Establishment of risk prediction model for postoperative liver injury after non-liver surgery based on different machine learning algorithms

Yizhu SUN; Yujie LI; Hao LIANG; Xiang LIU; Jiahao HUANG; Xin SHU; Ailin SONG; Zhiyong YANG; Bin YI

doi:10.16016/j.2097-0927.202312018

您当前的位置：

首页 >

文章列表页 >

Establishment of risk prediction model for postoperative liver injury after non-liver surgery based on different machine learning algorithms

Clinical Medicine | 更新时间：2026-01-09

- Establishment of risk prediction model for postoperative liver injury after non-liver surgery based on different machine learning algorithms
- Journal of Army Medical University Vol. 46, Issue 7, Pages: 760-767(2024)
- 作者机构：
  
  400038 重庆, 陆军军医大学(第三军医大学)第一附属医院麻醉科
- 作者简介：
- 基金信息：
  
  the National Key Research and Development Program(2018YFC0116702)
- DOI：10.16016/j.2097-0927.202312018
  CLC： R181.32;R319;R575
- Received：04 December 2023，
  
  Revised：2024-02-09，
  
  Published：15 April 2024
- 稿件说明：
移动端阅览
Yizhu SUN, Yujie LI, Hao LIANG, et al. Establishment of risk prediction model for postoperative liver injury after non-liver surgery based on different machine learning algorithms[J]. Journal of Army Medical University, 2024, 46(7): 760-767.
DOI：

Yizhu SUN, Yujie LI, Hao LIANG, et al. Establishment of risk prediction model for postoperative liver injury after non-liver surgery based on different machine learning algorithms[J]. Journal of Army Medical University, 2024, 46(7): 760-767. DOI： 10.16016/j.2097-0927.202312018.

摘要

目的

基于术前及术中用药指标建立一个预测非肝脏手术患者术后肝损伤的机器学习预测模型。

方法

采用病例-对照研究设计方案

选择国内3家大型综合医院开发的数据库中2014年1月至2022年9月行非肝脏手术术后出现肝损伤患者315例

按阳性: 阴性为1:3的比例随机匹配同期非肝脏手术术后未发生肝损伤的928例病例作为阴性对照病例。将这1 243例患者根据R语言设置的代码以7:3的比例随机拆分为建模组(

=869)和验证组(

=374)。利用患者术前临床指标(基本信息、病史、相关评分表、手术相关信息和实验室检查)及术中用药情况基于k-近邻(k-nearest neighbor

KNN)、支持向量机(support vector machine linear

SVM)、逻辑回归(logistic regression

LR)、极端梯度提升(extreme gradient boosting

XGBoost)4种机器学习算法建立预测非肝脏手术术后肝损伤的模型

并在验证组通过受试者工作特征曲线(receiver operating characteristic

ROC)、精准-召回曲线(percision-recall curve

P-R)、临床决策分析曲线(decision curve analysis

DCA)、Kappa值、灵敏度、特异度、Brier分数、F

分数对模型效能进行评估。

结果

4种机器学习算法建立的预测非肝脏手术术后肝损伤的模型以XGBoost算法最优

验证组中受试者工作特征曲线下面积(area under the receiver operating characteristic curve

AUROC)为0.916(95%

: 0.883~0.949)

精准-召回曲线下面积(area under the precision-recall curve

AUPRC)为0.841

Brier评分0.097

灵敏度和特异度分别为78.95%和87.10%。

结论

基于XGBoost算法建立的非肝脏手术术后肝损伤预测模型对术后肝损伤的发生有良好的预测能力。

Abstract

Objective

To construct a machine learning prediction model for postoperative liver injury in patients with non-liver surgery based on preoperative and intraoperative medication indicators.

Methods

A case-control study was conducted on 315 patients with liver injury after non-liver surgery selected from the databases developed by 3 large general hospitals from January 2014 to Sept

ember 2022.With the positive/negative ratio of 1:3

928 cases in corresponding period with non-liver surgery and without liver injury were randomly matched as negative control cases.These 1 243 patients were randomly divided into the modeling group (

=869) and the validation group (

=374) in a ratio of 7:3 using the R language setting code.Preoperative clinical indicators (basic information

medical history

relevant scale score

surgical information and results of laboratory tests) and intraoperative medication were used to construct the prediction model for liver injury after non-liver surgery based on 4 machine learning algorithms

k-nearest neighbor (KNN)

support vector machine linear (SVM)

logic regression (LR) and extreme gradient boosting (XGBoost).In the validation group

receiver operating characteristic (ROC) curve

precision-recall curve (P-R)

decision curve analysis (DCA) curve

Kappa value

sensitivity

specificity

Brier score

and F

score were applied to evaluate the efficacy of model.

Results

The model established by 4 machine learning algorithms to predict postoperative liver injury after non-liver surgery was optimal using the XGBoost algorithm.The area under the receiver operating characteristic curve (AUROC) was 0.916(95%

: 0.883~0.949)

area under the precision-recall curve (AUPRC) was 0.841

Brier score was 0.097

and sensitivity and specificity was 78.95% and 87.10%

respectively.

Conclusion

The postoperative liver injury prediction model for non-liver surgery based on the XGBoost algorithm has effective prediction for the occurrence of postoperative liver injury.

关键词

Keywords

references

栾正刚, 马晓春. 围手术期肝功能障碍类型及处理[J]. 中国实用外科杂志, 2014, 34(2): 123-126.

LUAN Z G, MA X C. Perioperative liver dysfunction: types and management[J]. Chin J Pract Surg, 2014, 34(2): 123-126.

CHETTY G, SHARPE D A, NANDI J, et al . Liver blood flow during cardiac surgery[J ] . Perfusion, 2004, 19(3): 153-156. DOI: 10.1191/0267659104pf735oa.

ALSHAHRANI A S, GONG G S, YOO M W. Comparison of long-term survival and immediate postoperative liver function after laparoscopic and open distal gastrectomy for early gastric cancer patients with liver cirrhosis[J]. Gastric Cancer, 2017, 20(4): 744-751. DOI: 10.1007/s10120-016-0675-4.

WANG H J, BAO Q Q, CAO D H, et al . Characteristics of computed tomography images for patients with acute liver injury caused by sepsis under deep learning algorithm[J ] . Contrast Media Mol Imaging, 2022, 2022: 9322196. DOI: 10.1155/2022/9322196.

ESMAILY H, TAYEFI M, DOOSTI H, et al . A comparison between decision tree and random forest in determining the risk factors associated with type 2 diabetes[J ] . J Res Health Sci, 2018, 18(2): e00412. [PubMed ]

WANG J T, ZHENG T L, LIAO Y, et al . Machine learning prediction model for post-hepatectomy liver failure in hepatocellular carcinoma: a multicenter study[J ] . Front Oncol, 2022, 12: 986867. DOI: 10.3389/fonc.2022.986867.

DIAB M, SPONHOLZ C, VON LOEFFELHOLZ C, et al . Impact of perioperative liver dysfunction on in-hospital mortality and long-term survival in infective endocarditis patients[J ] . Infection, 2017, 45(6): 857-866. DOI: 10.1007/s15010-017-1064-6.

GARCÍA-ORDÁS M T, BENAVIDES C, BENÍTEZ-ANDRADES J A, et al . Diabetes detection using deep learning techniques with oversampling and feature augmentation[J ] . Comput Methods Programs Biomed, 2021, 202: 105968. DOI: 10.1016/j.cmpb.2021.105968.

ÇAGLı K, BAȘAR F N, TOK D, et al . How to interpret liver function tests in heart failure patients?[J ] . Turk J Gastroenterol, 2015, 26(3): 197-203. DOI: 10.5152/tjg.2015.0086.

LIANG S H, LIU Y X, ZHANG B W, et al . Postoperative hepatic dysfunction after frozen elephant trunk for type A aortic dissection[J ] .Front Cardiovasc Med, 2021, 8: 739606. DOI: 10.3389/fcvm.2021.739606.

LIN X F, XIE L F, JIANG D B, et al . Hepatic dysfunction and adverse outcomes after total arch repair of acute type A aortic dissection: application of the MELD-XI score[J ] . BMC Cardiovasc Disord, 2022, 22(1): 491. DOI: 10.1186/s12872-022-02934-w.

黄海嵘, 景华. 心肺转流术后肝损伤机制的研究进展[J]. 医学研究生学报, 2007, 20(8): 885-889. DOI: 10.16571/j.cnki.1008-8199.2007.08.011.

HUANG H R, JING H. Research progress of the mechanism of hepatic injury after cardiopulmonary bypass[J]. J Med Postgrad, 2007, 20(8): 885-889. DOI: 10.16571/j.cnki.1008-8199.2007.08.011.

CHEUNG A T, CRUZ-SHIAVONE G E, MENG Q C, et al . Cardiopulmonary bypass, hemolysis, and nitroprusside-induced cyanide production[J ] . Anesth Analg, 2007, 105(1): 29-33. DOI: 10.1213/01.ane.0000264078.34514.32.

LIN J, MOORE D, HOCKEY B, et al . Drug-induced hepatotoxicity: incidence of abnormal liver function tests consistent with volatile anaesthetic hepatitis in trauma patients[J ] . Liver Int, 2014, 34(4): 576-582. DOI: 10.1111/liv.12278.

NICOLL A, MOORE D, NJOKU D, et al . Repeated exposure to modern volatile anaesthetics may cause chronic hepatitis as well as acute liver injury[J ] . BMJ Case Rep, 2012, 2012: bcr2012006543. DOI: 10.1136/bcr-2012-006543.

WANG Y, JI C P, WANG Y, et al . Predicting postoperative liver cancer death outcomes with machine learning[J ] . Curr Med Res Opin, 2021, 37(4): 629-634. DOI: 10.1080/03007995.2021.1885361.

HOLLINGER A, SIEGEMUND M, GOETTEL N, et al . Postoperative delirium in cardiac surgery: an unavoidable menace?[J ] . J Cardiothorac Vasc Anesth, 2015, 29(6): 1677-1687. DOI: 10.1053/j.jvca.2014.08.021.

YOUNG P Y, KHADAROO R G. Surgical site infections[J]. Surg Clin N Am, 2014, 94(6): 1245-1264. DOI: 10.1016/j.suc.2014.08.008.

BRATZLER D W, HOUCK P M. Antimicrobial prophylaxis for surgery: an advisory statement from the National Surgical Infection Prevention Project[J]. Am J Surg, 2005, 189(4): 395-404. DOI: 10.1016/j.amjsurg.2005.01.015.

BENKE K, BENKE G. Artificial intelligence and big data in public health[J]. Int J Environ Res Public Health, 2018, 15(12): E2796. DOI: 10.3390/ijerph15122796.

FERRARESE A, SARTORI G, ORRÙ G, et al . Machine learning in liver transplantation: a tool for some unsolved questions?[J ] . Transpl Int, 2021, 34(3): 398-411. DOI: 10.1111/tri.13818.

WU C C, YEH W C, HSU W D, et al . Prediction of fatty liver disease using machine learning algorithms[J ] . Comput Methods Programs Biomed, 2019, 170: 23-29. DOI: 10.1016/j.cmpb.2018.12.032.

YANG M, PENG B, ZHUANG Q, et al . Models to predict the short-term survival of acute-on-chronic liver failure patients following liver transplantation[J ] . BMC Gastroenterol, 2022, 22(1): 80. DOI: 10.1186/s12876-022-02164-6.

CHEN T Q, GUESTRIN C. XGBoost: a scalable tree boosting system, proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2016[C]. California: ACM, 2016: 785-794. DOI: 10.1145/2939672.2939785.

VICKERS A J, HOLLAND F. Decision curve analysis to evaluate the clinical benefit of prediction models[J]. Spine J, 2021, 21(10): 1643-1648. DOI: 10.1016/j.spinee.2021.02.024.

Views

654

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Biomarkers for hemorrhagic shock at high-altitude based on metabolomics and machine learning: Identification and validation of D-inositol-4-phosphate, phosalone and methionine

Risk factors for postoperative respiratory failure in patients undergoing cardiovascular surgery and construction of a prediction model

Development of a postoperative recurrence prediction model for stage Ⅰ non-small cell lung cancer patients using multimodal data based on machine learning

Construction of machine learning-based prognostic model of centrosome amplification-related genes for esophageal squamous cell carcinoma

Related Author

XU Yan

HOU Wenqing

TANG Yuxin

QIN Zhongyi

WANG Tao

XIANG Junyu

WANG Bin

WEN Liangzhi

Related Institution

Department of Gastroenterology， Chongqing Key Laboratory of Precision Prevention and Treatment for Digestive System Tumors， Chongqing， Daping Hospital， Army Medical University（Third Military Medical University）

National Key Laboratory of Trauma and Chemical Poisoning， Department of War Trauma， Shock and Blood Transfusion， Daping Hospital， Army Medical University（Third Military Medical University）

Department of Anesthesiology， First Affiliated Hospital， Army Medical University （Third Military Medical University）

Jinfeng Laboratory

Yu-Yue Pathology Scientific Research Center

⁰