• 首页
  • 期刊简介
  • 编委会
  • 投稿指南
  • 收录情况
  • 杂志订阅
  • 联系我们
引用本文:徐乃岳,凌 晨,刘 坤.基于可解释贝叶斯加权模型的ICU 急性肾损伤患者死亡风险预测[J].软件工程,2024,27(6):53-58.【点击复制】
【打印本页】   【下载PDF全文】   【查看/发表评论】  【下载PDF阅读器】  
←前一篇|后一篇→ 过刊浏览
分享到: 微信 更多
基于可解释贝叶斯加权模型的ICU 急性肾损伤患者死亡风险预测
徐乃岳1, 凌 晨2, 刘 坤1
(1.上海理工大学健康科学与工程学院, 上海 200093;
2.上海健康医学院医疗器械学院, 上海 201318)
xunaiyue21@163.com; lingc@sumhs.edu.cn; lkun11111@163.com
摘 要: 基于贝叶斯网络构建贝叶斯加权模型,进行重症监护病房(Intensive Care Unit,ICU)急性肾损伤患者死亡风险预测。以MIMIC-Ⅲ(Medical Information Mark for Intensive Care Ⅲ)数据库中急性肾损伤患者为研究对象,建立基础贝叶斯分类器,采用AUC(Area Under Curve)和Accuracy进行混合加权计算的集成策略构建贝叶斯加权模型。实验结果表明,贝叶斯加权模型的AUC值为80.8%、Accuracy值为73.2%、F1-score值为72.4%,预测效果优于单独的贝叶斯网络模型、逻辑回归、支持向量机和随机森林。贝叶斯加权模型具有可解释的概率推理流程,对ICU急性肾损伤患者的死亡风险预测有一定的参考价值。
关键词: 贝叶斯网络;急性肾损伤;死亡风险;模型解释;集成模型
中图分类号: TP391    文献标识码: A
基金项目: 科技部国家重点研发计划“主动健康和老龄化科技应对”重点专项(2020YFC2008700);国家自然科学基金项目:基于贝叶斯网络预测ICU术后患者死亡风险的方法研究(82072228)
Prediction of Mortality Risk for ICU Patients with Acute Kidney Injury Based on an Interpretable Bayesian Weighted Model
XU Naiyue1, LING Chen2, LIU Kun1
(1.School of Health Science and Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China;
2.Medical Instrumentation College, Shanghai University of Medicine and Health Sciences, Shanghai 201318, China)
xunaiyue21@163.com; lingc@sumhs.edu.cn; lkun11111@163.com
Abstract: This paper proposes to predict the mortality risk of patients with acute kidney injury in the Intensive Care Unit (ICU) with a Bayesian weighted model constructed based on Bayesian network. Taking acute kidney injury patients from the MIMIC-Ⅲ (Medical Information Mart for Intensive Care Ⅲ) database as the research subjects, a basic Bayesian classifier is established. An integrated strategy combining AUC (Area Under Curve) and Accuracy through mixed weighting calculation is employed to build the Bayesian weighted model. Experimental results show that the Bayesian weighted model achieves an AUC value of 80.8% , an Accuracy value of 73.2% , and an F1-score value of 72.4% , outperforming standalone Bayesian network models, Logistic Regression, Support Vector Machines (SVM), and Random Forests. The Bayesian weighted model has an interpretable probability inference process and provides valuable insights for predicting mortality risk for ICU patients with acute kidney injury.
Keywords: Bayesian network; acute kidney injury; mortality risk; model interpretation; integrated model


版权所有:软件工程杂志社
地址:辽宁省沈阳市浑南区新秀街2号 邮政编码:110179
电话:0411-84767887 传真:0411-84835089 Email:semagazine@neusoft.edu.cn
备案号:辽ICP备17007376号-1
技术支持:北京勤云科技发展有限公司

用微信扫一扫

用微信扫一扫