mlsanityData to 14 Oct 2026

SUDO: a framework for evaluating clinical artificial intelligence systems without ground-truth annotations

arXiv 2403.17011 ↗First posted 2 Jan 2024cs.LG cs.AI cs.CYcited by 4

Not on the map. Why?

Abstract (arXiv)

A clinical artificial intelligence (AI) system is often validated on a held-out set of data which it has not been exposed to before (e.g., data from a different hospital with a distinct electronic health record system). This evaluation process is meant to mimic the deployment of an AI system on data in the wild; those which are currently unseen by the system yet are expected to be encountered in a clinical setting. However, when data in the wild differ from the held-out set of data, a phenomenon referred to as distribution shift, and lack ground-truth annotations, it becomes unclear the extent to which AI-based findings can be trusted on data in the wild. Here, we introduce SUDO, a framework for evaluating AI systems without ground-truth annotations. SUDO assigns temporary labels to data points in the wild and directly uses them to train distinct models, with the highest performing model indicative of the most likely label. Through experiments with AI systems developed for dermatology images, histopathology patches, and clinical reports, we show that SUDO can be a reliable proxy for model performance and thus identify unreliable predictions. We also demonstrate that SUDO informs the selection of models and allows for the previously out-of-reach assessment of algorithmic bias for data in the wild without ground-truth annotations. The ability to triage unreliable predictions for further inspection and assess the algorithmic bias of AI systems can improve the integrity of research findings and contribute to the deployment of ethical AI systems in medicine.

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Builds on 16 of its 38 references that are in the map. First posted on arXiv 2 Jan 2024.

Builds on

16 of 38 references.

References · 16

  1. 2017On Calibration of Modern Neural Networkscited by 4,960
  2. 2016Simple and Scalable Predictive Uncertainty Estimation using Deep Ensemblescited by 3,926
  3. 2016A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networkscited by 2,385
  4. 2016Equality of Opportunity in Supervised Learningcited by 2,007
  5. 2019A Survey on Bias and Fairness in Machine Learningcited by 1,917
  6. 2020WILDS: A Benchmark of in-the-Wild Distribution Shiftscited by 1,375
  7. 2019Can You Trust Your Model's Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shiftcited by 1,287
  8. 2021Disparities in Dermatology AI: Assessments Using Diverse Clinical Imagescited by 134
  9. 2022Uncertainty-Informed Deep Learning Models Enable High-Confidence Predictions for Digital Histopathologycited by 37
  10. 2017Reverse Classification Accuracy: Predicting Segmentation Performance in the Absence of Ground Truthcited by 35
  11. 2020Can I Trust My Fairness Metric? Assessing Fairness with Unlabeled Data and Bayesian Inferencecited by 31
  12. 2021Single Layer Predictive Normalized Maximum Likelihood for Out-of-Distribution Detectioncited by 23
4 more
  1. 2022Doubly Robust Augmented Model Accuracy Transfer Inference with High Dimensional Featurescited by 13
  2. 2017Semi-Supervised Approaches to Efficient Evaluation of Model Prediction Performancecited by 9
  3. 2015Assessing binary classifiers using only positive and unlabeled datacited by 2
  4. 2022Semi-supervised Transfer Learning for Evaluation of Model Classification Performancecited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 1; 2025: 2; 2026: 1.

4 papers.

  1. 2024Rethinking Scale: The Efficacy of Fine-Tuned Open-Source LLMs in Large-Scale Reproducible Social Science Researchcited by 6
  2. 2025Automated Real-time Assessment of Intracranial Hemorrhage Detection AI Using an Ensembled Monitoring Model (EMM)cited by 3
  3. 2025MARIA: A Framework for Marginal Risk Assessment without Ground Truth in AI Systemscited by 0
  4. 2026Label-Free Foundational Model Selection for Medical Image Classification under Distribution Shift via Pseudo Label Discrepancycited by 0