Artificial intelligence in emergency medicine: a narrative review

Document Type

Article

Publication Date

1-20-2026

Institution/Department

Emergency Medicine

Journal Title

The American journal of emergency medicine

Abstract

INTRODUCTION: Artificial intelligence (AI) is increasingly embedded in healthcare, with expanding applications in emergency medicine (EM). OBJECTIVE: This focused narrative review provides an overview of key applications, limitations, and future directions of AI that are most relevant for emergency clinicians. DISCUSSION: AI refers broadly to computational systems capable of performing tasks that typically require human intelligence, including pattern recognition, prediction, and language understanding. Within EM, AI is being deployed across the care continuum. Prehospital uses include triage, dispatch, patient assessment, protocol adherence, and decision support. There are several uses of AI in radiology, including diagnosis as well as improving efficiency, safety, and education. AI can be used to assist with point-of-care ultrasound, including imaging indication, acquisition, interpretation, and medical decision-making. AI may be used to improve emergency department patient registration, redirection, triage, waiting time, and data entry, as well as diagnostic support. Use of AI in pediatric EM may improve triage, clinical decision-making and diagnostic support, documentation, prognostication, and managing patient disposition and bed availability. AI implications in trauma include improved triage, decision-making, management, and prognostication. Prognosis and predictive analysis uses include prognostic modeling, operational, and system-level prediction, optimizing triage, and data integrity and synthesis. AI shows promise for enhancing clinical education, particularly in helping to individualize the learning process. Potential limitations include data quality and bias, reporting standards, interpretability and transparency, external validation and generalizability, clinical workflow integration, and human factors. CONCLUSIONS: AI demonstrates promise across multiple domains for enhancing the practice of EM. Clinicians should understand its potential uses and limitations.

First Page

155

Last Page

165

Link to Full Text

Share

COinS