Performance Disparity Reduction (MR$2$), a theoretically principled regularization for classification.<n>Our analysis reveals how per-class feature variability contributes to error, motivating the use of larger margins for hard classes.
  arXiv  Detail & Related papers  (2026-01-30T12:56:08Z)
• SG-CLDFF: A Novel Framework for Automated White Blood Cell Classification and Segmentation [0.0]
  Saliency-Guided Cross-Layer Deep Feature Fusion framework (SG-CLDFF)<n>A lightweight hybrid backbone (Swin-style) produces multi-resolution representations, which are fused by a ResNeXt-CCinspired cross-layer fusion module.<n>Interpretability is enforced through Grad-CAM visualizations and saliency consistency checks, allowing model decisions to be inspected at the regional level.
  arXiv  Detail & Related papers  (2025-10-20T08:07:39Z)
• Addressing Bias in VLMs for Glaucoma Detection Without Protected Attribute Supervision [6.381985108365988]
  This work focuses on automated glaucoma screening from retinal fundus images.<n>We introduce an attribute-agnostic debiasing method that infers proxy subgroups via unsupervised clustering of image-image embeddings.<n>We evaluate our method on the Harvard FairVLMed glaucoma subset, reporting Equalized Odds Distance (EOD), Equalized Subgroup AUC (ES AUC), and Groupwise AUC to demonstrate equitable performance across inferred demographic subgroups.
  arXiv  Detail & Related papers  (2025-08-12T17:07:58Z)
• Class-attribute Priors: Adapting Optimization to Heterogeneity and Fairness Objective [54.33066660817495]
  Modern classification problems exhibit heterogeneities across individual classes. We propose CAP: An effective and general method that generates a class-specific learning strategy. We show that CAP is competitive with prior art and its flexibility unlocks clear benefits for fairness objectives beyond balanced accuracy.
  arXiv  Detail & Related papers  (2024-01-25T17:43:39Z)
• Balanced Classification: A Unified Framework for Long-Tailed Object Detection [74.94216414011326]
  Conventional detectors suffer from performance degradation when dealing with long-tailed data due to a classification bias towards the majority head categories. We introduce a unified framework called BAlanced CLassification (BACL), which enables adaptive rectification of inequalities caused by disparities in category distribution. BACL consistently achieves performance improvements across various datasets with different backbones and architectures.
  arXiv  Detail & Related papers  (2023-08-04T09:11:07Z)
• ARHNet: Adaptive Region Harmonization for Lesion-aware Augmentation to Improve Segmentation Performance [61.04246102067351]
  We propose a foreground harmonization framework (ARHNet) to tackle intensity disparities and make synthetic images look more realistic. We demonstrate the efficacy of our method in improving the segmentation performance using real and synthetic images.
  arXiv  Detail & Related papers  (2023-07-02T10:39:29Z)
• Fairness via Adversarial Attribute Neighbourhood Robust Learning [49.93775302674591]
  We propose a principled underlineRobust underlineAdversarial underlineAttribute underlineNeighbourhood (RAAN) loss to debias the classification head.
  arXiv  Detail & Related papers  (2022-10-12T23:39:28Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.