本文已被：浏览 984次 下载 560次
投稿时间：2020-04-16 网络发布日期：2020-11-20
投稿时间：2020-04-16 网络发布日期：2020-11-20
中文摘要: 目的 基于数据挖掘与可视化技术构建2型糖尿病疾病关联网络,揭示2型糖尿病相关疾病关联关系。方法 通过山西医科大学第一医院医疗大数据平台最终整理得到33 723条记录，210 861条有效诊断。利用共词分析方法,构建2型糖尿病疾病共现网络,并运用Gephi软件进行可视化展示。结果 构建的2型糖尿病疾病共现网络包含442个节点,2 304条连线,网络密度为0.024,平均最短路径为1.979;经模块化分析,网络聚类为11个社区,其中节点数大于10,词频占比大于5%的社区有6个,分别为2型糖尿病为中心的社区(可细分为高血压/高血脂集群、甲状腺相关疾病集群、骨关节病集群)、高血压及其脑血管病集群、冠心病集群、心律失常集群、2型糖尿病并发症集群、肺心病及其症状集群,其中2型糖尿病与高血压3级的关联最密切,网络连线最粗。结论 通过数据挖掘与可视化技术,展示了2型糖尿病关联疾病集群,有机会为2型糖尿病综合诊疗提供新的思路。
Abstract:Objective To Construct an association network for type 2 diabetes based on data mining and visualization technology,revealing the association relationship for type 2 diabetes.Methods Through the Health and Medical Big Data platform of the First Hospital of Shanxi Medical University,33 723 records and 210 861 valid diagnoses were finally sorted out.The co-word analysis method was used to construct the co-occurrence network of type 2 diabetes,and Gephi software was used for visual display.Results The constructed type 2 diabetes co-occurrence network contained 442 nodes,2 304 connections,a network density of 0.024,and an average shortest path of 1.979.After modular analysis,the network was clustered into 11 communities,of which the number of nodes was greater than 10.There were 6 communities whose word frequency accounts for more than 5%,namely:type 2 diabetes-centered communities (which can be subdivided into hypertension/hyperlipidemia clusters,thyroid-related diseases clusters,osteoarthropathy clusters),hypertension and its cerebrovascular disease cluster,coronary heart disease cluster,arrhythmia cluster,type 2 diabetes complication cluster,pulmonary heart disease and its symptom cluster,among which type 2 diabetes was the closest to hypertension level 3,and the network connection is the thickest.ConclusionThrough data mining and visualization technology,it shows the clusters of type 2 diabetes related diseases,and has the opportunity to provide new ideas for the comprehensive diagnosis and treatment of type 2 diabetes.
文章编号： 中图分类号： 文献标志码：B
基金项目:山西省重点研发计划项目(201803D31099)
附件
引用文本：
杨志清,张靓,郭玲玲,等.基于数据挖掘与可视化技术的2型糖尿病疾病关联性分析[J].中国临床研究,2020,33(11):1490-1494.
杨志清,张靓,郭玲玲,等.基于数据挖掘与可视化技术的2型糖尿病疾病关联性分析[J].中国临床研究,2020,33(11):1490-1494.