Comparative evaluation of large language models and clinicians in real-world glaucoma clinical reasoning
Graefe's Archive for Clinical and Experimental Ophthalmology · 06.08.2026
DOI: 10.1007/s00417-026-07442-7 ↗Kurz gefasst
Fachartikel in der Zeitschrift Graefe's Archive for Clinical and Experimental Ophthalmology, online erschienen am 06.08.2026. Verfasst von Houfa Yin und 5 weiteren Autor:innen. Beteiligte Einrichtung: Eye Center of Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, China. Die Arbeit nennt 25 Literaturangaben. Gefördert von: Natural Science Foundation of China, Education Reform Project of the Second Affiliated Hospital Zhejiang University School of Medicine.
Abstract (Originalsprache)
What is known Artificial intelligence has shown strong performance in image-based glaucoma tasks, but evidence for specialty-specific glaucoma clinical reasoning remains limited. Direct case-based comparisons between large language models and clinicians at different training stages remain scarce. What is new In this 34-case exploratory evaluation, AI model scores overlapped with those of attending ophthalmologists but should not be interpreted as proof of clinical equivalence. AI outputs were relatively structured and often included safety-critical diagnostic and management points, whereas human performance varied across clinicians. The findings support cautious study of LLMs as supervised decision-support and educational tools, not as replacements for specialist judgement.
Purpose
Clinical decision-making in glaucoma is complex and requires integration of heterogeneous information, including patient history, examination findings, and risk stratification. While artificial intelligence (AI) has shown strong performance in image-based ophthalmic tasks, its capability in specialty-specific clinical reasoning remains insufficiently explored.
Methods
Performance was evaluated by glaucoma specialists using a predefined rubric across three clinically oriented domains: medical accuracy (40%), key-point recall (30%), and logical completeness (30%). The weighted composite score was used as a descriptive summary of case-based reasoning quality.
Results
AI models showed structured clinical reasoning performance in this case-based dataset, with weighted mean scores overlapping with those of attending ophthalmologists and exceeding those of some lower-performing trainees. These findings should be interpreted as exploratory performance patterns rather than evidence of equivalence. Inter-individual variability was substantial among human clinicians, particularly residents. AI systems often included safety-critical diagnostic and management elements, while the best-performing human clinician achieved the highest individual score overall.
Conclusion
In this limited 34-case evaluation, large language model-based AI systems produced structured glaucoma-related reasoning with performance that overlapped with attending ophthalmologists but did not establish clinical equivalence. These systems require specialist oversight and further validation before clinical use, but may have potential as supervised decision-support and educational tools.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Eye Center of Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, China.
- Zhejiang Provincial Key Laboratory of Ophthalmology, Zhejiang Provincial Clinical Research Center for Eye Diseases, Zhejiang Provincial Engineering Institute on Eye Diseases, Hangzhou, China.
- Department of Ophthalmology, University of Warmia and Mazury, Olsztyn, Poland.
- Institute for Research in Ophthalmology, Foundation for Ophthalmology Development, Poznan, Poland.
- Eye Center of Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, China. jinkai@zju.edu.cn.
- Zhejiang Provincial Key Laboratory of Ophthalmology, Zhejiang Provincial Clinical Research Center for Eye Diseases, Zhejiang Provincial Engineering Institute on Eye Diseases, Hangzhou, China. jinkai@zju.edu.cn.
Förderung
- Natural Science Foundation of China
- Education Reform Project of the Second Affiliated Hospital Zhejiang University School of Medicine
Zitieren
APA
Yin, H., Shen, L., Cai, H., Wu, W., Grzybowski, A., & Jin, K. (2026). Comparative evaluation of large language models and clinicians in real-world glaucoma clinical reasoning. Graefe's Archive for Clinical and Experimental Ophthalmology. https://doi.org/10.1007/s00417-026-07442-7
Vancouver
Yin H, Shen L, Cai H, Wu W, Grzybowski A, Jin K. Comparative evaluation of large language models and clinicians in real-world glaucoma clinical reasoning. Graefe's Archive for Clinical and Experimental Ophthalmology. 2026. doi:10.1007/s00417-026-07442-7
BibTeX
@article{Yin2026,
author = {Yin, Houfa and Shen, Lixia and Cai, Haiyan and Wu, Wei and Grzybowski, Andrzej and Jin, Kai},
title = {Comparative evaluation of large language models and clinicians in real-world glaucoma clinical reasoning},
journal = {Graefe's Archive for Clinical and Experimental Ophthalmology},
year = {2026},
doi = {10.1007/s00417-026-07442-7}
}
EUVIMED