Data-driven based four examinations in TCM:a survey

基于数据驱动的中医四诊综述

Traditional Chinese medicine(TCM)diagnosis is a unique disease diagnosis method with thousands of years of TCM theory and effective experience.Its thinking mode in the process is different from that of modern medicine,which includes the essence of TCM theory.From the perspective of clinical application,the four diagnostic methods of TCM,including inspection,auscultation and olfaction,inquiry,and palpation,have been widely accepted by TCM practitioners worldwide.With the rise of artificial intelligence(AI)over the past decades,AI based TCM diagnosis has also grown rapidly,marked by the emerging of a large number of data-driven deep learning models.In this paper,our aim is to simply but systematically review the development of the data-driven technologies applied to the four diagnostic approaches,i.e.the four examinations,in TCM,including data sets,digital signal acquisition devices,and learning based computational algorithms,to better analyze the development of AI-based TCM diagnosis,and provide references for new research and its applications in TCM settings in the future.

中医诊断是基于中医学几千年的理论与有效经验而发展成的独特的疾病诊断方法,其过程中的思维模式区别于现代医学,囊括了中医理论的精髓。从临床应用的角度来说,中医望、闻、问、切四诊法目前已经在世界范围内从事中医相关工作者中得到了广泛认可。随着人工智能技术的快速发展,为实现中医四诊客观化,基于计算机辅助的中医诊断方法逐渐被人们所关注。尤其在深度学习模型大量应用的时代,以数据驱动为核心工作模式的工程范式,逐渐取代了人工操作,而成为工业界乃至学术界主要的研究方向。本文主要针对基于数据驱动模式的计算机辅助中医四诊方法开展调研,包括数据集、信号处理方法、深度学习算法等方面的内容,为相关研究人员提供一个综合的视角,并对未来可能的发展方向进行了简要论述。