人工智能应用于中医药临床研究的现状

Current status of artificial intelligence in clinical research of traditional Chinese medicine

目的梳理人工智能在中医药临床研究中的应用现状,为充分发挥人工智能在中医药研究领域的作用提供参考依据。方法检索中国知网、万方、中国科技期刊数据库、PubMed和Web of Science核心数据库,收集公开发表的针对人工智能在中医药临床研究中应用的文献,进行分析。结果最终纳入32篇文献,均来自中国,发文年限在2014—2023年。研究主要来自高校和医疗机构。纳入的文献均为观察性研究,共包含病例47564例。人工智能在中医药临床研究中的应用包含10个方面,排名前三的分别为:(1)根据中医证型预测疾病发生风险,预测准确率均在80%以上;(2)预测疾病的中医证型,预测准确率高达99.16%;(3)利用人工智能探索疾病的中医证型与理化指标的关联,准确率最高达96.4%。其余7个分别为利用人工智能挖掘并构建了舌图像与疾病的关联模型、辅助诊断、根据中医人格及体质预测疾病发生风险、预测疾病预后、根据中医脉象信号预测疾病发生风险、预测中成药不良反应发生率。结论人工智能在中医药临床研究中赋能效果明显,可有效提高患者的诊疗质量,避免医疗资源的浪费。

Objective To analyze and sort out the current situation and effects of the application of artificial intelligence(AI)in the clinical research of traditional Chinese medicine(TCM),and put forward corresponding suggestions to provide reference for giving full play to the role of artificial intelligence in the field of TCM research.Methods The databases of CNKI,Wanfang,VIP,PubMed and Web of Science were searched to collect publicly published relevant researches on the application of AI in the clinical research of TCM for scoping review.Results A total of 32 studies were finally included,all of them were from China,and the publication years were between 2014 and 2023.The studies mainly came from universities,followed by medical institutions.The included studies were all observational studies,totaling 47,564 cases.The application of AI in the clinical research of TCM encompasses 10 aspects,and the top 3 are:(1)Predicting the risk of disease occurrence based on TCM syndromes,with a prediction accuracy of over 80%;(2)Predicting the TCM syndrome of the disease,with a prediction accuracy as high as 99.16%;(3)Using artificial intelligence to explore the association between the TCM syndrome of the disease and physicochemical indicators,with an accuracy up to 96.4%.The remaining seven are using AI to mine and construct the association model between tongue images and diseases,assisting in diagnosis,predicting the risk of disease occurrence based on TCM personality and constitution,predicting disease prognosis,predicting the risk of disease occurrence based on TCM pulse signals,and predicting the incidence of adverse reactions to Chinese patent medicines.Conclusions The effect of AI in the clinical research of TCM is obvious.After being applied in clinical practice,it can effectively improve the quality of diagnosis and treatment for patients.