Related Experiment Video
Updated: Jun 3, 2025
Noninvasive and Invasive Renal Hypoxia Monitoring in a Porcine Model of Hemorrhagic Shock
Published on: October 28, 2022
Prediction of Acute Kidney Injury for Critically Ill Cardiogenic Shock Patients with Machine Learning Algorithms
Xiaofei Zhang1, Yonghong Xiong2, Huilan Liu3
1Department of Gerontology, China Aerospace Science & Industry Corporation 731 hospital, Beijing, People's Republic of China.
Insights
Machine learning models effectively predict acute kidney injury (AKI) in critically ill patients with cardiogenic shock (CS). Ensemble models demonstrated superior performance over traditional logistic regression, offering a promising tool for early detection and management.
Area of Science:
- Critical Care Medicine
- Nephrology
- Data Science
Background:
- Critically ill patients with cardiogenic shock (CS) are at high risk of developing acute kidney injury (AKI).
- Accurate prediction of AKI in this population is crucial for timely intervention and improved outcomes.
- Existing prediction models may not fully capture the complexity of AKI development in CS patients.
Purpose of the Study:
- To develop and validate machine learning (ML) models for predicting AKI in critically ill CS patients.
- To compare the predictive performance of various ML algorithms against conventional logistic regression.
- To identify key features influencing AKI prediction using SHAP analysis.
Main Methods:
- Retrospective analysis of clinical data from MIMIC-IV, eICU, and a university hospital database.
- Application of five ML algorithms (LightGBM, decision tree, XGBoost, random forest, ensemble model) and logistic regression.
- Validation of models using independent datasets and assessment via ROC curves and AUC values.
Main Results:
- The ensemble model achieved the highest predictive accuracy (AUC: 0.91-0.92 across datasets).
- Random forest and XGBoost also showed strong performance (AUC: 0.89-0.90).
- Logistic regression demonstrated the lowest predictive performance (AUC: 0.61-0.62).
Conclusions:
- ML algorithms, particularly ensemble models, significantly outperform logistic regression in predicting AKI in critically ill CS patients.
- These ML models offer a valuable tool for early AKI risk stratification in this vulnerable population.
- Further research can refine these models for clinical implementation.
Background:
The aim of this study was to use five machine learning approaches and logistic regression to design and validate the acute kidney injury (AKI) prediction model for critically ill individuals with cardiogenic shock (CS).
Methods:
All patients who diagnosed with CS from the MIMIC-IV database, the eICU database, and Zhongnan hospital of Wuhan university were included in this study. Clinical information, including demographics, comorbidities, vital signs, critical illness scores and laboratory tests was retrospectively collected. Five machine learning algorithms (LightGBM, decision tree, XGBoost, random forest, and ensemble model) and one conventional logistic regression were applied for the prediction of AKI in critically ill individuals with CS. ROC curves were generated via python software to assess the overall performance of machine learning algorithms and the SHAP analysis was adopted to reveal the impact of prediction for each feature.
Results:
The ensemble model exhibited the best predictive ability (AUC:0.91, 95% CI, 0.88-0.94), followed by random forest (AUC:0.90, 95% CI, 0.86-0.94) and XGBoost (AUC:0.89, 95% CI, 0.84-0.92). While the logistic regression model obtained the worst predictive performance (AUC:0.62, 95% CI, 0.56-0.68). When validated the prediction models with eICU database, the ensemble model exhibited the best predictive ability (AUC:0.92, 95% CI, 0.89-0.96), while the logistic model obtained the worst predictive performance (AUC:0.61, 95% CI, 0.56-0.67). Finally, we verified the prediction models using the data from our hospital and ensemble model still exhibited the best predictive ability (AUC:0.74, 95% CI, 0.62-0.86), while the decision tree model obtained the worst predictive performance (AUC:0.52, 95% CI 0.35-0.70).
Conclusion:
Machine learning algorithms could be utilized for the AKI prediction among critically ill CS patients, and exhibit superior predictive performance compared to the conventional logistic regression analysis.
