Interpretable Machine Learning Model for Early Prediction of Acute Kidney Injury in Critically Ill Patients with Cirrhosis: A Retrospective Study

2025-08-13

AI Summary

This study addresses the low predictive accuracy, poor interpretability, and limited clinical integration of existing models for early acute kidney injury (AKI) prediction in critically ill patients with cirrhosis. Leveraging the MIMIC-IV database, we developed a highly interpretable and deployable machine learning model using routinely collected clinical variables from the first 48 hours after ICU admission. Feature selection was performed via LASSO regression, and class imbalance was mitigated using SMOTE. Six algorithms were systematically evaluated; LightGBM achieved optimal performance (AUROC = 0.808, accuracy = 0.704, negative predictive value = 0.911). Key predictors—including pH and coagulation time—possess well-established pathophysiological relevance, directly supporting clinical decision-making. The model’s high negative predictive value enables safe de-escalation of monitoring for low-risk patients, thereby enhancing clinician trust and facilitating precision prevention of AKI in this vulnerable population.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Humans and AI: Human-in-the-loop Machine Learning
• Data Mining & Knowledge Management: Knowledge Acquisition from the Web

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Background: Cirrhosis is a progressive liver disease with high mortality and frequent complications, notably acute kidney injury (AKI), which occurs in up to 50% of hospitalized patients and worsens outcomes. AKI stems from complex hemodynamic, inflammatory, and metabolic changes, making early detection essential. Many predictive tools lack accuracy, interpretability, and alignment with intensive care unit (ICU) workflows. This study developed an interpretable machine learning model for early AKI prediction in critically ill patients with cirrhosis. Methods: We conducted a retrospective analysis of the MIMIC-IV v2.2 database, identifying 1240 adult ICU patients with cirrhosis and excluding those with ICU stays under 48 hours or missing key data. Laboratory and physiological variables from the first 48 hours were extracted. The pipeline included preprocessing, missingness filtering, LASSO feature selection, and SMOTE class balancing. Six algorithms-LightGBM, CatBoost, XGBoost, logistic regression, naive Bayes, and neural networks-were trained and evaluated using AUROC, accuracy, F1-score, sensitivity, specificity, and predictive values. Results: LightGBM achieved the best performance (AUROC 0.808, 95% CI 0.741-0.856; accuracy 0.704; NPV 0.911). Key predictors included prolonged partial thromboplastin time, absence of outside-facility 20G placement, low pH, and altered pO2, consistent with known cirrhosis-AKI mechanisms and suggesting actionable targets. Conclusion: The LightGBM-based model enables accurate early AKI risk stratification in ICU patients with cirrhosis using routine clinical variables. Its high negative predictive value supports safe de-escalation for low-risk patients, and interpretability fosters clinician trust and targeted prevention. External validation and integration into electronic health record systems are warranted.

Problem

• Early prediction of acute kidney injury in cirrhosis patients
• Interpretable machine learning model for ICU workflows
• Addressing lack of accurate and actionable predictive tools

Innovation

• Interpretable LightGBM model for AKI prediction
• LASSO and SMOTE for feature selection
• Routine clinical variables for risk stratification