Interpretable machine learning model for early prediction of acute kidney injury in patients with rhabdomyolysis

Abstract. Background. Rhabdomyolysis (RM) is a complex set of clinical syndromes. RM-induced acute kidney injury (AKI) is a common illness in war and military operations. This study aimed to develop an interpretable and generalizable model for early AKI prediction in patients with RM. Methods. Retro...

Full description

Saved in:
Bibliographic Details
Main Authors: Ximu Zhang, Xiuting Liang, Zhangning Fu, Yibo Zhou, Yao Fang, MD, Xiaoli Liu, BS, Qian Yuan, Rui Liu, Quan Hong, Chao Liu
Format: Article
Language:English
Published: Wolters Kluwer Health/LWW 2024-12-01
Series:Emergency and Critical Care Medicine
Online Access:http://journals.lww.com/10.1097/EC9.0000000000000126
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850151659686068224
author Ximu Zhang
Xiuting Liang
Zhangning Fu
Yibo Zhou
Yao Fang, MD
Xiaoli Liu, BS
Qian Yuan
Rui Liu
Quan Hong
Chao Liu
author_facet Ximu Zhang
Xiuting Liang
Zhangning Fu
Yibo Zhou
Yao Fang, MD
Xiaoli Liu, BS
Qian Yuan
Rui Liu
Quan Hong
Chao Liu
author_sort Ximu Zhang
collection DOAJ
description Abstract. Background. Rhabdomyolysis (RM) is a complex set of clinical syndromes. RM-induced acute kidney injury (AKI) is a common illness in war and military operations. This study aimed to develop an interpretable and generalizable model for early AKI prediction in patients with RM. Methods. Retrospective analyses were performed on 2 electronic medical record databases: the eICU Collaborative Research Database and the Medical Information Mart for Intensive Care III database. Data were extracted from the first 24 hours after patient admission. Data from the two datasets were merged for further analysis. The extreme gradient boosting (XGBoost) model with the Shapley additive explanation method (SHAP) was used to conduct early and interpretable predictions of AKI. Results. The analysis included 938 eligible patients with RM. The XGBoost model exhibited superior performance (area under the receiver operating characteristic curve [AUC] = 0.767) compared to the other models (logistic regression, AUC = 0.711; support vector machine, AUC = 0.693; random forest, AUC = 0.728; and naive Bayesian, AUC = 0.700). Conclusion. Although the XGBoost model performance could be improved from an absolute perspective, it provides better predictive performance than other models for estimating the AKI in patients with RM based on patient characteristics in the first 24 hours after admission to an intensive care unit. Furthermore, including SHAP to elucidate AKI-related factors enables individualized patient treatment, potentially leading to improved prognoses for patients with RM.
format Article
id doaj-art-e2173f3ddfbb489389d8a3516a5e077a
institution OA Journals
issn 2097-0617
2693-860X
language English
publishDate 2024-12-01
publisher Wolters Kluwer Health/LWW
record_format Article
series Emergency and Critical Care Medicine
spelling doaj-art-e2173f3ddfbb489389d8a3516a5e077a2025-08-20T02:26:09ZengWolters Kluwer Health/LWWEmergency and Critical Care Medicine2097-06172693-860X2024-12-014415516210.1097/EC9.0000000000000126202412000-00004Interpretable machine learning model for early prediction of acute kidney injury in patients with rhabdomyolysisXimu Zhang0Xiuting Liang1Zhangning Fu2Yibo Zhou3Yao Fang, MD4Xiaoli Liu, BS5Qian Yuan6Rui Liu7Quan Hong8Chao Liu9a Department of Critical Care Medicine, Hainan Hospital of Chinese PLA General Hospital, Sanya, Hainan, Chinab Department of Nursing, The First Medical Center of Chinese People’s Liberation Army General Hospital, Beijing, Chinac Chinese PLA Institute of Nephrology, State Key Laboratory of Kidney Diseases, National Clinical Research Center for Kidney Diseases, Beijing, Chinad Department of Critical Care Medicine, The First Medical Center of Chinese People’s Liberation Army General Hospital, Beijing, Chinae Department of Respiratory and Critical Care Medicine, General Hospital of Center Theater of PLA, Wuhan, Hubei, Chinaf Center for Artificial Intelligence in Medicine, The Chinese PLA General Hospital, Beijing, Chinah Honor Device Co., Ltd., Beijing, Chinai Department of Critical Care Medicine, Tangdu Hospital, Air Force Military Medical University, Xi’an, Shaanxi, China.c Chinese PLA Institute of Nephrology, State Key Laboratory of Kidney Diseases, National Clinical Research Center for Kidney Diseases, Beijing, Chinad Department of Critical Care Medicine, The First Medical Center of Chinese People’s Liberation Army General Hospital, Beijing, ChinaAbstract. Background. Rhabdomyolysis (RM) is a complex set of clinical syndromes. RM-induced acute kidney injury (AKI) is a common illness in war and military operations. This study aimed to develop an interpretable and generalizable model for early AKI prediction in patients with RM. Methods. Retrospective analyses were performed on 2 electronic medical record databases: the eICU Collaborative Research Database and the Medical Information Mart for Intensive Care III database. Data were extracted from the first 24 hours after patient admission. Data from the two datasets were merged for further analysis. The extreme gradient boosting (XGBoost) model with the Shapley additive explanation method (SHAP) was used to conduct early and interpretable predictions of AKI. Results. The analysis included 938 eligible patients with RM. The XGBoost model exhibited superior performance (area under the receiver operating characteristic curve [AUC] = 0.767) compared to the other models (logistic regression, AUC = 0.711; support vector machine, AUC = 0.693; random forest, AUC = 0.728; and naive Bayesian, AUC = 0.700). Conclusion. Although the XGBoost model performance could be improved from an absolute perspective, it provides better predictive performance than other models for estimating the AKI in patients with RM based on patient characteristics in the first 24 hours after admission to an intensive care unit. Furthermore, including SHAP to elucidate AKI-related factors enables individualized patient treatment, potentially leading to improved prognoses for patients with RM.http://journals.lww.com/10.1097/EC9.0000000000000126
spellingShingle Ximu Zhang
Xiuting Liang
Zhangning Fu
Yibo Zhou
Yao Fang, MD
Xiaoli Liu, BS
Qian Yuan
Rui Liu
Quan Hong
Chao Liu
Interpretable machine learning model for early prediction of acute kidney injury in patients with rhabdomyolysis
Emergency and Critical Care Medicine
title Interpretable machine learning model for early prediction of acute kidney injury in patients with rhabdomyolysis
title_full Interpretable machine learning model for early prediction of acute kidney injury in patients with rhabdomyolysis
title_fullStr Interpretable machine learning model for early prediction of acute kidney injury in patients with rhabdomyolysis
title_full_unstemmed Interpretable machine learning model for early prediction of acute kidney injury in patients with rhabdomyolysis
title_short Interpretable machine learning model for early prediction of acute kidney injury in patients with rhabdomyolysis
title_sort interpretable machine learning model for early prediction of acute kidney injury in patients with rhabdomyolysis
url http://journals.lww.com/10.1097/EC9.0000000000000126
work_keys_str_mv AT ximuzhang interpretablemachinelearningmodelforearlypredictionofacutekidneyinjuryinpatientswithrhabdomyolysis
AT xiutingliang interpretablemachinelearningmodelforearlypredictionofacutekidneyinjuryinpatientswithrhabdomyolysis
AT zhangningfu interpretablemachinelearningmodelforearlypredictionofacutekidneyinjuryinpatientswithrhabdomyolysis
AT yibozhou interpretablemachinelearningmodelforearlypredictionofacutekidneyinjuryinpatientswithrhabdomyolysis
AT yaofangmd interpretablemachinelearningmodelforearlypredictionofacutekidneyinjuryinpatientswithrhabdomyolysis
AT xiaoliliubs interpretablemachinelearningmodelforearlypredictionofacutekidneyinjuryinpatientswithrhabdomyolysis
AT qianyuan interpretablemachinelearningmodelforearlypredictionofacutekidneyinjuryinpatientswithrhabdomyolysis
AT ruiliu interpretablemachinelearningmodelforearlypredictionofacutekidneyinjuryinpatientswithrhabdomyolysis
AT quanhong interpretablemachinelearningmodelforearlypredictionofacutekidneyinjuryinpatientswithrhabdomyolysis
AT chaoliu interpretablemachinelearningmodelforearlypredictionofacutekidneyinjuryinpatientswithrhabdomyolysis