基于数据挖掘的慢性肺源性心脏病常见证候特征的临床研究

Clinical Research of TCM Common Syndrome for Chronic Pulmonary Heart Disease Based on the Data Mining Technology

目的:基于国内4家中医院呼吸科719例慢性肺源性心脏病临床调查数据,探讨其中医常见证候本质特征。方法:采用临床流行病学方法,采集4所三甲医院719例慢性肺源性心脏病患者的临床资料,使用Epidata软件建立数据库,采用MATLAB 6.5软件,运用数据挖掘方法,分析慢性肺源性心脏病的常见证候及其特征。结果:719例患者中最终获取23个模糊规则,通过规则转换及设定的证候主次症标准,辨证为6个常见证候:痰热壅肺证、痰湿壅肺证、痰瘀阻肺证、阳虚水犯证、肺肾气虚证、肺肾气阴两虚等,常见证候诊断标准检验结果符合率为72.2%。结论:慢性肺源性心脏病的常见证候的属性有一定规律性,数据挖掘技术可应用于中医证候特征的研究。

Objective：To explore the essential characteristic of TCM syndromes for Chronic pulmonary heart disease （CPHD） based on the 719 CPHD patients clinical date from the domestic four hospitals respiratory department. Methods：The 719 CPHD subjects were recruited from four TCM hospitals respirations department by means of clinical epidemiological method. The research dates were inputted with Epidata ver. 3. 1 and analyzed with data mining technology adopted MATLAB6. 5 software in order to analysis the common syndrome for CPHD. Results：Twenty three fuzzy rules were extracted from 719 CPHD patients. There were six common rules before rules transformed and main symptom and minor symptom for common syndrome setted. The TCM common syndrome for CPHD included phlegm-heat obstructing lungs syndrome,phlegm-damp stagnating lung syndrome,phlegm and blood stasis obstructing lung syndrome,water overflowing due to yang deficiency,insufficiency of lung and kidney qi syndrome,deficiency of both qi and yin of lung and kidney. The diagnostic accuracy rate was72. 2% before the common syndrome diagnostic criteria tested. Conclusion：The common syndromes for CPHD were of certain regularity. The data mining technology could been used into the TCM syndrome characteristic research.