Algorithms in action: Artificial Intelligence (AI) at the frontlines of anaesthesia and intensive care 

The world of Artificial Intelligence (AI) is moving fast. It seems like it was only a short time ago that healthcare was dreaming about the possibilities of how AI and Machine Learning (ML) could support clinical practice. Physicians dreamed of being relieved of the administrative burdens of clinical practice and having more time for human connection with patients. They saw opportunities to perfect diagnosis and the delivery of treatment. This was especially true for anaesthesiologists. 

The dream of an anaesthetic machine that could continuously monitor the patient, without fatigue or distraction, monitoring depth of anaesthesia, haemodynamics, neuromuscular function and, in response, titrate drugs appropriately appealed deeply to them. The application of closed-loop devices to anaesthetic agents and depth of anaesthesia represented the pinnacle of anaesthesia research and innovation. (1) An early attempt at semi-autonomous anaesthesia was the Sedasys pharmacological robot from Johnson & Johnson. It was designed for procedural sedation during colonoscopy without an anaesthetist in the room. After years of development, Sedasys was withdrawn from the market following fierce opposition from anaesthesiology professional societies. It showed the potential for automation, but limitations in adaptability and responsiveness led to it being discontinued. (2) 

AI has now evolved sufficiently to be on the verge of entering perioperative care. Its scope is broader than just automation. Anaesthesiology and critical care can benefit directly from AI support. It can facilitate risk assessment in the preoperative stages, interpreting physiological data and allowing for better planning for improved outcomes. (3) In intraoperative care, AI can drive mechanical ventilation, adjusting ventilation directly to the patient’s needs and responding to adjust parameters in real time. (4) Automated sonography and prediction models can also support the clinician in their decision-making by predicting complications, optimising the treatment pathway, and relieving the burdens of time-sensitive care. (5) In post-operative care, AI can recognise deterioration and provide early warnings to initiate treatment. (6) These sound so ideal, but do these aspirations safely measure up to clinical realities?  

In this symposium session “Algorithms in action: Artificial Intelligence (AI) at the frontlines of anaesthesia and intensive care” the speakers will explore how AI is enhancing pre-anaesthesia evaluations by improving risk assessment and perioperative planning, recent advancements in AI-driven mechanical ventilation, including its impact on respiratory mechanics, patient-tailored ventilation, and real-time adjustment of ventilation parameters, and finally, integration of AI in automated sonography and prediction models, and their implications on patient care.  

Dr. Rachele Simonte is an anaesthesiologist from Santa Maria della Misericordia Hospital in Perugia, Italy. Her research focuses on respiratory mechanics, patient-ventilator interactions, and non-invasive and invasive mechanical ventilation support. She has co-authored studies on advanced respiratory monitoring during ECMO, point-of-care bedside monitoring in acute respiratory failure, intraoperative lung protection strategies, comfort during non-invasive ventilation, and PEEP-induced alveolar recruitment. (7)(8)(9) In her presentation “Pimp my ventilator: AI driven mechanical ventilation” she will explore how AI/ML can enhance real-time ventilator management, and how AI can support patient-specific tailoring of settings based on respiratory mechanics, closed-loop adjustments, or integration with monitoring data for safer assisted ventilation.  

Dr. Regina Pikman Gavriely is an anaesthesiologist from Tel Aviv Medical Centre, Israel. She has published on airway management during procedural sedation/bronchoscopy and trauma haemorrhage control techniques (often incorporating ultrasound guidance). (10)(11) Her work includes randomised comparisons of respiratory support devices and case reports on manual compression techniques for vascular injury. In her presentation “Automated sonography applications” she will explore how automated or AI-assisted sonography can offer real-time guidance for vascular access, perioperative respiratory assessment, or POCUS tools in anaesthesia/ICU that depend less on the operator’s skill and experience.  

Dr. Carlos Ferrando Ortola is head of the Surgical Intensive Care Unit, Anesthesiology and Critical Care Department, Hospital Clínic of Barcelona in Spain. He will focus on how AI and ML can be used in “Deterioration prediction in the postoperative period”. He has extensive research experience in perioperative respiratory medicine, including large trials on lung-protective ventilation and extensive involvement in ML-based predictive modelling for outcomes in acute hypoxemic respiratory failure (AHRF)/ARDS, mechanical ventilation duration, and ICU mortality. (12)(13) Dr. Ortola will demonstrate evidence to show that ML models can be extremely useful in predicting early postoperative deterioration using preoperative/intraoperative data and including postoperative vital signs and laboratory results to forecast the chances of respiratory failure, re-intubation, or the likely need to escalate patient care to the ICU. He will share data from studies that used ML to predict prolonged mechanical ventilation, ICU mortality, and outcomes in AHRF/ARDS patients, all of which are highly relevant to predicting postoperative deterioration. (14)(15)(16) 

The symposium session “Algorithms in action: Artificial Intelligence (AI) at the frontlines of anaesthesia and intensive care” will take place at the Euroanaesthesia Congress on Sunday, June 7 at 17:00–18:00 CEST in room DOCK 1. 

References 

1. Wingert T, Lee C, Cannesson M. Machine Learning, Deep Learning, and Closed Loop Devices-Anesthesia Delivery. Anesthesiol Clin. 2021;39(3):565-581. doi:10.1016/j.anclin.2021.03.012 https://www.anesthesiology.theclinics.com/article/S1932-2275(21)00031-8/abstract  

2. Alexander JC, Joshi GP. Anesthesiology, automation, and artificial intelligence. Proc (Bayl Univ Med Cent). 2017;31(1):117-119. Published 2017 Dec 5. doi:10.1080/08998280.2017.1391036 https://www.tandfonline.com/doi/full/10.1080/08998280.2017.1391036  

3. Rajkomar, A., Oren, E., Chen, K. et al. Scalable and accurate deep learning with electronic health records. npj Digital Med 1, 18 (2018). https://doi.org/10.1038/s41746-018-0029-1 

4. Misseri G, Piattoli M, Cuttone G, Gregoretti C, Bignami EG. Artificial Intelligence for Mechanical Ventilation: A Transformative Shift in Critical Care. Ther Adv Pulm Crit Care Med. 2024;19:29768675241298918. Published 2024 Nov 11. doi:10.1177/29768675241298918 https://journals.sagepub.com/doi/10.1177/29768675241298918  

5. Topol, E.J. High-performance medicine: the convergence of human and artificial intelligence. Nat Med 25, 44–56 (2019). https://doi.org/10.1038/s41591-018-0300-7  

6. Escobar GJ, Liu VX, Schuler A, Lawson B, Greene JD, Kipnis P. Automated Identification of Adults at Risk for In-Hospital Clinical Deterioration. N Engl J Med. 2020 Nov 12;383(20):1951-1960. doi: 10.1056/NEJMsa2001090. PMID: 33176085; PMCID: PMC7787261. https://www.nejm.org/doi/10.1056/NEJMsa2001090?url_ver=Z39.88-2003&rfr_id=ori:rid:crossref.org&rfr_dat=cr_pub%20%200pubmed  

7. Simonte R, Cammarota G, Vetrugno L, De Robertis E, Longhini F, Spadaro S. Advanced Respiratory Monitoring during Extracorporeal Membrane Oxygenation. J Clin Med. 2024 Apr 26;13(9):2541. doi: 10.3390/jcm13092541. PMID: 38731069; PMCID: PMC11084162. https://www.mdpi.com/2077-0383/13/9/2541  

8. Simonte, Rachelea; Cammarota, Gianmariab; De Robertis, Edoardoa. Intraoperative lung protection: strategies and their impact on outcomes. Current Opinion in Anaesthesiology 37(2):p 184-191, April 2024. | DOI: 10.1097/ACO.0000000000001341 https://journals.lww.com/co-anesthesiology/abstract/2024/04000/intraoperative_lung_protection__strategies_and.16.aspx  

9. Cammarota G, Simonte R, Longhini F, Spadaro S, Vetrugno L, De Robertis E. Advanced Point-of-care Bedside Monitoring for Acute Respiratory Failure. Anesthesiology. 2023 Mar 1;138(3):317-334. doi: 10.1097/ALN.0000000000004480. PMID: 36749422. https://journals.lww.com/anesthesiology/fulltext/2023/03000/advanced_point_of_care_bedside_monitoring_for.18.aspx  

10. Pikman Gavriely R, Freund O, Tiran B, Perluk TM, Kleinhendler E, Matot I, Bar-Shai A, Gershman E. Laryngeal mask airway or high-flow nasal cannula versus nasal cannula for advanced bronchoscopy: a randomised controlled trial. ERJ Open Res. 2025 Feb 10;11(1):00421-2024. doi: 10.1183/23120541.00421-2024. PMID: 39931666; PMCID: PMC11808932. https://pmc.ncbi.nlm.nih.gov/articles/PMC11808932/  

11. Pikman Gavriely R, Lior Y, Gelikas S, Levy S, Ahimor A, Glassberg E, Shapira S, Benov A, Avital G. Manual Pressure Points Technique for Massive Hemorrhage Control-A Prospective Human Volunteer Study. Prehosp Emerg Care. 2023;27(5):586-591. doi: 10.1080/10903127.2022.2122644. Epub 2022 Sep 28. PMID: 36074122. https://www.tandfonline.com/doi/10.1080/10903127.2022.2122644?url_ver=Z39.88-2003&rfr_id=ori:rid:crossref.org&rfr_dat=cr_pub%20%200pubmed  

12. Villar J, González-Martín JM, Fernández C, et al. Predicting the Length of Mechanical Ventilation in Acute Respiratory Disease Syndrome Using Machine Learning: The PIONEER Study. J Clin Med. 2024;13(6):1811. Published 2024 Mar 21. doi:10.3390/jcm13061811 https://pmc.ncbi.nlm.nih.gov/articles/PMC10971349/  

13. Villar J, González-Martín JM, Fernández C, et al. Predicting Outcome and Duration of Mechanical Ventilation in Acute Hypoxemic Respiratory Failure: The PREMIER Study. J Clin Med. 2025;14(22):7903. Published 2025 Nov 7. doi:10.3390/jcm14227903 https://pmc.ncbi.nlm.nih.gov/articles/PMC12653734/ 

14. Villar J, González-Martín JM, Fernández C, et al. Early Prediction of ICU Mortality in Patients with Acute Hypoxemic Respiratory Failure Using Machine Learning: The MEMORIAL Study. J Clin Med. 2025;14(5):1711. Published 2025 Mar 4. doi:10.3390/jcm14051711 https://www.mdpi.com/2077-0383/14/5/1711 

15. Villar J, González-Martín JM, Fernández C, et al. Predicting Outcome and Duration of Mechanical Ventilation in Acute Hypoxemic Respiratory Failure: The PREMIER Study. J Clin Med. 2025;14(22):7903. Published 2025 Nov 7. doi:10.3390/jcm14227903 https://www.mdpi.com/2077-0383/14/22/7903  

16. de Haro C, Santos-Pulpón V, Telías I, et al. Flow starvation during square-flow assisted ventilation detected by supervised deep learning techniques. Crit Care. 2024;28(1):75. Published 2024 Mar 14. doi:10.1186/s13054-024-04845-y https://link.springer.com/article/10.1186/s13054-024-04845-y