Performance of Foundation Models vs Physicians in Textual and Multimodal Ophthalmological Questions

Results of this cross-sectional study suggest that for textual questions, current FMs exhibited notable improvements in ophthalmological knowledge reasoning when compared with older LLMs and ophthalmology trainees, with performance comparable with that of expert ophthalmologists. These models demonstrated potential for medical assistance for answering ophthalmological textual queries, but their multimodal abilities remain limited. Further research or fine-tuning models with diverse ophthalmic multimodal data may lead to more capable applications with multimodal functionalities.

Authors

Institutions

Publication Details

Journal
JAMA Ophthalmology
Published
2025-11-13
DOI
https://doi.org/10.1001/jamaophthalmol.2025.4255
Primary Topic
Multimodal Machine Learning Applications
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Performance of Foundation Models vs Physicians in Textual and Multimodal Ophthalmological Questions

Matthew Azzopardi, Marcus Ang, Darren Shu Jeng Ting, Saaeha Rauz et al.
JAMA Ophthalmology
Multimodal Machine Learning Applications
article

Performance of Foundation Models vs Physicians in Textual and Multimodal Ophthalmological Questions

Matthew Azzopardi, Marcus Ang, Darren Shu Jeng Ting, Saaeha Rauz, Arun James Thirunavukarasu, Yin‐Hsi Chang, Sean X. Zhou, Daniel Shu Wei Ting, Andrew Malem, Anna V. Song, Benjamin Kye Jyn Tan, Jodhbir S. Mehta, Shiao Wei Wong, Humberto Rocha, Yu Jeat Chong, Yee Ling Wong, Nikhil Jain, Ting Fang Tan
article en

Abstract

Results of this cross-sectional study suggest that for textual questions, current FMs exhibited notable improvements in ophthalmological knowledge reasoning when compared with older LLMs and ophthalmology trainees, with performance comparable with that of expert ophthalmologists. These models demonstrated potential for medical assistance for answering ophthalmological textual queries, but their multimodal abilities remain limited. Further research or fine-tuning models with diverse ophthalmic multimodal data may lead to more capable applications with multimodal functionalities.

JAMA OphthalmologyVol. 144(1)
Moorfields Eye Hospital NHS Foundation Trust (GB), University of Nottingham (GB), Guy's and St Thomas' NHS Foundation Trust (GB), Singapore General Hospital (SG), University of Cambridge (GB), Singapore National Eye Center (SG), Norfolk and Norwich University Hospitals NHS Foundation Trust (GB), Luton and Dunstable University Hospital NHS Foundation Trust (GB), Manchester Royal Eye Hospital (GB), Chang Gung Memorial Hospital (TW), Cleveland Clinic Lerner College of Medicine (US), University of Oxford (GB), London School of Hygiene & Tropical Medicine (GB), Singapore Eye Research Institute (SG), Birmingham and Midland Eye Centre (GB), Duke-NUS Medical School (SG), Moorfields Eye Hospital (GB), Centre for Inflammation Research (GB), University College London (GB)
Openalex Percentile: Top 65%
Multimodal Machine Learning Applications
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.