基于机器学习建立腹部手术术后脓毒症患者死亡风险预测模型

舒欣; 李昊洋; 李雨捷; 宋艾璘; 胡小艳; 陈芋文; 张炬; 易斌

doi:10.16016/j.2097-0927.202212045

您当前的位置：

首页 >

文章列表页 >

基于机器学习建立腹部手术术后脓毒症患者死亡风险预测模型

专题报道：围术期大数据治理与AI算法研究 | 更新时间：2026-01-09

- 基于机器学习建立腹部手术术后脓毒症患者死亡风险预测模型
- Prediction of postoperative sepsis mortality risk based on machine learning in patients undergoing abdominal surgery
- 陆军军医大学学报 2023年45卷第8期页码：732-738
- 作者机构：
  
  1.400038 重庆, 陆军军医大学(第三军医大学)第一附属医院麻醉科
  2.400038 重庆, 陆军军医大学(第三军医大学)基础医学院学员五大队
  3.400714 重庆, 中国科学院重庆绿色智能技术研究院
- 作者简介：
  
  [易斌，E-mail: yibin1974@163.com YI Bin, E-mail: yibin1974@163.com
- 基金信息：
  
  国家重点研发计划(2018YFC0116702);重庆英才计划“包干制”项目(CQYC202103080)
- DOI：10.16016/j.2097-0927.202212045
  中图分类号： R319;R619.3;R656.07
- 收稿：2022-12-06，
  
  修回：2023-01-09，
  
  纸质出版：2023-04-30
- 稿件说明：
移动端阅览
舒欣, 李昊洋, 李雨捷, 等. 基于机器学习建立腹部手术术后脓毒症患者死亡风险预测模型[J]. 陆军军医大学学报, 2023,45(8):732-738.

Xin SHU, Haoyang LI, Yujie LI, et al. Prediction of postoperative sepsis mortality risk based on machine learning in patients undergoing abdominal surgery[J]. Journal of Army Medical University, 2023, 45(8): 732-738.
舒欣, 李昊洋, 李雨捷, 等. 基于机器学习建立腹部手术术后脓毒症患者死亡风险预测模型[J]. 陆军军医大学学报, 2023,45(8):732-738. DOI： 10.16016/j.2097-0927.202212045.

Xin SHU, Haoyang LI, Yujie LI, et al. Prediction of postoperative sepsis mortality risk based on machine learning in patients undergoing abdominal surgery[J]. Journal of Army Medical University, 2023, 45(8): 732-738. DOI： 10.16016/j.2097-0927.202212045.

摘要

目的

探讨机器学习算法构建腹部手术术后脓毒症患者死亡风险预测模型的可行性。

方法

采用病例-对照研究设计方案，从公共重症监护医学信息数据库(Medical Information Mart for Intensive Care Ⅳ，MIMIC-Ⅳ v1.0)中筛选出行腹部手术后发生脓毒症的患者，研究终点事件定义为患者入院后90 d内死亡。根据死亡与否将数据集随机拆分为训练数据集(70%)与测试数据集(30%)，在训练数据集上基于Logistic回归(logistic regression，LR)、梯度提升树(gradient boosting decision tree，GBDT)、随机森林(random forest，RF)、支持向量机(support vector machine，SVM)和自适应提升(adaptive boosting，AdaBoost)算法构建预测死亡风险模型；在测试数据集上通过受试者工作曲线(receiver operating characteristic curve，ROC)和曲线下面积(area under the ROC curve，AUC)、敏感性、特异性、阳性预测值、阴性预测值、F1分数和准确率来评估模型效能。

结果

最终986例患者纳入本研究，其中251例(25.5%)患者入院后90 d内死亡，LR、GBDT、RF、SVM及AdaBoost模型的AUC依次为0.852、0.903、0.921、0.940和0.906，其中SVM的AUC最高，预测性能更好，而LR模型效能最差。

结论

基于GBDT、RF、SVM及AdaBoost这4种算法建立的腹部手术术后脓毒症死亡率预测模型的效能优于传统的LR模型，可能有助于临床决策，改善不良结局。

Abstract

Objective

To explore the feasibility of constructing prediction models of postoperative sepsis mortality risk based on machine learning in patients undergoing abdominal surgery.

Methods

A case-control trial was designed and conducted on the patients diagnosed with sepsis after abdominal surgery from Medical Information Mart for Intensive Care Ⅳ (MIMIC-Ⅳ) database

and 90-day mortality was defined as the primary endpoint event after hospitalization. The dataset was ramdomly split into training (70%) and test (30%) datasets according to wether diagnosed with postopertive sepsis or not. On the training dataset

logistic regression (LR)

gradient boosting decision tree (GBDT)

random forest (RF)

support vector machine (SVM) and adaptive boosting (AdaBoost) were used to develop the prediction model for death. The area under the receiver operating characteristic curve (AUC)

sensitivity

specificity

positive predictive value

negative predictive value

accuracy and F1 score were used for model evaluation on the test dataset.

Results

A total of 986 patients were finally analyzed

of whom 251 patients (25.5%) died within 90 d after hospitalization. The AUC values of LR

GBDT

SVM and AdaBoost prediction models were 0.852

0.903

0.921

0.940 and 0.906

respectively. The model based on SVM yielded the best AUC value

higher differentiation and better prediction performance

while LR performed the worst among them.

Conclusion

The performances of the prediction model of postoperative sepsis mortality based on GBDTT

SVM and AdaBoost are all better than that of traditional LR model

which may help to assist clinical decision making and improve adverse outcomes.

关键词

Keywords

references

SINGER M, DEUTSCHMAN C S, SEYMOUR C W, et al . The third international consensus definitions f or sepsis and septic shock (Sepsis-3)[J ] . JAMA, 2016, 315(8): 801-810. DOI: 10.1001/jama.2016.0287.

LIU V, ESCOBAR G J, GREENE J D, et al . Hospital deaths in patients with sepsis from 2 independent cohorts[J ] . JAMA, 2014, 312(1): 90-92.

PLAEKE P, DE MAN J G, COENEN S, et al . Clinical- and surgery-specific risk factors for post-operative sepsis: a systematic review and meta-analysis of over 30 million patients[J ] . Surg Today, 2020, 50(5): 427-439. DOI: 10.1007/s00595-019-01827-4.

WELEDJI E P, NGOWE M N. The challenge of intra-abdominal sepsis[J]. Int J Surg, 2013, 11(4): 290-295. DOI: 10.1016/j.ijsu.2013.02.021.

NEMATI S, HOLDER A, RAZMI F, et al . An interpretable machine learning model for accurate prediction of sepsis in the ICU[J ] . Crit Care Med, 2018, 46(4): 547-553. DOI: 10.1097/CCM.0000000000002936.

LAURITSEN S M, THIESSON B, JØRGENSEN M J, et al . The Framing of machine learning risk prediction models illustrated by evaluation of sepsis in general wards[J ] . NPJ Digit Med, 2021, 4(1): 158. DOI: 10.1038/s41746-021-00529-x.

CILLONIZ C, WARD L, MOGENSEN M L, et al . Machine-learning model for mortality prediction in patients with community-acquired pneumonia: development and validation study[J ] . Chest, 2022: S0012-S3692(22)01243-0. DOI: 10.1016/j.chest.2022.07.005.

JOHNSON A B L, POLLARD T, HORNG S, et al . MIMIC-IV (version 1.0). PhysioNet2021. Available from: https://doi.org/10.13026/7vcr-e114 https://doi.org/10.13026/7vcr-e114 ..

WANG J, ZHANG H Q, WANG J Z, et al . Feature selection using a neural network with group lasso regularization and controlled redundancy[J ] . IEEE Trans Neural Netw Learn Syst, 2021, 32(3): 1110-1123. DOI: 10.1109/TNNLS.2020.2980383.

BATEMAN B T, SCHMIDT U, BERMAN M F, et al . Temporal trends in the epidemiology of severe postoperative sepsis after elective surgery: a large, nationwide sample[J ] . Anesthesiology, 2010, 112(4): 917-925. DOI: 10.1097/ALN.0b013e3181cea3d0.

FLEISCHMANN-STRUZEK C, MELLHAMMAR L, ROSE N, et al . Incidence and mortality of hospital- and ICU-treated sepsis: results from an updated and expanded systematic review and meta-analysis[J ] . Intensive Care Med, 2020, 46(8): 1552-1562. DOI: 10.1007/s00134-020-06151-x.

刘蓬然, 霍彤彤, 陆林, 等. 人工智能在医学中的应用现状与展望[J]. 中华医学杂志, 2021, 101(44): 3677-3683. DOI: 10.3760/cma.j.cn112137-20210313-00628

LIU P R, HUO T T, LU L, et al . Application status and prospect of artificial intelligence in medicine[J ] . Natl Med J China, 2021, 101(44): 3677-3683. DOI: 10.3760/cma.j.cn112137-20210313-00628

LAURITSEN S M, KALØR M E, KONGSGAARD E L, et al . Early detection of sepsis utilizing deep learning on electronic health record event sequences[J ] . Artif Intell Med, 2020, 104: 101820. DOI: 10.1016/j.artmed.2020.101820.

BUNN C, KULSHRESTHA S, BOYDA J, et al . Applicationof machine learning to the prediction of postoperative sepsis after appendectomy[J ] . Surgery, 2021, 169(3): 671-677. DOI: 10.1016/j.surg.2020.07.045.

DONG J Z, FENG T, THAPA-CHHETRY B, et al . Machine learning model for early prediction of acute kidney injury (AKI) in pediatric critical care[J ] . Crit Care, 2021, 25(1): 288. DOI: 10.1186/s13054-021-03724-0.

UENO R, XU L Y, UEGAMI W, et al . Value of laboratory results in addition to vital signs in a machine learning algorithm to predict in-hospital cardiac arrest: a single-center retrospective cohort study[J ] . PLoS One, 2020, 15(7): e0235835. DOI: 10.1371/journal.pone.0235835.

D'ASCENZO F, DE FILIPPO O, GALLONE G, et al . Machine learning-based prediction of adverse events following an acute coronary syndrome (PRAISE): a modelling study of pooled datasets[J ] . Lancet, 2021, 397(10270): 199-207. DOI: 10.1016/S0140-6736(20)32519-8.

THORSEN-MEYER H C, NIELSEN A B, NIELSEN A P, et al . Dynamic and explainable machine learning prediction of mortality in patients in the intensive care unit: a retrospective study of high-frequency data in electronic patient records[J ] . Lancet Digit Health, 2020, 2(4): e179-e191. DOI: 10.1016/S2589-7500(20)30018-2.

齐霜, 徐浩然, 胡婕, 等. 基于机器学习的重症监护病房脓毒症患者早期死亡风险预测模型[J]. 解放军医学院学报, 2021, 42(2): 150-155, 181. DOI: 10.3969/j.issn.2095-5227.2021.02.006.

QI S, XU H R, HU J, et al . Early mortality risk prediction model for sepsis patients in intensive care unit based on machine learning[J ] . Acad J Chin Pla Med Sch, 2021, 42(2): 150-155, 181. DOI: 10.3969/j.issn.2095-5227.2021.02.006.

张颖莹, 刘怡果, 赵丹, 等. 基于机器学习建立脓毒症心肾综合征患者早期死亡风险预测模型[J]. 中华肾脏病杂志, 2022, 38(9)785-793. DOI: 10.3760/cma.j.cn441217-20211126-00113

ZHANG Y Y, LIU Y G, ZHAO D, et al . Early mortality risk prediction models for patients with sepsis-induced cardiorenal syndrome based on machine learning[J ] . Chin J Nephrol, 2022, 38(9): 785-793. DOI: 10.3760/cma.j.cn441217-20211126-00113

XIAO J, DING R F, XU X L, et al . Comparison and development of machine learning tools in the prediction of chronic kidney disease progression[J ] . J Transl Med, 2019, 17(1): 119. DOI: 10.1186/s12967-019-1860-0.

RYOO S M, LEE J, LEE Y S, et al . Lactate level versus lactate clearance for predicting mortality in patients with septic shock defined by Sepsis-3[J ] . Crit Care Med, 2018, 46(6): e489-e495. DOI: 10.1097/CCM.000000000000 3030.

JONES A E, SHAPIRO N I, TRZECIAK S, et al . Lactate clearance vs central venous oxygen saturation as goals of early sepsis therapy: a randomized clinical trial[J ] . JAMA, 2010, 303(8): 739-746. DOI: 10.1001/jama.2010.158.

浏览量

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