Methodological Considerations on Constructing Intelligent Diagnosis and Treatment Agent for Integrated Chinese and Western Medicine Diagnosis and Treatment Based on Clinical Practice Guidelines
10.13288/j.11-2166/r.2026.17.009
- VernacularTitle:基于临床实践指南的中西医诊疗智能体构建的方法学思考
- Author:
Feng ZHOU
1
;
Tengfei CHEN
2
;
Wandi ZHANG
1
;
Haoyuan LI
1
;
Xingyu ZONG
1
;
Jiahao LIN
1
;
Qingquan LIU
2
;
Guozhen ZHAO
1
Author Information
1. Institute of Basic Research in Clinical Medicine，China Academy of Chinese Medical Sciences，Beijing，100700
2. Beijing Hospital of Traditional Chinese Medicine，Capital Medical University
- Publication Type:Journal Article
- Keywords:
integrated Chinese and western medicine diagnosis and treatment;
intelligent agent;
clinical practice guidelines;
knowledge graph;
methodology
- From:
Journal of Traditional Chinese Medicine
2026;67(17):1853-1857
- CountryChina
- Language:Chinese
-
Abstract:
Developing an intelligent agent for integrated traditional Chinese and western medicine diagnosis and treatment based on clinical practice guidelines is a key approach to advancing the standardization and intelligent deve-lopment of traditional Chinese medicine （TCM） and to supporting clinical decision-making. This paper argues for the necessity of building a knowledge base using guidelines as the sole data source， systematically identifies five core methodological challenges across the entire process， and proposes corresponding solutions. An ontology framework capable of distinguishing between TCM and western medicine concepts should be designed to address the difficulty of knowledge integration. A closed-loop "algorithm-based extraction-expert review" model is adopted to reduce data extraction errors. Expert consensus is incorporated to address decision gaps in specific clinical scenarios outlined in the guidelines， while a multi-stage， standardized output process for the AI system is established to eliminate model hallucinations. Furthermore， an evaluation framework reflecting clinical applicability is developed to enhance the AI system's accuracy and stability. Throughout the research process， it is essential to ensure the in-depth and continuous participation of multidisciplinary experts， strictly control the selection and quality evaluation of guidelines， and rely on the expert consensus method to address decision gaps in the knowledge base. Additionally， continuous validation and iterative optimization of intelligent agents are required to ensure the accuracy and stability of generated outputs. Limitations remain in the mechanisms for real-time knowledge synchronization in intelligent diagnostic and treatment systems， and further validation through large-scale real-world studies is warranted.
