System Maintenance Notice

Please note that maintenance work is scheduled between 4 AM and 5 AM - CET. During this time, you may experience reduced performance. Thanks for your patience.

Publications

Journal Article Epublish

PubMed: 42816865 DOI: 10.1186/s40560-026-00947-7

Development and validation of VentPilot: an AI-based recommendation system for mechanical ventilation.

J Intensive Care 14(1)

Authors: Hong Yeul Lee, Gaon An, Yeonwoo Jeong, Edward J Schenck, Ho Geol Ryu, Brian W Pickering, Vitaly Herasevich, Daniel A Diedrich, Seungyong Moon, Hyun-Lim Yang, Hyung-Chul Lee, Wun Hur, Jaewon Choi, Seokjun Lee, Jeongjae Lee, Won-Seok Hur, Sebin Kim, Yoonshik Kim, Rajesh Ranganath, Kyunghyun Cho, Sang-Min Lee, Hyun Oh Song

Affiliations

  • Department of Critical Care Medicine, Seoul National University Hospital, 101 Daehak-ro, Jongno-gu, Seoul, 03080, Republic of Korea.
  • Department of Computer Science and Engineering, Seoul National University, 1 Gwanak-ro, Gwanak-gu, Seoul, 08826, Republic of Korea.
  • Division of Pulmonary and Critical Care Medicine, Department of Medicine, Weill Cornell Medicine, 1300 York Avenue, New York, NY, 10065, USA.
  • Department of Anesthesiology and Pain Medicine, Seoul National University College of Medicine, 103 Daehak-ro, Jongno-gu, Seoul, 03080, Republic of Korea.
  • Department of Anesthesiology and Perioperative Medicine, Mayo Clinic, 200 First Street SW, Rochester, MN, 55905, USA.
  • Department of Biomedical Engineering, College of Medicine, Chungnam National University, 266 Munhwa-ro, Jung-gu, Daejeon, 35015, Republic of Korea.
  • Department of Data Science Research, Innovative Medical Technology Research Institute, Seoul National University Hospital, 214 Yulgok-ro, Jongno-gu, Seoul, 03122, Republic of Korea.
  • Department of Electrical and Computer Engineering, Seoul National University, 1 Gwanak-ro, Gwanak-gu, Seoul, 08826, Republic of Korea.
  • Center for Data Science, New York University, 60 5th Avenue, New York, NY, 10011, USA.
  • Courant Institute of Mathematical Sciences, New York University, 251 Mercer Street, New York, NY, 10012-1185, USA.
  • Department of Critical Care Medicine, Seoul National University Hospital, 101 Daehak-ro, Jongno-gu, Seoul, 03080, Republic of Korea.
  • Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Seoul National University Hospital, 101 Daehak-ro, Jongno-gu, Seoul, 03080, Republic of Korea.
  • Department of Computer Science and Engineering, Seoul National University, 1 Gwanak-ro, Gwanak-gu, Seoul, 08826, Republic of Korea.

Abstract

BACKGROUND

Mechanical ventilation requires repeated adjustment to changing patient physiology, but consistent individualized management remains challenging. We developed VentPilot, an artificial intelligence-based system for recommending ventilator settings, and evaluated it in multicenter retrospective validation cohorts.

METHODS

VentPilot was developed using offline reinforcement learning on high-resolution physiologic and ventilator trajectories from Seoul National University Hospital (SNUH), with a reward combining ventilator-free days and intensivist-derived preference feedback. We evaluated VentPilot in an internal validation cohort from SNUH and an external validation cohort from Mayo Clinic. Fitted Q-evaluation assessed the estimated return of VentPilot relative to observed clinician behavior under the prespecified reward function. In a complementary inverse probability-weighted analysis, clinical outcomes were compared between patients with higher versus lower concordance between observed ventilator settings and VentPilot recommendations.

RESULTS

Among 4296 mechanically ventilated adults, 3002 were included in the derivation cohort, 283 in the internal validation cohort, and 1011 in the external validation cohort. Under the prespecified reward function, fitted Q-evaluation estimated higher returns for VentPilot than for observed clinician behavior in both validation cohorts, with differences in overall reward of 1.6 (95% CI 0.8-2.5) and 1.3 (95% CI 0.6-2.0), respectively. In inverse probability-weighted analyses, higher concordance with VentPilot recommendations was associated with more ventilator-free days within 28 days, with mean differences of 5.0 days (95% CI 2.4-7.6) and 3.1 days (95% CI 1.7-4.6), respectively. Higher concordance was also associated with lower 28-day mortality and shorter ICU length of stay.

CONCLUSIONS

In multicenter retrospective validation cohorts, fitted Q-evaluation estimated higher returns for VentPilot than for observed clinician behavior under the prespecified reward function, while greater concordance between observed care and VentPilot recommendations was associated with more favorable clinical outcomes. Further clinical evaluation is warranted to establish VentPilot's safety, usability, and clinical impact.

MeSH

Probability, Reward, Ventilation