中医智能推理辨证模型探索
Exploration of Intelligent Inference Model for Syndrome Differentiations of Traditional Chinese Medicine
江启煜 1孙晓生 1谢波 2罗广波2
作者信息
- 1. 广州中医药大学,广州 510405
- 2. 广州中医药大学第一附属医院,广州 510405
- 折叠
摘要
虽然中医人工智能辨证诊断技术不断发展,但现有的智能辨证模型只能输出辨证诊断结果,无法展现辨证推理过程,不具有中医诊断学知识的可解释性,难以真正体现中医的辨证思维过程,这可能是困惑已久的难题.本文阐述了中医智能辨证诊断的发展历程与技术特点,探讨了当前中医智能辨证诊断研究在中医辨证思维推理解析问题上遇到的关键技术及瓶颈,提出了一种可推理解释中医辨证思维过程的新型智能辨证诊断模型架构,并阐述了模型与算法的实现过程,以及进行了详细的数据实例分析.本文将为中医的智能辨证发展提供了新的方法,为中医临床辅助决策打开了新的研究思路.
Abstract
Although the artificial intelligence diagnosis technology of traditional Chinese medicine(TCM)is constantly developing,the existing intelligent syndrome differentiation models can only output the diagnosis results and cannot display the reasoning process of TCM diagnosis.They do not have the interpretability of TCM diagnostic knowledge and are difficult to truly reflect the process of TCM's syndrome differentiation.This can be a long-standing problem that has puzzled the academic community.This article elaborates on the development process and technical characteristics of intelligent diagnosis in traditional Chinese medicine,explores multiple key technologies and bottlenecks in current research in reasoning of TCM's Syndrome Differentiation,and factors affecting diagnosis.A novel intelligent diagnosis model architecture that can infer and explain the process of TCM'ssyndrome differentiation was proposed,and the implementation process of the model and algorithm was explained,and detailed data instance analysis was conducted.This article will provide new methods for the development of intelligent syndrome differentiation of TCM and open up new research ideas for clinical decision-making assistance of TCM.
关键词
人工智能/辨证/中医诊断/推理/ChatGPT/深度学习/语言大模型Key words
Artificial intelligence/Syndrome differentiation/TCM diagnosis/Inference/ChatGPT/Deep learning/Largelanguage model引用本文复制引用
基金项目
国家中医药管理局(国中医药人教函[2022]75号)
出版年
2024
NETL
NSTL
万方数据