基于关联规则的数据挖掘在超重肥胖患者心脏构型超声数据模型中的初步研究  

Data mining in echocardiographic parameter model of overweight and obese patients based on association rules

作　　者：孙婷婷 朱向明[2] 刘冬[3] 张霞[2] 张健[1] SUN Tingting;ZHU Xiangming;LIU Dong(Department of Ulstrasonography,Zhejiang Provincial People’s Hospital,Hangzhou 310014,China)

机构地区：[1]浙江省人民医院超声科,杭州310014 [2]皖南医学院附属弋矶山医院超声科 [3]皖南医学院附属弋矶山医院信息科

出　　处：《浙江医学》2019年第21期2256-2259,共4页Zhejiang Medical Journal

基　　金：浙江省医药卫生科研项目(2017KY020);安徽省2017年公益性技术应用研究联动计划项目(1704f0804048)

摘　　要：目的通过对超重肥胖患者心脏构型参数与临床信息的数据挖掘,揭示出数据之间潜在、有价值的关联信息。方法选取受检者212例(正常体重受检者69例,超重患者84例,肥胖患者59例)作为研究对象,获取临床信息及心脏构型参数,采用关联规则(Apriori算法)进行挖掘分析,提取有效关联规则并优选出能反映超重肥胖人群心脏构型及其变化的特征性指标及相关影响因素,建立心脏构型超声数据关联规则模型。结果根据提取的有效关联规则显示,超重肥胖患者收缩期末左心房前后径、左心房面积、左心房容积、左心室心肌重量与正常体重受检者相比倾向于变大,年龄、肥胖程度以及肥胖病程与心脏构型的改变密切相关。结论数据挖掘在医疗大数据中提取隐藏的关联信息,有助于超重肥胖患者心脏构型变化的早期检测、早期预防及早期干预。Objective To screen influencing factors of cardiac geometry from basic clinical information of overweight/obese patients using data mining techniques and to reveal the potential relationship between influencing factors and the change of cardiac structure. Methods Two hundred and twelve patients were recruited in the study, including 69 overweight patients, 84 obese patients and 59 normal subjects. Data of basic clinical information and cardiac configuration indexes were collected and analyzed with association rules(Apriori algorithm). The influencing factors were screened with data mining and the effective association rules were examined. Then built the association rule model of cardiac geometry in overweight/obese patients. Results According to the effective association rules extracted, the left atrial diameter, left atrial area, left atrial volume, left ventricular mass tended to be changed in overweight and obese patients. Age, obesity degree and obesity duration were closely related to changes in cardial configuration. Conclusion Data mining is useful for early detection,early prevention and early intervention of cardiac configuration changes in overweight and obese patients by extracting hidden association information from a large number of data.

关 键 词：数据挖掘 关联规则 超重肥胖 心脏 构型 

分 类 号：R54[医药卫生—心血管疾病]