基于数据挖掘的脑白质疏松症相关因素分析

doi:10.3969/j.issn.1673-5765.2018.10.004

摘要/Abstract

摘要：

目的利用数据挖掘技术分析脑白质疏松症相关因素。

方法回顾性地收集2015年4月-2017年2月中南医院神经内科符合标准的1197例住院患者临床资料，利用数据挖掘技术4种预测模型和卡方检验基础上的特征选择方法分析脑白质疏松症相关因素。

结果 4种数据挖掘模型中，逻辑回归模型预测性能最佳[特征数为9，受试者工作特征曲线下面积（area under the curve，AUC）=0.825±0.012]。特征选择方法选出9种与脑白质疏松症相关的因素。利用决策树模型（特征数为4，AUC=0.788±0.017）得到该决策树的可视化结果图。

结论数据挖掘方法选出的因素中与脑白质疏松呈正相关的因素有年龄、高血压病史、颅内动脉狭窄、贫血、2型糖尿病、肌酐、红细胞分布宽度；呈负相关的因素有红细胞计数、血红蛋白浓度。

文章导读： 本文用数据挖掘技术分析血管源性脑白质疏松症的相关因素，为血管源性脑白质疏松症预防及危险因素研究提出一定方向，也探索了数据挖掘技术在临床研究中的应用。

关键词: 脑白质疏松症; 相关因素; 数据挖掘; 特征选择

Abstract:

Objective To analyze the related factors of leukoaraiosis by using data mining technology. Methods The clinical data of 1197 inpatients who were eligible for inclusion criteria in Zhongnan Hospital during April 2015 to February 2017 were included in the study. Among the data mining technology, 4 kinds of prediction models and chi square test based feature selection method were used to analyze the related factors of leukoaraiosis. Results Among the 4 kinds of data mining models, the logistic regression model had the best prediction performance (the number of features is 9, AUC=0.825±0.012). Feature selection methods were used to select the top 9 features as related factors of leukoaraiosis. Decision tree model (the number of features was 4, AUC=0.788±0.017) was used to obtain the visual result graph of the decision tree. Conclusions According to data mining methods in this paper, the positive relevant factors of leukoaraiosis were age, history of hypertension, arteriosclerosis, anemia, type 2 diabetes mellitus, creatinine, red blood cell count, and red blood cell distribution width; and the negative relevant factors were red blood cell count and hemoglobin concentration.

Key words: Leukoaraiosis; Relevant factors; Data mining; Feature selection

娜迪热·艾孜热提艾力, 封红亮, 张帅美, 王美瑶, 刘煜敏. 基于数据挖掘的脑白质疏松症相关因素分析[J]. 中国卒中杂志, 2018, 13(10): 1019-1024.

Nadire·Aiziretiaili, FENG Hong-Liang, ZHANG Shuai-Mei, WANG Mei-Yao, LIU Yu-Min. Analysis of Related Factors of Leukoaraiosis Based on Data Mining[J]. Chinese Journal of Stroke, 2018, 13(10): 1019-1024.

参考文献

[1] HACHINSKI V C，POTTER P，MERSKEY H.

Leuko-araiosis[J]. Arch Neurol，1987，44（1）：21-23.

[2] 戴书华，廖翠薇，杨柳，等. CT及MRI对脑白质疏松

症的诊断价值探讨[J]. 第三军医大学学报，2008，30

（9）：871-873.

[3] GRUETER B E，SCHULZ U G. Age-related

cerebral white matter disease（leukoaraiosis）：a

review[J]. Postgrad Med J，2012，88（1036）：79-87.

[4] 封红亮，张帅美，罗征进，等. 严重脑白质疏松症相

关因素的病例对照研究[J]. 中风与神经疾病，2017，

34（4）：324-327.

[5] ZHANG S，KANG X. Investigation of the risk

factors for leukoaraiosis（LA）[J]. Asia Pac J Public

Health，2013，25（4 Suppl）：64S-71S.

[6] LI J J，HUANG Y H，LIN Y Y，et al. The

association of uric acid with leukoaraiosis[J]. J Intl

Med Res，2016，45（1）：75-81.

[7] 高叶萌，张立，高潇，等. 缺血性脑白质病变的研究

进展[J]. 中风与神经疾病，2016，33（9）：857-858.

[8] FAZEKAS F，CHAWLUK J B，ALAVI A，et al.

MR signal abnormalities at 1. 5 T in Alzheimer's

dementia and normal aging[J]. AJR Am J Roentgenol，

1987，149（2）：351-356.

[9] 魏娜，王拥军，张玉梅. 脑白质病变4个分级量表的

信度研究[J]. 中国康复理论与实践，2012，18（6）：

562-565.

[10] LI X，LIU H，DU X，et al. Integrated Machine

Learning Approaches for Predicting Ischemic Stroke

and Thromboembolism in Atrial Fibrillation[J/OL].

AMIA Annu Symp Proc，2017，2016：799-807. https：

//www.ncbi.nlm.nih.gov/pmc/articles/PMC5333223/

[11] SAEYS Y，INZA I，LARRAÑAGA P. A review of

feature selection techniques in bioinformatics[J].

Bioinformatics，2007，23（19）：2507-2517.

[12] BASILE A M，PANTONI L，PRACUCCI G，et al.

Age，hypertension，and lacunar stroke are the major

determinants of the severity of age-related white

matter changes. The LADIS（Leukoaraiosis and

Disability in the Elderly）Study[J]. Cerebrovasc Dis，

2006，21（5-6）：315-322.

[13] 黄焕. 缺血性脑白质病变与动态血压的相关性研究

[D]. 杭州：浙江大学，2015.

[14] LEE S J，KIM J S，LEE K S，et al. The leukoaraiosis

is more prevalent in the large artery atherosclerosis

stroke subtype among Korean patients with ischemic

stroke[J/OL]. BMC Neurol，2008，8：31. https：//doi.

org/10.1186/1471-2377-8-31

[15] LUCATELLI P，MONTISCI R，SANFILIPPO R，

et al. Is there an association between leukoaraiosis

volume and diabetes?[J]. J Neuroradiol，2016，43（4）：

273-279.

[16] VAN HARTEN B，DE LEEUW F E，WEINSTEIN

H C，et al. Brain imaging in patients with diabetes：a

systematic review[J]. Diabetes Care，2006，29（11）：

2539-2548.

[17] TOYODA K. Cerebral small vessel disease and

chronic kidney disease[J]. J Stroke，2015，17（1）：

31-37.

[18] INZITARI M，STUDENSKI S，ROSANO C，et

al. Anemia is associated with the progression of

white matter disease in older adults with high blood

pressure：the cardiovascular health study[J]. J Am

Geriatr Soc，2008，56（10）：1867-1872.

[19] 彭想想，费世早，陈向宇，等. 急性缺血性脑卒中患者

合并脑白质疏松症的RDW变化[J]. 皖南医学院学报，

2017，36（1）：16-19.