EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Uncertainty-Aware AI-Assisted Diabetic Retinopathy Grading from Fundus Images with Ordinal Conformal Prediction

Umar Hasan iD, Muhammad Ali Nayeem iD, Turki G Alghamdi iD

Diagnostics · 19.08.2026 · 16 (16) · 2645

DOI: 10.3390/diagnostics16162645 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Diagnostics, online erschienen am 19.08.2026. Verfasst von Umar Hasan und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Mathematics and Physics, School of Engineering and Physical Sciences, North South University, Dhaka 1229, Bangladesh. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 50 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Qassim University.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Artificial intelligence (AI) systems for diabetic retinopathy (DR) grading require reliable uncertainty estimates when applied to fundus images outside the development dataset. We evaluated whether conformal prediction can provide structured set-valued outputs and whether internal uncertainty calibration remains reliable during external evaluation.

Methods

EfficientNet-B0, ResNet-50, and Vision Transformer (ViT-Base) classifiers were trained on APTOS 2019. Split-conformal predictors were calibrated exclusively on held-out APTOS images using three categorical scores, LAC, APS, and RAPS, and an ordinal score restricted to adjacent severity grades. Performance was assessed internally on APTOS and externally on IDRiD, with an independent five-seed ViT replication extending evaluation to Messidor-2, at target coverages of 90% and 95%.

Results

Coverage was approximately nominal internally but decreased on both external datasets. On IDRiD, the largest deficit occurred for severe DR (grade 3). The ordinal method produced contiguous intervals in 100% of cases and achieved the highest grade-3 coverage in every tested backbone–risk configuration. For ResNet-50 at 95% target coverage, grade-3 coverage increased from 0.750 with APS to 0.945 with the ordinal method, while average set size increased from 2.89 to 3.03. In the independent ViT replication on Messidor-2, ordinal marginal coverage exceeded APS at both targets (0.676 versus 0.632 and 0.741 versus 0.714).

Conclusions

Internal calibration did not ensure reliable class-specific uncertainty during external evaluation. Ordinal prediction sets improved structural coherence and mitigated severe-grade undercoverage, but did not restore formal coverage guarantees after dataset shift.

Schlagwörter

Stichwörter der Autor:innen

artificial intelligenceconformal predictiondataset shiftdeep learningdiabetic retinopathydiagnostic supportexternal validationfundus photographyordinal classificationuncertainty quantification

Autor:innen und Institutionen

  • Department of Mathematics and Physics, School of Engineering and Physical Sciences, North South University, Dhaka 1229, Bangladesh.
  • Department of Computer Engineering, College of Computer, Qassim University, Buraydah 51452, Saudi Arabia.
  • Department of Computer Engineering and Networks, College of Computer and Information Sciences, Jouf University, Sakaka 72388, Saudi Arabia.

ORCID-Profile: Umar Hasan, Muhammad Ali Nayeem, Turki G Alghamdi

Förderung

  • Qassim University

Zitieren

APA

Hasan, U., Nayeem, M. A., & Alghamdi, T. G. (2026). Uncertainty-Aware AI-Assisted Diabetic Retinopathy Grading from Fundus Images with Ordinal Conformal Prediction. Diagnostics, 16(16), 2645. https://doi.org/10.3390/diagnostics16162645

Vancouver

Hasan U, Nayeem MA, Alghamdi TG. Uncertainty-Aware AI-Assisted Diabetic Retinopathy Grading from Fundus Images with Ordinal Conformal Prediction. Diagnostics. 2026;16(16):2645. doi:10.3390/diagnostics16162645

BibTeX
@article{Hasan2026,
  author = {Hasan, Umar and Nayeem, Muhammad Ali and Alghamdi, Turki G},
  title = {Uncertainty-Aware AI-Assisted Diabetic Retinopathy Grading from Fundus Images with Ordinal Conformal Prediction},
  journal = {Diagnostics},
  year = {2026},
  volume = {16},
  number = {16},
  pages = {2645},
  doi = {10.3390/diagnostics16162645}
}
Kostenlos starten →