基于机器学习构建急性心力衰竭患者易损期死亡或再入院风险预测模型

曾竟; 何小龙; 胡华娟; 罗晓宇; 郭志念; 陈运龙; 王敏; 王江

doi:10.16016/j.2097-0927.202312147

您当前的位置：

首页 >

文章列表页 >

基于机器学习构建急性心力衰竭患者易损期死亡或再入院风险预测模型

临床医学 | 更新时间：2026-01-09

- 基于机器学习构建急性心力衰竭患者易损期死亡或再入院风险预测模型
- Construction of risk prediction model for predicting death or readmission in acute heart failure patients during vulnerable phase based on machine learning
- 陆军军医大学学报 2024年46卷第7期页码：738-745
- 作者机构：
  
  1.400037 重庆, 陆军军医大学(第三军医大学)第二附属医院心血管内科, 全军心血管病研究所
  2.611130 成都, 西南财经大学统计学院
- 作者简介：
  
  [王江，E-mail：1530579054@qq.comWANG Jiang, E-mail: 1530579054@qq.com
- 基金信息：
  
  重庆市科卫联合医学重点项目(2023ZDXM035)
- DOI：10.16016/j.2097-0927.202312147
  中图分类号： R181.32;R319;R541.6
- 收稿：2023-12-21，
  
  修回：2024-02-08，
  
  纸质出版：2024-04-15
- 稿件说明：
移动端阅览
曾竟, 何小龙, 胡华娟, 等. 基于机器学习构建急性心力衰竭患者易损期死亡或再入院风险预测模型[J]. 陆军军医大学学报, 2024,46(7):738-745.

Jing ZENG, Xiaolong HE, Huajuan HU, et al. Construction of risk prediction model for predicting death or readmission in acute heart failure patients during vulnerable phase based on machine learning[J]. Journal of Army Medical University, 2024, 46(7): 738-745.
曾竟, 何小龙, 胡华娟, 等. 基于机器学习构建急性心力衰竭患者易损期死亡或再入院风险预测模型[J]. 陆军军医大学学报, 2024,46(7):738-745. DOI： 10.16016/j.2097-0927.202312147.

Jing ZENG, Xiaolong HE, Huajuan HU, et al. Construction of risk prediction model for predicting death or readmission in acute heart failure patients during vulnerable phase based on machine learning[J]. Journal of Army Medical University, 2024, 46(7): 738-745. DOI： 10.16016/j.2097-0927.202312147.

摘要

目的

应用不同机器学习算法构建急性心力衰竭(acute heart failure

AHF)患者易损期死亡或再入院的风险预测模型

并筛选出最优模型。

方法

选取2019年10月至2021年7月于陆军军医大学第二附属医院心血管内科住院治疗的651例AHF患者为研究对象

收集入院生命体征、合并症和实验室检查结果等临床资料。复合终点事件定义为AHF患者出院后3个月内发生全因死亡或心衰加重再入院。采用简单随机抽样法将研究对象按8:2拆分为训练集(521例)和测试集(130例)

基于逻辑回归(logistic regression

LR)、随机森林(random forest

RF)、决策树(decision tree

DT)、轻量梯度提升(light gradient boosting machine

LGBM)、极端梯度提升(extreme gradient boosting

XGBoost)和神经网络(neural network

NN)6种机器学习算法分别构建预测模型。采用受试者工作特征曲线(receiver operating characteristic

ROC)和临床决策曲线(decision curve analysis

DCA)对模型的预测性能和临床获益进行评价

使用Shapley加性解释(Shapley additive explanation

SHAP)算法评估不同临床特征对模型的影响。

结果

651例AHF患者中发生复合终点事件203例(31.2%)。ROC曲线分析显示

LR、RF、DT、LGBM、XGBoost和NN模型的曲线下面积(area under curve

AUC)依次为0.707、0.756、0.616、0.677、0.768、0.681

XGBoost模型的AUC最高

DCA曲线中XGBoost模型的临床决策净获益也更大

整体预测效能最佳。SHAP算法分析得出

影响XGBoost模型输出结果的重要临床特征分别为血清尿酸、D-二聚体、平均动脉压、B型利钠肽、左房前后径、体质量指数和NYHA分级。

结论

XGBoost模型预测急性心力衰竭患者易损期死亡或再入院风险效果最佳。

Abstract

Objective

To construct risk prediction models of death or readmission in patients with acute heart failure (AHF) during the vulnerable phase based on machine learning algorithms and screen the optimal model.

Methods

A total of 651 AHF patients with admitted to Department of Cardiology of the Second Affiliated Hospital of Army Medical University from October 2019 to July 2021 were included.The clinical data consisting of admission vital signs

comorbidities and laboratory results were collected from electronic medical records.The composite endpoint was defined as all-cause death or readmission for worsening heart failure within 3 months after discharge.The patients were divided into a training set (521 patients) and a test set (130 patients) in a ratio of 8:2 through the simple random sampling.Six machine learning models were developed

including logistic regression (LR)

random forest (RF)

decision tree (DT)

light gradient boosting machine (LGBM)

extreme gradient boosting (XGBoost) and neural networks (NN).Receiver operating characteristic (ROC) curve and decision curve analysis (DCA) were used to evaluate the predictive performance and clinical benefit of the models.Shapley additive explanation (SHAP) was used to explain and evaluate the effect of different clinical characteristics on the models.

Results

A total of 651 AHF patients were included

of whom 203 patients (31.2%) died or were readmitted during the vulnerable phase.ROC curve analysis showed that the AUC values of the LR

LGBM

XGBoost and NN model were 0.707

0.756

0.616

0.677

0.768 and 0.681

respectively.The XGBoost model had the highest AUC value.DCA showed that the XGBoost model exhibited greater clinical net benefit compared with other models

with the best predictive performance.SHAP algorithm analysis showed that the clinical features that had the greatest impact on the output of the model were serum uric acid

D-dimer

mean arterial pressure

B-type natriuretic peptide

left atrial diameter

body mass index

and New York Heart Association (NYHA) classification.

Conclusion

The XGBoost model has the best predictive performance in predicting the risk of death or readmission of AHF patients during the vulnerable phase.

关键词

Keywords

references

ARRIGO M, JESSUP M, MULLENS W, et al . Acute heart failure[J ] . Nat Rev Dis Primers, 2020, 6(1): 16. DOI: 10.1038/s41572-020-0151-7.

HAO G, WANG X, CHEN Z, et al . Prevalence of heart failure and left ventricular dysfunction in China: the China hypertension survey, 2012-2015[J ] . Eur J Heart Fail, 2019, 21(11): 1329-1337. DOI: 10.1002/ejhf.1629.

DESAI A S, STEVENSON L W. Rehospitalization for heart failure: predict or prevent?[J]. Circulation, 2012, 126(4): 501-506. DOI: 10.1161/CIRCULATIONAHA.112.125435.

GHEORGHIADE M, VADUGANATHAN M, FONAROW G C, et al . Rehospitalization for heart failure[J ] . J Am Coll Cardiol, 2013, 61(4): 391-403. DOI: 10.1016/j.jacc.2012.09.038.

GREENE S J, FONAROW G C, VADUGANATHAN M, et al . The vulnerable phase after hospitalization for heart failure[J ] . Nat Rev Cardiol, 2015, 12(4): 220-229. DOI: 10.1038/nrcardio.2015.14.

中华医学会心血管病学分会心力衰竭学组, 中国医师协会心力衰竭专业委员会, 中华心血管病杂志编辑委员会. 中国心力衰竭诊断和治疗指南2018[J]. 中华心血管病杂志, 2018, 46(10): 760-789. DOI: 10.3760/cma.j.issn.0253-3758.2018.10.004.

Heart Failure Group of Chinese Society of Cardiology of Chinese Medical Association, Chinese Heart Failure Association of Chinese Medical Doctor Association, Editorial Board of Chinese Journal of Cardiology . China guidelines for diagnosis and treatment of heart failure 2018[J ] . Chin J Cardiol, 2018, 46(10): 760-789. DOI: 10.3760/cma.j.issn.0253-3758.2018.10.004.

石珂, 贾永平. 心力衰竭易损期再入院率和死亡评估新进展[J]. 中西医结合心脑血管病杂志, 2019, 17(18): 2783-2786. DOI: 10.12102/j.issn.1672-1349.2019.18.016.

SHI K, JIA Y P. New progress of readmission rate and death assessment in vulnerable period of heart failure[J]. Chin J Integr Med Cardio Cerebrovasc Dis, 2019, 17(18): 2783-2786. DOI: 10.12102/j.issn.1672-1349.2019.18.016.

RU B S, TAN X, LIU Y, et al . Comparison of machine learning algorithms for predicting hospital readmissions and worsening heart failure events in patients with heart failure with reduced ejection fraction: mode ling study[J ] . JMIR Form Res, 2023, 7: e41775. DOI: 10.2196/41775.

SABOURI M, RAJABI A B, HAJIANFAR G, et al . Machine learning based readmission and mortality prediction in heart failure patients[J ] . Sci Rep, 2023, 13(1): 18671. DOI: 10.1038/s41598-023-45925-3.[PubMed ]

伍亚舟, 陈锡程, 易东. 人工智能在临床领域的研究进展及前景展望[J]. 陆军军医大学学报, 2022, 44(1): 89-102. DOI: 10.16016/j.1000-5404.202112092.

WU Y Z, CHEN X C, YI D. Advances and perspective of artificial intelligence in clinical area[J]. J Army Med Univ, 2022, 44(1): 89-102. DOI: 10.16016/j.1000-5404.202112092.

DESAI R J, WANG S V, VADUGANATHAN M, et al . Comparison of machine learning methods with traditional models for use of administrative claims with electronic medical records to predict heart failure outcomes[J ] . JAMA Netw Open, 2020, 3(1): e1918962. DOI: 10.1001/jamanetworkopen.2019.18962.

SHIN S, AUSTIN P C, ROSS H J, et al . Machine learning vs . conventional statistical models for predicting heart failure readmission and mortality[J ] . ESC Heart Fail, 2021, 8(1): 106-115. DOI: 10.1002/ehf2.13073..

LUNDBERG S M, ERION G, CHEN H, et al . From local explanations to global understanding with explainable AI for trees[J ] . Nat Mach Intell, 2020, 2(1): 56-67. DOI: 10.1038/s42256-019-0138-9.

PETCH J, DI S, NELSON W. Opening the black box: the promise and limitations of explainable machine learning in cardiology[J]. Can J Cardiol, 2022, 38(2): 204-213. DOI: 10.1016/j.cjca.2021.09.004.

SOLOMON S D, DOBSON J, POCOCK S, et al . Influence of nonfatal hospitalization for heart failure on subsequent mortality in patients with chronic heart failure[J ] . Circulation, 2007, 116(13): 1482-1487. DOI: 10.1161/CIRCULATIONAHA.107.696906.

王薇, 雷璐碧, 赵倩, 等. 急性心力衰竭再入院及死亡风险预测模型研究进展[J]. 中华流行病学杂志, 2023, 44(12): 2005-2011. DOI: 10.3760/cma.j.cn112338-20230527-00336.

WANG W, LEI L B, ZHAO Q, et al. Progress in research of models for predicting the risk of readmission and mortality among patients with acute heart failure[J ] . Chin J Epidemiol, 2023, 44(12): 2005-2011. DOI: 10.3760/cma.j.cn112338-20230527-00336.

ZHUANG B, SHEN T, LI D J, et al . A model for the prediction of mortality and hospitalization in Chinese heart failure patients[J ] . Front Cardiovasc Med, 2021, 8: 761605. DOI: 10.3389/fcvm.2021.761605.

YANG M X, TAO L Y, AN H, et al . A novel nomogram to predict all-cause readmission or death risk in Chinese elderly patients with heart failure[J ] . ESC Heart Fail, 2020, 7(3): 1015-1024. DOI: 10.1002/ehf2.12703.

BAZOUKIS G, STAVRAKIS S, ZHOU J D, et al . Machine learning versus conventional clinical methods in guiding management of heart failure patients—a systematic review[J ] . Heart Fail Rev, 2021, 26(1): 23-34. DOI: 10.1007/s10741-020-10007-3.

王齐, 陈康玉, 严激. 基于人工智能的心力衰竭预后研究[J]. 中华心血管病杂志, 2022, 50(7): 637-645. DOI: 10.3760/cma.j.cn112148-20210721-00606.

WANG Q, CHEN K Y, YAN J. Artificial intelligence based prognostic studies in patients with heart failure[J]. Chin J Cardiol, 2022, 50(7): 637-645. DOI: 10.3760/cma.j.cn112148-20210721-00606.

PALAZZUOLI A, RUOCCO G, DE VIVO O, et al . Prevalence of hyperuricemia in patients with acute heart failure with either reduced or preserved ejection fraction[J ] . Am J Cardiol, 2017, 120(7): 1146-1150. DOI: 10.1016/j.amjcard.2017.06.057.

PALAZZUOLI A, RUOCCO G, PELLEGRINI M, et al . Prognostic significance of hyperuricemia in patients with acute heart failure[J ] . Am J Cardiol, 2016, 117(10): 1616-1621. DOI: 10.1016/j.amjcard.2016.02.039.

WEI F F, CHEN X W, CHENG W, et al . Associations of long-term mortality with serum uric acid at admission in acute decompensated heart failure with different phenotypes[J ] . Nutr Metab Cardiovasc Dis, 2023, 33(10): 1998-2005. DOI: 10.1016/j.numecd.2023.06.007.

LEVY W C, MOZAFFARIAN D, LINKER D T, et al . The Seattle Heart Failure Model: prediction of survival in heart failure[J ] . Circulation, 2006, 113(11): 1424-1433. DOI: 10.1161/CIRCULATIONAHA.105.584102.

DOEHNER W, FRENNEAUX M, ANKER S D. Metabolic impairment in heart failure: the myocardial and systemic perspective[J]. J Am Coll Cardiol, 2014, 64(13): 1388-1400. DOI: 10.1016/j.jacc.2014.04.083.

王晓琪, 苏冠华. 高尿酸血症和心力衰竭的病理生理机制、治疗和预后价值[J]. 心血管病学进展, 2021, 42(9): 780-783. DOI: 10.16806/j.cnki.issn.1004-3934.2021.09.004.

WANG X Q, SU G H. Pathophysiological mechanism, treatment and prognostic value of hyperuricemia and heart failure[J]. Adv Cardiovasc Dis, 2021, 42(9): 780-783. DOI: 10.16806/j.cnki.issn.1004-3934.2021.09.004.

张加良, 高蓉蓉, 李新立, 等. D-二聚体评估急性心力衰竭短期预后的临床意义[J]. 上海大学学报(自然科学版), 2016, 22(3): 376-380. DOI: 10.3969/j.issn.1007-2861.2016.03.020.

ZHANG J L, GAO R R, LI X L, et al . Clinic significance of D-dimer in short-term prognosis of patients with acute heart failure[J ] . J Shanghai Univ Nat Sci Ed, 2016, 22(3): 376-380. DOI: 10.3969/j.issn.1007-2861.2016.03.020.

HUANG L Y, LIANG L, TIAN P C, et al . D-dimer and outcomes in hospitalized heart failure patients across the ejection fraction phenotypes[J ] . ESC Heart Fail, 2022, 9(5): 3060-3070. DOI: 10.1002/ehf2.14049.

LIU M H, CHEN X H, ZHANG S Z, et al . Long-term visit-to-visit mean arterial pressure variability and the risk of heart failure and all-cause mortality[J ] . Front Cardiovasc Med, 2021, 8: 665117. DOI: 10.3389/fcvm.2021.665117.

JONES N R, ORDÓÑEZ-MENA J M, ROALFE A K, et al . Body mass index and survival in people with heart failure[J ] . Heart, 2023, 109(20): 1542-1549. DOI: 10.1136/heartjnl-2023-322459.

浏览量

1053

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

心血管手术患者术后呼吸衰竭的危险因素及预测模型构建

基于增强静脉期CT放射组学的甲状腺乳头状癌中央区淋巴结转移预测模型构建与验证

基于5种机器学习算法构建的怀孕概率预测模型及其性能比较

高原脱适应症发生风险预测模型的构建与验证

基于机器学习算法构建组织学绒毛膜羊膜炎的预测模型