Clinically Interpretable Mortality Prediction for ICU Patients with Diabetes and Atrial Fibrillation: A Machine Learning Approach

2025-06-18

AI Summary

This study addresses the critical need for early, interpretable mortality prediction in ICU patients with comorbid diabetes mellitus and atrial fibrillation (DM+AF). Method: We propose a novel, comorbidity-aware modeling framework integrating two-stage feature selection (ANOVA F-test followed by random forest) and Accumulated Local Effects (ALE) analysis to uncover nonlinear clinical mechanisms—such as age-related risk acceleration and bilirubin threshold effects. The primary model is logistic regression, enhanced with SMOTE oversampling, z-score normalization, and median/mode imputation. Contribution/Results: The model achieves an AUROC of 0.825 (95% CI: 0.779–0.867), significantly outperforming black-box alternatives. Key interpretable predictors identified include plasma renin activity, age, total bilirubin, and endotracheal extubation status. These findings enable rapid clinical triage and support individualized, mechanism-informed decision-making for high-risk DM+AF ICU patients.

Technology Category

• Machine Learning: Transparent, Interpretable, Explainable ML
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Reasoning under Uncertainty: Other Foundations of Reasoning under Uncertainty

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Background: Patients with both diabetes mellitus (DM) and atrial fibrillation (AF) face elevated mortality in intensive care units (ICUs), yet models targeting this high-risk group remain limited. Objective: To develop an interpretable machine learning (ML) model predicting 28-day mortality in ICU patients with concurrent DM and AF using early-phase clinical data. Methods: A retrospective cohort of 1,535 adult ICU patients with DM and AF was extracted from the MIMIC-IV database. Data preprocessing involved median/mode imputation, z-score normalization, and early temporal feature engineering. A two-step feature selection pipeline-univariate filtering (ANOVA F-test) and Random Forest-based multivariate ranking-yielded 19 interpretable features. Seven ML models were trained with stratified 5-fold cross-validation and SMOTE oversampling. Interpretability was assessed via ablation and Accumulated Local Effects (ALE) analysis. Results: Logistic regression achieved the best performance (AUROC: 0.825; 95% CI: 0.779-0.867), surpassing more complex models. Key predictors included RAS, age, bilirubin, and extubation. ALE plots showed intuitive, non-linear effects such as age-related risk acceleration and bilirubin thresholds. Conclusion: This interpretable ML model offers accurate risk prediction and clinical insights for early ICU triage in patients with DM and AF.

Problem

• Predict mortality in ICU patients with diabetes and atrial fibrillation
• Develop interpretable machine learning model for clinical use
• Identify key risk factors from early-phase ICU data

Innovation

• Interpretable machine learning model development
• Two-step feature selection pipeline
• Early temporal feature engineering