Robust Uncertainty-Informed Glaucoma Classification Under Data Shift.

Journal: Translational vision science & technology
Published Date: Jun 2, 2025

Abstract

PURPOSE: Standard deep learning (DL) models often suffer significant performance degradation on out-of-distribution (OOD) data, where test data differs from training data, a common challenge in medical imaging due to real-world variations.

Authors

  • Homa Rashidisabet
    Department of Biomedical Engineering, University of Illinois Chicago, Chicago, IL, USA.
  • R V Paul Chan
    Ophthalmology, Illinois Eye and Ear Infirmary, Chicago, IL, United States.
  • Yannek I Leiderman
    Department of Ophthalmology and Visual Sciences - Illinois Eye and Ear Infirmary, University of Illinois at Chicago, Chicago, Illinois; Richard and Loan Hill Department of Biomedical Engineering, University of Illinois at Chicago, Chicago, Illinois. Electronic address: yannek@uic.edu.
  • Thasarat Sutabutr Vajaranant
    Artificial Intelligence in Ophthalmology (Ai-O) Center, University of Illinois Chicago, Chicago, IL, USA.
  • Darvin Yi
    Stanford University, Department of Radiology, Stanford, CA.

Keywords

External Resources

Popular Topics
Recent Journals