Article

Does Model Compression Amplify Clinical Bias? A Subgroup Fairness Audit of Quantized ICU Risk Models on MIMIC-IV

2026-07-30

Abstract excerpt

<title>Abstract</title> <p>Model compression is a prerequisite for deploying clinical risk-prediction models on resource-constrained hardware, but its effect on subgroup fairness is rarely audited alongside its effect on aggregate accuracy. We train a compact multilayer perceptron (5,313 parameters) to predict 72-hour acute kidney injury (AKI) onset in a MIMIC-IV cohort of 52,178 adult first ICU stays, audit its...

Topics

Open a Topic to create a Post that cites this publication.

Identifiers and source

Literature Corpus work: 1b8b4a17-ea48-5fb1-a07d-94090390ca91
DOI: 10.21203/rs.3.rs-10521968/v1

Open publication

Related research

Semantic proximity does not establish scientific evidence.

Click a neighbor to travelStep 1 · 12 closest

Does Model Compression Amplify Clinical Bias? A Subgroup Fairness Audit of Quantized ICU Risk Models on MIMIC-IVDOI 10.21203/rs.3.rs-10521968/v1Read

A Simulation-Based Validation Framework for Uncertainty-Aware Clinical Triage: Conformal Prediction and Ensemble Learning Applied to a Synthetic ICU Benchmark CohortSemantic proximity Read

Trustworthy LLM-Embedding Clinical Prediction: Calibrating Confidence and Transparency for Foundation Model-Based Disease Risk ScoresSemantic proximity Read

A Reproducible Clinical Decision-Support Suite on MIMIC-IVSemantic proximity Read

Deployment-readiness audit of calibration, clinical utility, and fairness in perioperative infection predictionSemantic proximity Read

The Deferral Frontier: Reliability-Gated Selective Prediction for Clinical AISemantic proximity Read

MIMIC-IF: Interpretability and Fairness Evaluation of Deep Learning Models on MIMIC-IV DatasetSemantic proximity Read

Unsupervised Cluster-Augmented Mortality Prediction in MIMIC-IV Using Tabular ICU Time-Series Aggregation and Gradient Boosted LearningSemantic proximity Read

Ensemble Post-hoc Explainable AI for Multilead ECG: Identifying Disease-Relevant Features in Single-Lead InterpretationsSemantic proximity Read

Pretraining Diversity and Clinical Metric Optimization Achieve State-of-the-Art Performance on ChestX-ray14Semantic proximity Read

Outcome Prediction Models for Critically Ill Patients Using Small Routine Laboratory DatasetsSemantic proximity Read

Perioperative Mortality Prediction Using a Prevalence-Adaptive Four-Model Bayesian Ensemble with Entropy-Based Uncertainty TriageSemantic proximity Read

A Simple, Privacy-Preserving Multimodal Stacking System for Clinical Progression Prediction in Low-Resource Rare DiseasesSemantic proximity Read