中医脉诊智能化设备数据标注的研究述评

Research review on data annotation for intelligent pulse diagnosis devices in traditional Chinese medicine

近年来中医诊断智能化的研究火热,其中脉诊的智能化更是研究重点。脉诊的智能化主要集中在采脉与数据处理两个方面,采脉的难点主要集中在传感器及采集指,其数据来源以及处理方式目前行业内无统一标准,目前相关研究主要聚焦在脉图变化以及桡动脉超声方面。介绍基于脉诊八要素(脉位、脉力、至数、脉律、脉长、脉宽、流利度及紧张度)的数据标注的方法,比较脉图及桡动脉超声的优缺点,并提出三维脉图作为脉诊数据标注形式的可行性,进一步探讨目前对于脉图的前端处理方法及人工智能、机器学习等应用于脉图数据标注的现状以及建议,以期为脉诊客观化提供新的研究思路。

In recent years,there has been a surge of interest in the research on the intelligentization of traditional Chinese medicine(TCM)diagnostics,with intelligent pulse diagnosis being a focal point.The intelligentization of pulse diagnosis primarily focuses on pulse acquisition and data processing.The challenges in pulse acquisition mainly lie in sensors and pulse acquisition tools,and there is currently no unified industry standard for data sources or processing methods.Present research predominantly focuses on pulse waveform variations and radial artery ultrasound data.This paper introduces data annotation methods based on the eight key elements of pulse diagnosis(pulse position,pulse strength,pulse beats,pulse rhythm,pulse length,pulse width,fluency and tension),compares the advantages and disadvantages of pulse waveform and radial artery ultrasound,and explores the feasibility of using threedimensional pulse waveforms as a form of pulse diagnosis data annotation.It further discusses the current front-end processing methods for pulse waveforms and the application of artificial intelligence and machine learning in pulse waveform data annotation,aiming to provide new research perspectives for the objectification of pulse diagnosis.