Large Language Models for Psychiatric Diagnosis Based on Multicenter Real-World Clinical Records: Comparative Study
LLMs-especially GPT-4.0-demonstrate promising capability in supporting psychiatric diagnosis using real-world EHRs. However, diagnostic performance varies by age group and disorder category. LLMs should be regarded as assistive tools rather than replacements for clinical judgment, and further validation is needed before routine clinical implementation.
Authors
- Jia Yu (ORCID: https://orcid.org/0000-0001-9423-485X)
- Huaili Deng
- Yunsheng Yang (ORCID: https://orcid.org/0000-0003-0361-7674)
- Jun Feng (ORCID: https://orcid.org/0000-0002-9790-5353)
- Jiaquan Liang (ORCID: https://orcid.org/0000-0003-4583-0203)
- Min Sun (ORCID: https://orcid.org/0000-0001-7527-6712)
- Tao Xiao (ORCID: https://orcid.org/0000-0003-2331-2034)
- Guoping Huang (ORCID: https://orcid.org/0000-0002-4081-8398)
- Zebo Long
Institutions
- Mianyang Third People's Hospital (CN)
- Ganzhou People's Hospital (CN)
- First People's Hospital of Foshan (CN)
- Taiyuan Iron and Steel Group (China) (CN)
- Foshan Second People's Hospital (CN)
- Chongqing Emergency Medical Center (CN)
- Panzhihua Central Hospital (CN)
- Third People's Hospital of Huzhou (CN)
- Changzhou Third People's Hospital (CN)
- Chongqing Medical University (CN)
Publication Details
- Journal
- JMIR Medical Informatics
- Published
- 2026-01-13
- DOI
- https://doi.org/10.2196/77699
- Primary Topic
- Machine Learning in Healthcare
- Type
- article
- Field-Weighted Citation Impact
- 0.00