Artificial Intelligence Applications in Anesthesiology and Intensive Care Medicine: Current Evidence, Clinical Readiness and Future Perspectives
The Anatolian Journal of Anesthesiology and Intensive Care · 2026
DOI: 10.71350/ajaic.20 ↗Kurz gefasst
Fachartikel in der Zeitschrift The Anatolian Journal of Anesthesiology and Intensive Care, 2026. Verfasst von Gürcan Güler. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Themen: Künstliche Intelligenz in der Pflege, Intensivpflege
Abstract (Originalsprache)
Background
Artificial intelligence (AI) is increasingly being investigated across anesthesiology and intensive care medicine (ICM), where complex data, continuous monitoring, and time-sensitive decisions create substantial opportunities for computational decision support. However, the clinical readiness of many AI applications remains uncertain.
Methods
This narrative review provides a structured and clinically oriented synthesis of AI applications in anesthesiology and ICM. PubMed and Google Scholar were searched for English-language publications from January 2019 to December 2025 using the terms “artificial intelligence,” “machine learning,” “deep learning,” “large language models,” “ChatGPT,” “anesthesiology,” “anesthesia,” “perioperative,” “intensive care,” and “critical care.” Studies with direct or potential clinical relevance to perioperative or intensive care practice were considered, and the selected literature was synthesized descriptively.
Results
Forty-seven publications were included, encompassing systematic reviews and meta-analyses, narrative and scoping reviews, retrospective and prospective studies, randomized trials, and model development, validation, and evaluation studies. In anesthesiology, AI applications included preoperative risk stratification, difficult airway assessment, hemodynamic monitoring, closed-loop anesthesia, ultrasound-assisted procedures, postoperative outcome prediction, and perioperative communication. In ICM, major applications involved mortality and sepsis risk assessment, respiratory failure and mechanical ventilation management, intelligent alarm systems, delirium detection, sedation monitoring, and large language model (LLM)-assisted documentation. Although several approaches demonstrated promising predictive or supportive capabilities, prospective and randomized evidence remained limited. External validation, real-world clinical evaluation, data quality, model transparency, privacy, accountability, and regulatory uncertainty were recurring challenges.
Conclusion
AI technologies have substantial potential to support clinical decision-making, situational awareness, individualized care, and workflow efficiency in anesthesiology and ICM. However, most applications remain in early stages of clinical translation. Current evidence supports their use as complementary, clinician-supervised tools rather than autonomous decision-makers. Robust prospective validation, standardized evaluation, ethical governance, and alignment with clinical needs are essential for safe and meaningful integration into routine practice.
Zitieren
APA
Güler, G. (2026). Artificial Intelligence Applications in Anesthesiology and Intensive Care Medicine: Current Evidence, Clinical Readiness and Future Perspectives. The Anatolian Journal of Anesthesiology and Intensive Care. https://doi.org/10.71350/ajaic.20
Vancouver
Güler G. Artificial Intelligence Applications in Anesthesiology and Intensive Care Medicine: Current Evidence, Clinical Readiness and Future Perspectives. The Anatolian Journal of Anesthesiology and Intensive Care. 2026. doi:10.71350/ajaic.20
BibTeX
@article{Gler2026,
author = {Güler, Gürcan},
title = {Artificial Intelligence Applications in Anesthesiology and Intensive Care Medicine: Current Evidence, Clinical Readiness and Future Perspectives},
journal = {The Anatolian Journal of Anesthesiology and Intensive Care},
year = {2026},
doi = {10.71350/ajaic.20}
}
EUVIMED