Evaluating large language model clinical reasoning in glaucoma using retrieval-augmented generation
Houfa Yin, Qi Miao, Wanshu Zhou, Chenyang Hu, Yongwei Guo, Lixia Shen, Haiyan Cai, Andrzej Grzybowski, Kai Jin · 2026
Read Evaluating large language model clinical reasoning in glaucoma using retrieval-augmented generation by Houfa Yin, Qi Miao, Wanshu Zhou, Chenyang Hu, Yongwei Guo, Lixia Shen, Haiyan Cai, Andrzej Grzybowski, Kai Jin (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13522369.txt.
Description
Houfa Yin and colleagues tested whether giving chatbots access to medical sources sharpens their reasoning about glaucoma care. Published in 2026, the study used forty real glaucoma cases spanning early, secondary, post-surgery, and end-stage disease. Six setups were scored: three language models, each with and without retrieval-augmented generation, which pulls in guideline text before answering.
Models with retrieval beat their plain twins on accuracy, coverage, and logical completeness, with gains holding up under sensitivity checks. Yet agreement between the two masked specialist raters was limited, so scores call for consensus review and careful reading. Safety was audited apart from the main score. The authors back retrieval-backed models as supervised decision aids, not as stand-alone clinicians.
Catalog record
| CREATOR | Houfa Yin, Qi Miao, Wanshu Zhou, Chenyang Hu, Yongwei Guo, Lixia Shen, Haiyan Cai, Andrzej Grzybowski, Kai Jin |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 38 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |