基于机械通气波形大数据的人机不同步自动监测方法  

Review on automatic detection methods of patient-ventilatory asynchrony based on big data of mechanical ventilation waveform

作　　者：潘清[1] 葛慧青[2] Pan Qing;Ge Huiqing(College of Information Engineering,Zhejiang University of Technology,Hangzhou 310023,China;Department of Respiratory Care,Sir Run Run Shaw Hospital,School of Medicine,Zhejiang University,Hangzhou 310016,China)

机构地区：[1]浙江工业大学信息工程学院,杭州310023 [2]浙江大学医学院附属邵逸夫医院呼吸治疗科国家呼吸区域医疗中心,杭州310016

出　　处：《中华重症医学电子杂志》2024年第4期399-403,共5页Chinese Journal Of Critical Care & Intensive Care Medicine(Electronic Edition)

基　　金：国家自然科学基金资助项目(32371372,82070087)

摘　　要：人机不同步(PVA)在机械通气中较为常见,与呼吸做功增加、机械通气时间延长、呼吸机诱发肺损伤以及不良的预后密切相关。识别PVA需要仔细观察患者及其呼吸机波形,但临床医护人员识别PVA的能力参差不齐,且难以在床旁持续监测,亟需自动化的监测手段。随着机械通气波形大数据的日趋成熟,PVA自动检测算法在近年快速发展,呈现出数据驱动与知识驱动协同发展的趋势。本文综述PVA自动检测方法的发展历程,概述基于规则、传统机器学习、深度学习、生理系统模型的技术的优缺点,介绍PVA实时检测与分析系统的发展和临床应用现状,并探讨基于机械通气波形大数据检测PVA的研究面临的缺乏标准数据集、算法泛化能力不足等挑战。Patient-ventilatory asynchrony(PVA)is common during mechanical ventilation,and is closely associated with elevated work of breath,prolonged mechanical ventilation,ventilator-induced lung injury,as well as worse clinical outcomes.Identifying PVA requires careful observation of the patient and their ventilator waveforms,but clinical healthcare providers vary in their ability to recognize PVA,and continuous bedside monitoring is challenging,urging the development of automated monitoring methods.PVA automatic detection algorithms have rapidly developed in recent years,showing a trend of synergistic development driven by data and knowledge.This article reviews the development history of PVA automatic detection methods,outlines the advantages and disadvantages of technologies based on rules,traditional machine learning,deep learning,and physiological system models,introduces the development and clinical application status of real-time PVA detection and analysis systems,and discusses the challenges faced in PVA detection based on mechanical ventilation waveform big data,such as the lack of standard datasets and insufficient algorithm generalization capability.

关 键 词：机械通气 人机不同步 机器学习 深度学习 生理系统建模 

分 类 号：R197.39[医药卫生—卫生事业管理] R318.6[医药卫生—公共卫生与预防医学]