MANDARIN: Mixture-of-Experts Framework for Dynamic Delirium and Coma Prediction in ICU Patients: Development and Validation of an Acute Brain Dysfunction Prediction Model

2025-03-08

AI Summary

To address delayed and inconsistent early warning of acute brain dysfunction (ABD)—including delirium and coma—in intensive care units (ICUs), caused by reliance on conventional intermittent clinical assessments (e.g., GCS, CAM, RASS), this study proposes the first mixture-of-experts (MoE) neural network model specifically designed for ABD’s phenotypic heterogeneity. The model integrates multicenter longitudinal physiological time-series data (from eICU, MIMIC-IV, and UFH databases) with static clinical features, enabling multitask, multi-branch modeling of dynamic brain states. It undergoes rigorous cross-institutional validation: external validation yields AUROCs of 75.5% for delirium and 87.3% for coma; prospective validation achieves 82.0% and 93.4%, respectively—both significantly outperforming standard clinical scales (p < 0.001). With 1.5 million parameters, the model supports real-time risk prediction 12–72 hours prior to clinical onset, offering a scalable, data-driven alternative for ABD surveillance in critical care.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Humans and AI: Brain-Sensing and Analysis

Application Category

• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Acute brain dysfunction (ABD) is a common, severe ICU complication, presenting as delirium or coma and leading to prolonged stays, increased mortality, and cognitive decline. Traditional screening tools like the Glasgow Coma Scale (GCS), Confusion Assessment Method (CAM), and Richmond Agitation-Sedation Scale (RASS) rely on intermittent assessments, causing delays and inconsistencies. In this study, we propose MANDARIN (Mixture-of-Experts Framework for Dynamic Delirium and Coma Prediction in ICU Patients), a 1.5M-parameter mixture-of-experts neural network to predict ABD in real-time among ICU patients. The model integrates temporal and static data from the ICU to predict the brain status in the next 12 to 72 hours, using a multi-branch approach to account for current brain status. The MANDARIN model was trained on data from 92,734 patients (132,997 ICU admissions) from 2 hospitals between 2008-2019 and validated externally on data from 11,719 patients (14,519 ICU admissions) from 15 hospitals and prospectively on data from 304 patients (503 ICU admissions) from one hospital in 2021-2024. Three datasets were used: the University of Florida Health (UFH) dataset, the electronic ICU Collaborative Research Database (eICU), and the Medical Information Mart for Intensive Care (MIMIC)-IV dataset. MANDARIN significantly outperforms the baseline neurological assessment scores (GCS, CAM, and RASS) for delirium prediction in both external (AUROC 75.5% CI: 74.2%-76.8% vs 68.3% CI: 66.9%-69.5%) and prospective (AUROC 82.0% CI: 74.8%-89.2% vs 72.7% CI: 65.5%-81.0%) cohorts, as well as for coma prediction (external AUROC 87.3% CI: 85.9%-89.0% vs 72.8% CI: 70.6%-74.9%, and prospective AUROC 93.4% CI: 88.5%-97.9% vs 67.7% CI: 57.7%-76.8%) with a 12-hour lead time. This tool has the potential to assist clinicians in decision-making by continuously monitoring the brain status of patients in the ICU.

Problem

• Predict acute brain dysfunction in ICU patients
• Overcome delays in traditional screening tools
• Provide real-time brain status monitoring

Innovation

• Mixture-of-experts neural network for real-time prediction
• Integrates temporal and static ICU data for brain status
• Outperforms traditional neurological assessment scores significantly