| Title |
|---|
| The Application Efficacy of Large Language Models in Health Consultation for Glaucoma Patients |
| Authors |
|---|
| huan xu |
| Presenting |
|---|
| huan xu |
| PURPOSE: |
|---|
| To evaluate the effectiveness and application potential of different large language models in health education consultation for glaucoma patients |
| METHODS: |
|---|
| The DeepSeek R1, Wenxin Yiyan X1, Doubao, and KIMI large language models were tested for 3 rounds using 18 common clinical issues of glaucoma and 2 typical clinical scenarios. Three ophthalmologists with senior titles were selected to evaluate the accuracy, comprehensiveness, understandability,humanistic care, and case analysis ability from five dimensions |
| RESULTS: |
|---|
| The overall performance of each model was good, with DeepSeek R1 scoring the highest, followed by KIMI2.5, Doubao, and Wenxin Yiyan X1. DeepSeek R1 and KIMI2.5 excelled in complex clinical reasoning, evidence-based medical knowledge transfer, humanistic care, and case analysis ability, while Doubao and Wenxin Yiyan X1 had advantages in Chinese context adaptability |
| CONCLUSIONS: |
|---|
| Large language models show positive auxiliary application value in the field of glaucoma health education, but they need to be combined with clinical review mechanisms to improve the reliability of information, and promote its clinical application by optimizing algorithm transparency and localization adaptability. |