Related Experiment Videos
Artificial Intelligence in Predicting Mortality Risk for Critically Ill Children in Pediatric Intensive Care Units: A
Seyedeh Narjes Ahmadizadeh1, Hasan Shamsi2, Neda Izadi3
1Department of Pediatric Intensive Care, Mofid Children's Hospital Shahid Beheshti University of Medical Sciences Tehran Iran.
Health Science Reports
|May 6, 2026
Summary
Artificial Intelligence (AI) significantly improves mortality prediction in Pediatric Intensive Care Units (PICUs). Machine learning models, particularly random forest, show high accuracy (AUROC > 0.8 in 88% of studies) for predicting outcomes in critically ill children.
Area of Science:
- Pediatric Critical Care Medicine
- Medical Informatics
- Artificial Intelligence in Healthcare
Background:
- Accurate mortality risk prediction in Pediatric Intensive Care Units (PICUs) is a significant clinical challenge.
- Artificial Intelligence (AI) offers potential for identifying complex patterns in clinical data to improve risk assessment.
- A systematic review is needed to synthesize evidence on AI's role in PICU mortality prediction.
Purpose of the Study:
- To systematically review and summarize the current evidence on the application of AI in predicting mortality among critically ill children in PICUs.
- To evaluate the performance and methodologies of AI-driven mortality prediction models.
Main Methods:
- A systematic literature search was performed on PubMed, Scopus, and Web of Science on December 2, 2024.
- Study selection involved title/abstract screening followed by full-text review, adhering to PRISMA guidelines.
- Risk of bias and applicability were assessed using the Prediction model study Risk Of Bias Assessment Tool (PROBAST).
Main Results:
- Seventeen articles were included, with 76% published in 2020 or later.
- Random forest was the most frequently used algorithm.
- The Area Under the Receiver Operating Characteristic (AUROC) exceeded 0.8 in 88% of the reviewed studies.
Conclusions:
- Machine learning and deep learning models demonstrate significant potential for enhancing mortality prediction accuracy in PICUs.
- Algorithm choice and feature selection critically influence prediction performance, as indicated by variable AUROC values.
- AI-powered tools can aid clinicians in better risk stratification and resource allocation for critically ill children.