基于主成分分析法的脑卒中后轻度认知障碍风险预测模型研究

Study on risk prediction model for post-stroke mild cognitive impairment based on principal component analysis method

目的:构建脑卒中后轻度认知障碍(PSMCI)预测模型,精准评估患者轻度认知障碍风险程度,为临床识别高危风险人群提供依据。方法:选择2020年1月至2022年6月在东南大学附属中大医院神经内科住院的脑卒中患者,采用听觉词语学习测验、逻辑记忆测验、Rey复杂图形测验、言语流畅性测验、图片命名测验、数字广度测验、数字符号转换测验、连线测验、Stroop色词测验、词语流畅测验等16个指标进行调查,采用主成分分析法对数据进行统计分析,确定脑卒中患者存在的认知障碍风险所包涵维度,形成风险预测模型。结果:脑卒中患者认知障碍风险存在3个维度,对总方差的累积贡献率是59.227%,分别为记忆功能和注意水平、记忆功能和执行能力、患者抑郁状态和日常生活自理能力,并形成基于3个维度在原变量中所占的百分比认知障碍风险的评价计算公式。结论:本研究构建的PSMCI患者风险预测模型,能客观评价患者存在的认知障碍风险并进行分级,有助于临床早期识别不同程度风险人群,针对性地提供预防与干预措施。

Objective:This study aims to construct a risk prediction model for post-stroke mild cognitive impairment(PSMCI)patients,to accurately assess the degree of risk of mild cognitive impairment,and to provide a basis for clinical identification of high-risk groups.Methods:Stroke patients admitted to the Department of Neurology,Zhongda Hospital Affiliated to Southeast University,from January 2020 to June 2022 were recruited.16 indicators such as the Auditory-Verbal Learning Test(AVLT),Logical Memory Test(LMT),Rey-Osterrieth Complex Figure Test(ROCF),Verbal Fluency Test(VFT),Picture-Naming Test,Digit Span Test(DST),Symbol Digit Modalities Test(SDMT),Trail Making Test(TMT),Stroop Color Word Test(SCWT)and Word Fluency Test were used to investigate patients.Principal component analysis method was used to determine the dimensions of risk for PSMCI patients and to develop a risk prediction model.Results:Three dimensions of risk for PSMCI patients were identified,with a cumulative contribution of 59.227%to the total variance:memory function and attention level,memory function and executive ability,and depressive state and activities of daily living.The evaluation formula of cognitive impairment risk was formed based on the percentage of the three dimensions in the original variable.Conclusion:The risk prediction model for PSMCI patients constructed in this study can objectively evaluate and grade the risk of cognitive impairment.This is helpful for the early clinical identification of stroke people with different levels of risk,and the targeted provision of preventive measures and intervention.