A machine learning-based prediction of hospital mortality in mechanically ventilated ICU patients
Hexin Li, Negin Ashrafi, Chris Kang, Guanlan Zhao, Yubing Chen, Maryam Pishgar
Abstract
Mechanical ventilation (MV) is vital for critically ill ICU patients but carries significant mortality risks. This study aims to develop a predictive model to estimate hospital mortality among MV patients, utilizing comprehensive health data to assist ICU physicians with early-stage alerts.
Introduction
In the United States, over one million patients receive mechanical ventilation (MV) annually in Intensive Care Units (ICU), occupying 24–41% of ICU beds at any given time [1]. Although MV is frequently considered a lifesaving intervention.
Method
The feature selection process in this study involved several stages. Initially, backward elimination and the Lasso method were employed to identify the most significant features [20, 21].
Results
Following our approach for feature selection and data preprocessing for ICU patients requiring mechanical ventilation, our final dataset included 25,202 patients from the MIMIC-III database.
Discussion
Several models have been concurrently developed to predict mortality for ICU patients, specifically those receiving mechanical ventilation (MV). These include machine learning frameworks using various techniques and databases.
Conclusion
This study significantly improves the prediction of hospital mortality for ICU patients on mechanical ventilation using advanced machine learning techniques.
Citation: Li H, Ashrafi N, Kang C, Zhao G, Chen Y, Pishgar M (2024) A machine learning-based prediction of hospital mortality in mechanically ventilated ICU patients. PLoS ONE 19(9): e0309383. https://doi.org/10.1371/journal.pone.0309383
Editor: Upaka Rathnayake, Atlantic Technological University, IRELAND
Received: July 12, 2024; Accepted: August 10, 2024; Published: September 4, 2024
Copyright: © 2024 Li et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The raw dataset is available in the MIMIC-III repository:https://physionet.org/content/mimiciii/1.4/.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Source: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0309383#abstract0
