Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Videos

Machine Learning Approaches for Mortality Prediction in ARDS.

Xingyue Huo¹, Christian Bime¹, Joseph Finkelstein¹

¹University of Arizona, Tucson, AZ, USA.

Studies in Health Technology and Informatics

|July 3, 2026

Summary

This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

A Multi-Modular Human-AI Workflow for LLM-Assisted Thematic Analysis: Application to COPD Telerehabilitation Interviews.

Studies in health technology and informatics·2026

Same author

Association of Remote Patient Monitoring with Care Utilization in Patients with Chronic Cardiopulmonary Conditions.

Studies in health technology and informatics·2026

Same author

Estimating Mosaic Loss of the Y Chromosome in Male Bladder Cancer Participants Using "All of Us" Data.

Studies in health technology and informatics·2026

Same author

Early Prediction of Delirium in ICU Patients Using Machine Learning Analysis of Admission Data.

Studies in health technology and informatics·2026

Same author

Multimodule Human-Artificial Intelligence Collaboration Pipeline for Large Language Model-Assisted Thematic Analysis Across Digital Health Interview Studies: Comparative Evaluation Study.

JMIR medical informatics·2026

Same author

Early Safety/Therapeutic Results of ALT-100 eNAMPT mAb Impact in the PUERTA Phase 2A ARDS Trial.

American journal of respiratory and critical care medicine·2026

Same journal

A GenAI Pipeline for Violinist Kinematic Data Management.

Studies in health technology and informatics·2026

Same journal

AMAL-For-Qatar: A Comprehensive AI Ecosystem for Fetal Ultrasound Analysis - Project Overview and Achievements.

Studies in health technology and informatics·2026

Same journal

Longitudinal Treatment-Aware Multimodal AI for Dermatology: A Scoping Review.

Studies in health technology and informatics·2026

Same journal

Predicting Postpartum Depression Using Imbalance-Aware Machine Learning.

Studies in health technology and informatics·2026

Same journal

Validation of Deep-Learning Models for Autosegmentation of Brain Metastases.

Studies in health technology and informatics·2026

Same journal

Delay-Dependent Gating in Modular RNNs.

Studies in health technology and informatics·2026

See all related articles

Accurate mortality prediction in acute respiratory distress syndrome (ARDS) aids patient care. XGBoost models demonstrated the best balance of performance metrics for predicting ARDS patient mortality.

Area of Science:

Critical Care Medicine
Medical Informatics
Machine Learning in Healthcare

Background:

Acute respiratory distress syndrome (ARDS) presents a significant challenge in critical care.
Predicting mortality in ARDS is crucial for timely clinical decision-making and resource allocation.
Existing prediction models may not fully capture the complexity of ARDS patient outcomes.

Purpose of the Study:

To compare the performance of various machine learning models in predicting mortality for patients with ARDS.
To identify the most clinically appropriate model for predicting both 90-day all-cause mortality and 28-day hospital mortality in ARDS.
To evaluate model discrimination (AUC) and other key metrics like recall and F1 score.

Main Methods:

Utilized multiple machine learning algorithms to analyze ARDS patient data.

Keywords:

Acute Respiratory Distress Syndrome Machine Learning Mortality

Related Experiment Videos

Trained and validated models to predict two distinct mortality outcomes: 90-day all-cause and 28-day hospital mortality.

Assessed model performance using Area Under the Curve (AUC), recall, and F1 score.

Main Results:

All evaluated machine learning models demonstrated good discrimination for predicting ARDS mortality (AUCs ranging from 0.693 to 0.753).
XGBoost exhibited a superior and consistent balance of AUC, recall, and F1 score across both mortality prediction tasks.
Random forest models achieved high AUC for the primary outcome but showed significantly lower recall, potentially missing high-risk patients.

Conclusions:

XGBoost emerges as the most clinically suitable machine learning model for predicting mortality in ARDS patients due to its balanced performance.
The findings highlight the potential of machine learning, particularly XGBoost, to improve risk stratification and clinical management in ARDS.
Careful consideration of multiple performance metrics, beyond AUC, is essential for selecting clinically relevant predictive models in critical care settings.