Article Dans Une Revue Annales françaises de médecine d'urgence Année : 2025

Artificial intelligence in emergency medicine by the Innovation Board of the French Society of Emergency Medicine

L’intelligence artificielle en médecine d’urgence par le Board Innovation de la Société française de médecine d’urgence

Résumé

Artificial intelligence (AI) is revolutionizing emergency medicine by optimizing diagnosis, triage, and therapeutic decisions. It enables rapid analysis of large volumes of data to assist healthcare professionals without replacing them. Historically, AI advanced significantly with deep learning in the 2010s, outperforming traditional methods through deep neural networks. Supervised, unsupervised, and hybrid models process structured and unstructured data to predict key variables such as hospitalization or mortality. In pre-hospital care, AI improves emergency call assessment, optimizes resource allocation, and detects critical situations. In hospitals, it assists with admission triage, diagnostics (fractures, heart attacks), and patient orientation, improving patient flow management. AI tools demonstrate increased sensitivity and specificity in interpreting ultrasounds and X-rays. However, integrating AI presents significant technical, ethical, security, and educational challenges. Ensuring system interoperability, data protection, and scientific validation of models is essential. Additionally, algorithmic biases and risks of errors remain concerns, requiring human oversight. AI offers promising opportunities to strengthen emergency medicine, but its deployment demands careful consideration of ethics, professional training, and liability in case of errors. Maintaining human predominance over technology is crucial to ensure optimal and safe patient care. While AI enhances efficiency and decision-making, healthcare professionals must remain central to the process, ensuring a balance between technological advancements and human judgment for better patient outcomes.

Fichier non déposé

Dates et versions

hal-05242651 , version 1 (05-09-2025)

Identifiants

Citer

E. Lansiaux, E. Arnaud, L. Arrouy, P.-H. Auboiroux, P.-A. Balaz, et al.. L’intelligence artificielle en médecine d’urgence par le Board Innovation de la Société française de médecine d’urgence. Annales françaises de médecine d'urgence, 2025, 15 (4), pp.259-267. ⟨10.1684/afmu.2025.0637⟩. ⟨hal-05242651⟩
218 Consultations
0 Téléchargements

Altmetric

Partager

  • More