Copyright: ©Author(s) 2027. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution-NonCommercial (CC BY-NC 4.0) license. No commercial re-use. See permissions. Published by Baishideng Publishing Group Inc.
World J Gastroenterol. Jan 7, 2027; 33(1): 123087
Published online Jan 7, 2027. doi: 10.3748/wjg.123087
Large language models with supervised fine-tuning for inflammatory bowel disease diagnosis and prognosis
Shi-Ming Zhou, Ming-Jun Zhang, Shu-Lin Zhang, Tao Chen, Zheng-Jie Wei, Hai-Bin Dong, Qi Cui, Meng Li, Dan Wang, Hui-Chao Wang, Yi-Ru Guan, Hai-Hua Yang, Bahabaike Jiangtulu, Bo Cui, Qian Xu, Yi Zhao, Xuan Jiang
Shi-Ming Zhou, Ming-Jun Zhang, Shu-Lin Zhang, Tao Chen, Zheng-Jie Wei, Hai-Bin Dong, Qi Cui, Xuan Jiang, Department of Gastroenterology, Beijing Tsinghua Changgung Hospital, School of Clinical Medicine, Tsinghua Medicine, Tsinghua University, Beijing 102218, China
Meng Li, Dan Wang, Hui-Chao Wang, Yi-Ru Guan, Hai-Hua Yang, Bahabaike Jiangtulu, Bo Cui, Qian Xu, Xiaohe Health, Bytedance, Beijing 100192, China
Yi Zhao, Department of Dermatology and Venereology, Beijing Tsinghua Changgung Hospital, School of Clinical Medicine, Tsinghua Medicine, Tsinghua University, Beijing 102218, China
Co-first authors: Shi-Ming Zhou and Ming-Jun Zhang.
Co-corresponding authors: Yi Zhao and Xuan Jiang.
Author contributions: Zhou SM and Zhang MJ share co-first authorship based on their equivalent substantive contributions to study design, research execution, and manuscript drafting and revision; Chen T, Wei ZJ, Dong HB and Cui Q participated in data collection and model evaluation; Zhang SL was responsible for data collection and statistical analysis; Li M, Wang D, Wang HC, Guan YR, Yang HH, Jiangtulu B, Cui B and Xu Q assisted with data collection and model training; Zhao Y and Jiang X serve as co-corresponding authors, with equivalent contributions to the overall study design, research supervision, and manuscript revision; all authors have read and approve the final manuscript.
AI contribution statement: AI tools were used solely for linguistic refinement and formatting assistance. No AI tool was involved in the generation of research data, interpretation of results, or formulation of conclusions. All AI-generated outputs were critically reviewed and revised by the authors.
Institutional review board statement: The study was reviewed and approved by Beijing Tsinghua Changgung Hospital Institutional Review Board (approval No. 23695-4-01).
Informed consent statement: The data we used were all de-identified information provided by the hospital’s information department, which did not involve patient privacy or any other sensitive information, and no informed consent was required.
Conflict-of-interest statement: The authors declare that they have no conflict of interest.
STROBE statement: The authors have read the STROBE Statement—a checklist of items, and the manuscript was prepared and revised according to the STROBE Statement-a checklist of items.
Data sharing statement: The data that support the findings of this study are available from the corresponding author, upon reasonable request.
Corresponding author: Xuan Jiang, MD, Doctor, Department of Gastroenterology, Beijing Tsinghua Changgung Hospital, School of Clinical Medicine, Tsinghua Medicine, Tsinghua University, No. 168 Litang Road, Beijing 102218, China.
jxa01998@btch.edu.cn
Received: May 8, 2026
Revised: July 3, 2026
Accepted: August 28, 2026
Published online: January 7, 2027
Processing time: 185 Days and 20.4 Hours
BACKGROUND
Inflammatory bowel disease (IBD) is a chronic, incurable condition characterized by complex complications. Effective management of IBD demands extensive physician expertise and time, presenting major challenges due to limited experience and healthcare resources. Large language models (LLMs) may serve as efficient auxiliary tools to alleviate workforce pressures and enhance the accessibility and efficiency of healthcare services.
AIM
To evaluate the capabilities of an LLM trained by supervised fine-tuning (SFT) in the diagnosis and treatment of IBD.
METHODS
In this observational study, the diagnostic and treatment capabilities of a general LLM for IBD were evaluated using cases derived from real-world electronic health records. High-quality medical records were filtered for SFT to construct a dedicated IBD-LLM, followed by a direct head-to-head comparison of IBD care performance between the IBD-LLM and resident physicians.
RESULTS
The general LLM demonstrated suboptimal performance in IBD diagnosis and treatment, although its response consistency and harmlessness were confirmed. In the head-to-head comparison, IBD-LLM vs resident physicians achieved scores of 2.88 vs 3.47 out of 5 in diagnostic accuracy (P = 0.001), 4.18 vs 3.47 out of 5 in humanistic care (P = 0.001), 4.39 vs 3.41 out of 5 in clinical judgment (P = 0.001), 3.88 vs 3.53 out of 5 in communication skills (P = 0.002), and 4.11 vs 3.74 out of 5 in organizational effectiveness (P = 0.003). Although the IBD-LLM demonstrated performance comparable to resident physicians in three parameters, it showed significantly lower diagnostic accuracy. No safety risks or artificial intelligence hallucinations were observed in outputs from either the general LLM or IBD-LLM.
CONCLUSION
A specialized LLM can be efficiently trained with limited samples using SFT, and demonstrated significant potential for assisting with medical tasks under the supervision of attending physicians.
Core Tip: This study constructed a specialized inflammatory bowel disease-large language model (LLM) using supervised fine-tuning (SFT) on a limited number of high-quality data from real-world electronic health records. In a head-to-head comparison with resident physicians during simulated clinical interactions, the model excelled in humanistic care, clinical judgment, and organizational effectiveness, but demonstrated a significant deficit in diagnostic accuracy. Notably, no artificial intelligence hallucinations were observed, attributed to rigorous structured training and a conservative learning rate. These findings establish SFT-trained LLMs as valuable clinical adjuncts for enhancing patient interaction in resource-constrained settings, although complex diagnostic reasoning still mandates human oversight.
