Health AI Central
Peer-Reviewed Publication
NPJ Digit Med2025;8(1):532.August 19, 2025Journal Article

Specialized curricula for training vision language models in retinal image analysis.

Robbie Holland1, Thomas R P Taylor2, Christopher Holmes3, Sophie Riedl4,5, Julia Mai4,5, Maria Patsiamanidi2, Dimitra Mitsopoulou2, Paul Hager5, Philip Müller5, Johannes C Paetzold6,7, Hendrik P N Scholl8,9,10, Hrvoje Bogunović11, Ursula Schmidt-Erfurth4, Daniel Rueckert6,7,12, Sobha Sivaprasad3, Andrew J Lotery2, Martin J Menten6,7,12,
1Biomedical Image Analysis, Department of Computing, Imperial College London, London, United Kingdom. robbie.holland@stanford.edu.
2Clinical and Experimental Sciences, Faculty of Medicine, University of Southampton, Southampton, United Kingdom.
3Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom.
4Ophthalmic Image Analysis Group (OPTIMA), Medical University of Vienna, Vienna, Austria.
5Department of Ophthalmology and Optometry, Medical University of Vienna, Vienna, Austria.
6Biomedical Image Analysis, Department of Computing, Imperial College London, London, United Kingdom.
7Chair for AI in Healthcare and Medicine, Technical University of Munich, Munich, Germany.
8Institute of Molecular and Clinical Ophthalmology Basel, Basel, Switzerland.
9Department of Ophthalmology, University of Basel, Basel, Switzerland.
10Department of Clinical Pharmacology, Medical University of Vienna, Vienna, Austria.
11Institute of Artificial Intelligence, Centre for Medical Data Science, Medical University of Vienna, Vienna, Austria.
12Munich Center for Machine Learning, Munich, Germany.
PubMed PMC DOI PDF

Abstract

Clinicians spend significant time reviewing medical images and transcribing findings. By integrating visual and textual data, foundation models have the potential to reduce workloads and boost efficiency, yet their practical clinical value remains uncertain. In this study, we find that OpenAI's ChatGPT-4o and two medical vision-language models (VLMs) significantly underperform ophthalmologists in…

Create a free account to keep reading

Free members get 10 full research views every month across publications, clinical trials, FDA clearances, adverse events, and NIH grants. No credit card required.

Create Free AccountSign In

Want unlimited research access? See Pro plans

Data Accuracy Notice: Research intelligence on Health AI Central is aggregated from public sources (PubMed, ClinicalTrials.gov, FDA, NIH, CMS, and others) and refreshed nightly. Classifications and derived metrics are produced by automated methods described in our Methodology. We recommend verifying critical data points against the primary sources before making decisions.