Page 84 - 南京医科大学自然版
P. 84

南京医科大学学报(自然科学版)                                  第44卷第5期
               ·672 ·                     Journal of Nanjing Medical University(Natural Sciences)   2024年5月


             ·临床研究·

              基于可解释机器学习构建脑卒中患者日常生活自理能力风险

              预测模型



              叶   倩 ,杨 云 ,徐文韬 ,刘玲玲             1*
                     1,2
                                       2
                              1,3
               南京医科大学附属第一医院康复医学中心,江苏                   南京    210029;南京中医药大学针灸推拿学院,江苏               南京   210029;
              1                                                    2
               南京师范大学心理学院,江苏 南京             210023
              3

             [摘    要] 目的:利用机器学习算法预测影响脑卒中患者日常生活自理能力(activities of daily living,ADL)的风险因素,为其
              ADL管理决策提供参考。方法:对2015年1月—2019年2月在南京医科大学附属第一医院康复医学中心治疗的423例脑卒中患
              者进行回顾性分析。根据Barthel指数(Barthel index,BI)评定量表,将患者分为ADL较好组(BI≥60分)和ADL较差组(BI<60分),
              并进行数据预处理。采用共线性诊断及最小绝对收缩和选择算子(least absolute shrinkage and selection operator,LASSO)筛选
              特征变量。选择逻辑回归、支持向量机、随机森林(random forest,RF)、极限梯度提升及K最近邻5种机器学习算法进行预测建
              模,十倍交叉验证后,使用受试者工作特征曲线、受试者工作特征曲线下面积(area under curve,AUC)、精确召回率曲线、精确
              召回率曲线下的面积(area under the precision recall curve,PRAUC)、准确率、灵敏度、特异度分别对模型进行综合评估,引入
              Shapley 加性解释(Shapley additive explanation,SHAP)对最优机器学习模型进行可解释化处理。结果:经LASSO 回归分析后,
              确定16个特征变量用于构建机器学习模型。RF模型具有最高的AUC(0.74)、PRAUC(0.64)、准确率(0.97)、灵敏度(0.75)和特
              异度(0.97)。SHAP 模型解释性分析显示,对 ADL 贡献度前 5 的特征中,Brunnstrom 分期(下肢)的影响最为显著,其次是
              Brunnstrom 分期(上肢)、D⁃二聚体、血清白蛋白水平及年龄。结论:RF 模型预测脑卒中患者 ADL 的效能最优,为脑卒中患者
              ADL管理决策提供了有价值的参考。
             [关键词] 机器学习;预测模型;脑卒中;日常生活自理能力
             [中图分类号] R743.3                    [文献标志码] A                      [文章编号] 1007⁃4368(2024)05⁃672⁃09
              doi:10.7655/NYDXBNSN240009



              Constructing a prediction model for stroke patients’activities of daily living risk based on
              interpretable machine learning

                                              2
                                 1,3
                     1,2
              YE Qian ,YANG Yun ,XU Wentao ,LIU Lingling  1*
               Center of Rehabilitation Medicine,the First Affiliated Hospital of Nanjing Medical University,Nanjing 210029;
              1
               College of Acupuncture and Massage,Nanjing University of Chinese Medicine,Nanjing 210029;School of
              2                                                                                       3
              Psychology,Nanjing Normal University,Nanjing 210023,China
             [Abstract] Objective:To utilize machine learning algorithms to predict risk factors affecting the activities of daily living(ADL)of
              stroke patients,providing references for their ADL management decisions. Methods:A retrospective analysis was conducted on 423
              stroke patients treated at the Rehabilitation Medicine Center of the First Affiliated Hospital of Nanjing Medical University from January
              2015 to February 2019. Patients were categorized into a better ADL group(BI ≥ 60 points)and a poorer ADL group(BI <60 points)
              based on the Barthel Index(BI)assessment scale,and data preprocessing was performed. Feature variables were selected using
              colinearity diagnostics and the least absolute shrinkage and selection operator(LASSO). Logistic regression(LR),support vector
              machine(SVM),random forest(RF),extreme gradient boosting(XGBoost),and K nearest neighbor(KNN)were selected as the five
              machine learning algorithms for predictive modeling. Afterten⁃fold cross⁃validation,the models were comprehensively evalutated using
              receiver operating characteristic(ROC)curves,area under aerue(AUC),precision recall(PR)curves,area under the precision recall

             [基金项目] 国家自然科学基金(82104993)
              ∗
              通信作者(Corresponding author),E⁃mail: ch600lll@163.com
   79   80   81   82   83   84   85   86   87   88   89