血管性认知障碍早期预测机器学习模型的构建

doi:10.3969/j.issn.1006-9771.2021.09.011

《中国康复理论与实践》 ›› 2021, Vol. 27 ›› Issue (9): 1072-1077.doi: 10.3969/j.issn.1006-9771.2021.09.011

血管性认知障碍早期预测机器学习模型的构建

张倩^1,²,卞敏洁^1,²,何琴^1,²,黄东锋^1,²()

1.中山大学附属第七医院康复医学科,广东深圳市 518000
2.广东省康复医学与临床转化工程技术研究中心,广东广州市 510000

收稿日期:2021-05-18 修回日期:2021-08-06 出版日期:2021-09-25 发布日期:2021-10-09
通讯作者: 黄东锋 E-mail:Huangdf_sysu@163.com
作者简介:张倩(1989-),女,汉族,内蒙古包头市人,硕士,医师,主要研究方向：神经康复。
基金资助:
中山大学临床医学研究5010计划项目(2014001)

Machine Learning Models for Early Prediction of Vascular Cognitive Impairment

ZHANG Qian^1,²,BIAN Min-jie^1,²,HE Qin^1,²,HUANG Dong-feng^1,²()

1. Department of Rehabilitation Medicine, the Seventh Affiliated Hospital, Sun Yat-sen University, Shenzhen, Guangdong 518000, China
2. Guangdong Engineering and Technology Research Center for Rehabilitation Medicine and Translation, Guangzhou, Guangdong 510000, China

Received:2021-05-18 Revised:2021-08-06 Published:2021-09-25 Online:2021-10-09
Contact: HUANG Dong-feng E-mail:Huangdf_sysu@163.com
Supported by:
Sun Yat-sen University Clinical Medicine Reseach Plan 5010(2014001)

摘要/Abstract

摘要：

目的探索以血管性高危因素构建的机器学习模型早期预测血管性认知障碍的预测性能。方法 2020年4月至9月,收集本院住院患者及陪护人员70例的人口学资料、血管性高危因素,行蒙特利尔认知评估量表(MoCA)评估,根据评估结果将受试者分为正常组、血管性轻度认知障碍(VaMCI)组和痴呆组;单因素方差分析筛选组间存在显著性差异的血管性高危因素,采用支持向量机(SVM)和极限学习机(ELM)构建预测模型;采用接受者操作特征曲线比较两种模型的预测性能。结果根据MoCA评估结果,正常组32例,VaMCI组23例,痴呆组15例;三组间收缩压、空腹血糖、总胆固醇、低密度脂蛋白、血尿酸有显著性差异(F > 3.318, P < 0.05);SVM模型预测VaMCI的曲线下面积最高,为0.911 (P < 0.01),SVM模型优于ELM模型。结论基于血管性高危因素构建的SVM预测模型优于ELM模型。

关键词: 血管性认知障碍, 支持向量机, 极限学习机, 机器学习, 预测模型

Abstract:

Objective To explore the predictive performance of machine learning model based on vascular risk factors in early prediction of vascular cognitive impairment.Methods From April to September, 2020, 70 subjects were enrolled and collected information of the demographics and vascular risk factors. They were assessed with Montreal Cognitive Assessment (MoCA), and then divided into normal group, vascular mild cognitive impairment (VaMCI) group and dementia group. The differences of vascular risk factors among the three groups were detected with one-way ANOVA, and the significant factors were selected to establish predictive models with support vector machine (SVM) and extreme learning machine (ELM). The predictive performance of two models was compared with Receiver Operating Characteristic Curve.Results There were 32 cases in the normal group, 23 in VaMCI group and 15 in dementia group. Systolic blood pressure, fasting blood glucose, total cholesterol, low density lipoprotein and blood uric acid were significantly different among the three groups (F > 3.318, P < 0.05). The area under curve was the most (0.911) in SVM model predicting for VaMCI (P < 0.01), and the predictive efficacy was better for SVM model.Conclusion SVM predictive model based on vascular risk factors may be more effective for predicting VaMCI.

Key words: vascular cognitive impairment, support vector machines, extreme learning machines, machine learning, predictive model

张倩,卞敏洁,何琴,黄东锋. 血管性认知障碍早期预测机器学习模型的构建[J]. 《中国康复理论与实践》, 2021, 27(9): 1072-1077.

ZHANG Qian,BIAN Min-jie,HE Qin,HUANG Dong-feng. Machine Learning Models for Early Prediction of Vascular Cognitive Impairment[J]. 《Chinese Journal of Rehabilitation Theory and Practice》, 2021, 27(9): 1072-1077.

图/表 6

表1

表2

表3

图1

图2

表4

参考文献 37

[1]	CAMPBELL B C V, DE SILVA D A, MACLEOD M R, et al. Ischaemic stroke[J]. Nat Rev Dis Primers, 2019, 5(1):70. doi: 10.1038/s41572-019-0118-8
[2]	KALARIA R N, AKINYEMI R, IHARA M. Stroke injury, cognitive impairment and vascular dementia[J]. Biochim Biophys Acta, 2016, 1862(5):915-925.
[3]	IADECOLA C, DUERING M, HACHINSKI V, et al. Vascular Cognitive Impairment and Dementia: JACC Scientific Expert Panel[J]. J Am Coll Cardiol, 2019, 73(25):3326-3344. doi: 10.1016/j.jacc.2019.04.034
[4]	SACHDEV P S, BLACKER D, BLAZER D G, et al. Classifying neurocognitive disorders: the DSM-5 approach[J]. Nat Rev Neurol, 2014, 10(11):634-642. doi: 10.1038/nrneurol.2014.181
[5]	SKROBOT O A, BLACK S E, CHEN C, et al. Progress toward standardized diagnosis of vascular cognitive impairment: Guidelines from the Vascular Impairment of Cognition Classification Consensus Study[J]. Alzheimers Dement, 2018, 14(3):280-292. doi: 10.1016/j.jalz.2017.09.007
[6]	BARTHELS D, DAS H. Current advances in ischemic stroke research and therapies[J]. Biochim Biophys Acta Mol Basis Dis, 2020, 1866(4):165260. doi: 10.1016/j.bbadis.2018.09.012
[7]	YAO Q, ZHANG J, YAN K, et al. Development and validation of a 2-year new-onset stroke risk prediction model for people over age 45 in China[J]. Medicine (Baltimore), 2020, 99(41):e22680. doi: 10.1097/MD.0000000000022680
[8]	GORELICK P B, SCUTERI A, BLACK S E, et al. Vascular contributions to cognitive impairment and dementia: a statement for healthcare professionals from the american heart association/american stroke association[J]. Stroke, 2011, 42(9):2672-2713. doi: 10.1161/STR.0b013e3182299496
[9]	DICHGANS M, ZIETEMANN V. Prevention of vascular cognitive impairment[J]. Stroke, 2012, 43(11):3137-3146. doi: 10.1161/STROKEAHA.112.651778
[10]	PALTA P, SCHNEIDER A L, BIESSELS G J, et al. Magnitude of cognitive dysfunction in adults with type 2 diabetes: a meta-analysis of six cognitive domains and the most frequently reported neuropsychological tests within domains[J]. J Int Neuropsychol Soc, 2014, 20(3):278-291. doi: 10.1017/S1355617713001483
[11]	YANG M, WILLIAMSON J. Blood pressure and statin effects on cognition: a review[J]. Curr Hypertens Rep, 2019, 21(9):70. doi: 10.1007/s11906-019-0973-4
[12]	YANG Z, WANG H, EDWARDS D, et al. Association of blood lipids, atherosclerosis and statin use with dementia and cognitive impairment after stroke: a systematic review and meta-analysis[J]. Ageing Res Rev, 2020, 57:100962. doi: 10.1016/j.arr.2019.100962
[13]	黄菲芸, 王艳红, 李娟娟, 等. 蒙特利尔认知评估量表在中国中老年人群筛查轻度认知功能障碍中的截断值的系统评价[J]. 中国循证医学杂志, 2017, 17(4):450-457.
	HUANG F Y, WANG Y H, LI J J, et al. Diagnostic value of montreal cognitive assessment for mild cognitive impairment in Chinese middle-aged adults: a meta-analysis[J]. Chin J Evid Based Med, 2017, 17(4):450-457.
[14]	王伊龙, 韩尚容, 曹勇, 等. 中国脑血管病临床管理指南(节选版):脑血管病高危人群管理[J]. 中国卒中杂志, 2019, 14(7):700-708.
	WANG Y L, HAN S R, CAO Y, et al. Chin J Stroke, 2019, 14(7):700-708.
[15]	KAMAL H, LOPEZ V, SHETH S A. Machine learning in acute ischemic stroke neuroimaging[J]. Front Neurol, 2018, 9:945. doi: 10.3389/fneur.2018.00945
[16]	SIRSAT M S, FERMÉ E, CÂMARA J. Machine learning for brain stroke: a review[J]. J Stroke Cerebrovasc Dis, 2020, 29(10):105162. doi: 10.1016/j.jstrokecerebrovasdis.2020.105162
[17]	KISS T, BALASUBRAMANIAN P, VALCARCEL-ARES M N, et al. Nicotinamide mononucleotide (NMN) treatment attenuates oxidative stress and rescues angiogenic capacity in aged cerebromicrovascular endothelial cells: a potential mechanism for the prevention of vascular cognitive impairment[J]. Geroscience, 2019, 41(5):619-630. doi: 10.1007/s11357-019-00074-2
[18]	GREGORY M A, MANUEL-APOLINAR L, SÁNCHEZ-GARCIA S, et al. Soluble intercellular adhesion molecule-1 (sICAM-1) as a biomarker of vascular cognitive impairment in older adults[J]. Dement Geriatr Cogn Disord, 2019, 47(4-6):243-253. doi: 10.1159/000500068
[19]	ARMSTRONG N M, BANGEN K J, AU R, et al. Associations between midlife (but not late-life) elevated coronary heart disease risk and lower cognitive performance: results from the Framingham Offspring Study[J]. Am J Epidemiol, 2019, 188(12):2175-2187. doi: 10.1093/aje/kwz210
[20]	BANERJEE G, CHAN E, AMBLER G, et al. Effect of small-vessel disease on cognitive trajectory after atrial fibrillation-related ischaemic stroke or TIA[J]. J Neurol, 2019, 266(5):1250-1259. doi: 10.1007/s00415-019-09256-6
[21]	QIAO M, CHEN C, LIANG Y, et al. The influence of serum uric acid level on Alzheimer's disease: a narrative review[J]. Biomed Res Int, 2021, 2021:5525710.
[22]	姚辉, 钱园园, 刘金颖, 等. 血管性认知障碍的治疗进展[J]. 医学综述, 2019, 25(13):2617-2621.
	YAO H, QIAN Y Y, LIU J Y, et al. Progress in treatment of vascular cognitive impairment[J]. Med Recapitulate, 2019, 25(13):2617-2621.
[23]	王凌雪, 李双阳, 梁岚, 等. 血管性认知障碍的中西医治疗进展[J]. 实用中医内科杂志, 2020, 34(3):90-94.
	WANG L X, LI S Y, LIANG L, et al. Progress in treatment of vascular cognitive impairment with traditional Chinese and western medicine[J]. J Prac Tradit Chin Inter Med, 2020, 34(3):90-94.
[24]	PILIPOVICH A A, VOROB'EVA O V. [Mild cognitive impairment: modern aspects of diagnostics and therapy][J]. [in Russian]. Zh Nevrol Psikhiatr Im S S Korsakova, 2020, 120(11):124-130.
[25]	JIA L, QUAN M, FU Y, et al. Dementia in China: epidemiology, clinical management, and research advances[J]. Lancet Neurol, 2020, 19(1):81-92. doi: 10.1016/S1474-4422(19)30290-X
[26]	GHOLAMNEZHAD Z, BOSKABADY M H, JAHANGIRI Z. Exercise and dementia[J]. Adv Exp Med Biol, 2020, 1228:303-315.
[27]	DEMURTAS J, SCHOENE D, TORBAHN G, et al. Physical activity and exercise in mild cognitive impairment and dementia: an umbrella review of intervention and observational studies[J]. J Am Med Dir Assoc, 2020, 21(10):1415-1422, e6. doi: 10.1016/j.jamda.2020.08.031
[28]	JI S, DUAN J, HOU X, et al. The role of acupuncture improving cognitive deficits due to Alzheimer's disease or vascular diseases through regulating neuroplasticity[J]. Neural Plast, 2021, 2021:8868447.
[29]	DJABELKHIR L, WU Y H, VIDAL J S, et al. Computerized cognitive stimulation and engagement programs in older adults with mild cognitive impairment: comparing feasibility, acceptability, and cognitive and psychosocial effects[J]. Clin Interv Aging, 2017, 12:1967-1975. doi: 10.2147/CIA
[30]	GRAFF-RADFORD J. Vascular cognitive impairment[J]. Continuum (Minneap Minn), 2019, 25(1):147-164.
[31]	HE J, ZHAO C, LIU W, et al. Neurochemical changes in the hippocampus and prefrontal cortex associated with electroacupuncture for learning and memory impairment[J]. Int J Mol Med, 2018, 41(2):709-716.
[32]	HARVEY P D. Domains of cognition and their assessment[J]. Dialogues Clin Neurosci, 2019, 21(3):227-237. doi: 10.31887/DCNS
[33]	AL-AIDROOS N, SAID C P, TURK-BROWNE N B. Top-down attention switches coupling between low-level and high-level areas of human visual cortex[J]. Proc Natl Acad Sci U S A, 2012, 109(36):14675-14680. doi: 10.1073/pnas.1202095109
[34]	VAN DER FLIER W M, SKOOG I, SCHNEIDER J A, et al. Vascular cognitive impairment[J]. Nat Rev Dis Primers, 2018, 4:18003. doi: 10.1038/nrdp.2018.3
[35]	XIONG L, REIJMER Y D, CHARIDIMOU A, et al. Intracerebral hemorrhage and cognitive impairment[J]. Biochim Biophys Acta, 2016, 1862(5):939-944.
[36]	LO COCO D, LOPEZ G, CORRAO S. Cognitive impairment and stroke in elderly patients[J]. Vasc Health Risk Manag, 2016, 12:105-116.
[37]	PARK Y H, JANG J W, PARK S Y, et al. Executive function as a strong predictor of recovery from disability in patients with acute stroke: a preliminary study[J]. J Stroke Cerebrovasc Dis, 2015, 24(3):554-561. doi: 10.1016/j.jstrokecerebrovasdis.2014.09.033

变量	正常组 (n = 32)	VaMCI组 (n = 23)	痴呆组 (n = 15)	χ²/F值^a	P值^a	χ²/t值^b	P值^b	χ²/t值^c	P值^c
性别(男/女, n)	16/16	12/11	8/7	0.053	0.979
年龄(岁)	34.9±2.2	53.9±2.7	59.7±3.6	24.310	< 0.001	5.277	< 0.001	1.282	0.210
脑梗死(男/女, n)	0	5/6	6/4	0.444	0.505
脑出血(男/女, n)	0	2/2	2/3	0.090	0.764
非卒中(男/女, n)	16/16	5/3	0	0.401	0.527
教育年限(年)	13.6±0.7	9.4±0.9	10.1±1.0	7.965	< 0.001	3.631	< 0.001	0.519	0.608
身高(m)	1.6±0.01	1.6±0.02	1.7±0.02	0.453	0.638
体质量(kg)	62.5±2.4	65.1±2.6	64.6±1.8	0.348	0.708
BMI (kg/m²)	23.4±0.7	24.4±0.7	23.7±0.7	0.456	0.636

项目	正常组 (n = 32)	VaMCI组 (n = 23)	痴呆组 (n = 15)	F值^a	P值^a	t值^b	P值^b	t值^c	P值^c
总分	27.5±0.4	21.8±0.3	13.5±0.9	150.600	< 0.001	9.071	< 0.001	8.481	< 0.001
视空间与执行功能	4.6±0.1	3.8±0.3	1.6±0.3	41.430	< 0.001	2.720	0.010	4.783	< 0.001
命名功能	3.0±0.0	2.9±0.1	2.3±0.2	16.140	< 0.001	1.335	0.189	3.199	0.003
注意功能	6.0±0.0	4.4±0.2	2.8±0.3	90.480	< 0.001	10.040	< 0.001	4.298	< 0.001
语言功能	2.5±0.1	1.8±0.2	0.9±0.2	17.630	< 0.001	2.857	0.007	2.725	0.011
抽象功能	1.4±0.1	0.9±0.2	0.5±0.2	8.405	< 0.001	2.408	0.020	1.394	0.174
延迟回忆功能	3.8±0.2	2.2±0.3	0.5±0.3	41.230	<0.001	4.180	< 0.001	4.365	< 0.001
定向功能	5.9±0.1	4.8±0.2	4.1±0.4	16.140	< 0.001	5.746	< 0.001	1.383	0.1773

因素	正常组 (n = 32)	VaMCI组 (n = 23)	痴呆组 (n = 15)	F值^a	P值^a	t值^b	P值^b	t值^c	P值^c
收缩压(mmHg)	122.7±2.0	142.2±4.1	133.4±2.8	12.580	< 0.001	4.737	< 0.001	1.736	0.093
舒张压(mmHg)	77.1±1.4	80.6±1.4	78.4±1.6	1.333	0.272	1.588	0.112	1.041	0.306
空腹血糖(mmol/L)	5.2±0.2	6.5±0.3	6.2±0.4	8.925	< 0.001	4.488	< 0.001	0.743	0.464
总胆固醇(mmol/L)	4.1±0.2	4.8±0.3	4.135±0.2603	3.318	0.044	2.489	0.017	1.921	0.065
低密度脂蛋白(mmol/L)	3.1±0.1	3.3±0.3	2.3±0.2	6.644	0.003	0.786	0.436	2.846	0.008
血尿酸(μmol/L)	363.5±14.4	381.9±25.6	488.2±17.1	11.900	< 0.001	0.677	0.502	3.404	0.002

项目	SVM			ELM			P值
项目	正常组	VaMCI组	痴呆组	正常组	VaMCI组	痴呆组	P值
AUC	0.8662	0.9109	0.8723	0.8289	0.8634	0.8336	0.006
敏感性	0.7928	0.8068	0.8953	0.8837	0.8595	0.7653	0.953
特异性	0.8272	0.8583	0.7823	0.6389	0.7500	0.7823	0.212

血管性认知障碍早期预测机器学习模型的构建

Machine Learning Models for Early Prediction of Vascular Cognitive Impairment

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 6

参考文献 37

相关文章 13

Metrics

本文评价

推荐阅读 0

[1]	朱旭,刘静,董泽萍,仇大伟. 基于表面肌电图手势动作意图识别的系统综述[J]. 《中国康复理论与实践》, 2022, 28(9): 1032-1038.
[2]	张博超,杨朝,郭立泉,陈静,熊大曦. 基于机器学习的慢性阻塞性肺疾病急性加重预测模型的研究[J]. 《中国康复理论与实践》, 2022, 28(6): 678-683.
[3]	王博,袁永学,张庆苏. 非经口进食吞咽障碍脑卒中患者预后的相关因素及预测模型[J]. 《中国康复理论与实践》, 2022, 28(4): 453-460.
[4]	俞快,张利,叶祥明. 慢性意识障碍患者结局预测模型的系统综述[J]. 《中国康复理论与实践》, 2022, 28(2): 190-198.
[5]	丁妍怡,张胜行,刘雨露,余燕,杨敏光,梁胜祥,柳维林,陶静. 电针对血管性认知障碍大鼠脑功能局部一致性的效果[J]. 《中国康复理论与实践》, 2022, 28(1): 55-61.
[6]	蔡亚飞, 洪毅, 王方永. 脊髓损伤神经功能变化随访与预测模型[J]. 《中国康复理论与实践》, 2019, 25(10): 1120-1124.
[7]	崔尧, 丛芳, 李建军, 高峰, 杜良杰, 杨明亮. 成年脊髓损伤患者功能结局的预测模型研究[J]. 《中国康复理论与实践》, 2019, 25(10): 1125-1132.
[8]	张文冉, 王正则, 许晓丹, 张丽莹, 田沈. 血管性认知障碍脑CT灌注成像脑血流特点[J]. 《中国康复理论与实践》, 2018, 24(8): 942-949.
[9]	李静, 刘阳, 李晓芳, 代瑞廷. 愉快家庭康复对血管性认知障碍患者的效果[J]. 《中国康复理论与实践》, 2017, 23(3): 263-265.
[10]	张华a;王昊b;李多多b;贾宝林b;谭中建c;陈红b. 颈椎病慢性疼痛患者的默认网络研究[J]. 《中国康复理论与实践》, 2015, 21(01): 69-073.
[11]	王安睿;费树岷. 粗糙集和支持向量机应用于帕金森病辅助诊断[J]. 《中国康复理论与实践》, 2009, 15(11): 1086-1088.
[12]	周爱红;贾建平;闵保权 . 血管性认知障碍的神经心理学特征及存在的问题 [J]. 《中国康复理论与实践》, 2008, 14(03): 214-216.
[13]	杜万良综述;王荫华审校. 血管性认知障碍 [J]. 《中国康复理论与实践》, 2004, 10(04): 238-240.