Development and Comparative Analysis of Machine Learning Models for Hypoxemia Severity Triage in CBRNE Emergency Scenarios Using Physiological and Demographic Data from Medical-Grade Devices

TL;DR

This study develops and compares machine learning models, particularly gradient boosting and sequential models, for predicting hypoxemia severity during CBRNE emergencies using physiological data, emphasizing real-time applicability and interpretability.

Contribution

The paper introduces a novel NEWS2+ feature set and demonstrates that gradient boosting models outperform sequential models in speed, interpretability, and reliability for hypoxemia prediction in emergency scenarios.

Findings

Gradient Boosting Models outperformed sequential models in accuracy and speed.

NEWS2+ features significantly improved prediction performance.

Gradient Boosting Models captured key patterns without relying heavily on temporal dependencies.

Abstract

This paper presents the development of machine learning (ML) models to predict hypoxemia severity during emergency triage, especially in Chemical, Biological, Radiological, Nuclear, and Explosive (CBRNE) events, using physiological data from medical-grade sensors. Gradient Boosting Models (XGBoost, LightGBM, CatBoost) and sequential models (LSTM, GRU) were trained on physiological and demographic data from the MIMIC-III and IV datasets. A robust preprocessing pipeline addressed missing data, class imbalances, and incorporated synthetic data flagged with masks. Gradient Boosting Models (GBMs) outperformed sequential models in terms of training speed, interpretability, and reliability, making them well-suited for real-time decision-making. While their performance was comparable to that of sequential models, the GBMs used score features from six physiological variables derived from the enhanced National Early Warning Score (NEWS) 2, which we termed NEWS2+. This approach significantly improved prediction accuracy. While sequential models handled temporal data well, their performance gains did not justify the higher computational cost. A 5-minute prediction window was chosen for timely intervention, with minute-level interpolations standardizing the data. Feature importance analysis highlighted the significant role of mask and score features in enhancing both transparency and performance. Temporal dependencies proved to be less critical, as Gradient Boosting Models were able to capture key patterns effectively without relying on them. This study highlights ML's potential to improve triage and reduce alarm fatigue. Future work will integrate data from multiple hospitals to enhance model generalizability across clinical settings.