基于贝叶斯网络分析重症肺炎中医证候规律

Analysis of Chinese Medicine Syndrome Rules of Severe Pneumonia Based on Bayesian Network

目的:利用贝叶斯网络技术分析重症肺炎中医证候规律,为重症肺炎证候标准化研究提供客观依据。方法:首先收集中国中医科学院望京医院重症医学科247例重症肺炎患者的中医证候信息,然后利用贝叶斯网络结构学习技术建立证候间的贝叶斯网络概率模型,最后结合专家经验,提取重症肺炎的常见证型。结果:提取的证候要素包括风热、痰热、痰湿、气虚、阴虚、阳虚6个病性证素和肺、脾、肾3个病位证素。在此基础上提取6个常见证型,分别为风热犯肺证、痰湿阻肺证、痰热蕴肺证、痰湿阻肺兼肺脾气虚证、痰热蕴肺兼气阴两虚证、邪陷正脱证。结论:贝叶斯网络技术为重症肺炎证候规律标准化研究提供了科学性及客观性依据。

Objective: To analyze the rules of Chinese medicine syndrome of severe pneumonia based on Bayesian network technology so as to provide an objective basis for the standardization of syndrome of severe pneumonia.Methods: Firstly, the Chinese medicine syndrome information of 247 patients with severe pneumonia in the Department of Critical Medicine, Wangjing Hospital, China Academy of Chinese Medical Sciences was collected, and then the Bayesian network probability model between syndromes was established by using Bayesian network structure learning technology.Finally, combined with expert experience, the common syndromes of severe pneumonia were extracted.Results: The extracted syndrome elements include six disease syndrome elements of wind-heat, phlegm-heat, phlegm-dampness, qi deficiency, yin deficiency and yang deficiency, and three disease location syndrome elements of lung, spleen and kidney.On this basis, six common syndrome types were extracted, which were syndrome of wind-heat invading lung, syndrome of phlegm-dampness blocking lung, syndrome of phlegm-heat accumulating lung, syndrome of phlegm-dampness blocking lung and lung spleen qi deficiency, syndrome of phlegm-heat accumulation in lung and deficiency of qi and yin, syndrome of interior invasion of pathogen and vital qi collapse.Conclusion: Bayesian network technology provides a scientific and objective basis for the standardization of syndrome law of severe pneumonia.