Format

Send to

Choose Destination
Nat Med. 2018 Nov;24(11):1716-1720. doi: 10.1038/s41591-018-0213-5. Epub 2018 Oct 22.

The Artificial Intelligence Clinician learns optimal treatment strategies for sepsis in intensive care.

Komorowski M1,2,3, Celi LA3,4, Badawi O3,5,6, Gordon AC7, Faisal AA8,9,10,11.

Author information

1
Department of Surgery and Cancer, Imperial College London, London, UK.
2
Department of Bioengineering, Imperial College London, London, UK.
3
Laboratory of Computational Physiology, Harvard-MIT Division of Health Sciences & Technology, Cambridge, MA, USA.
4
Beth Israel Deaconess Medical Center, Boston, MA, USA.
5
Department of eICU Research and Development, Philips Healthcare, Baltimore, MD, USA.
6
Department of Pharmacy Practice and Science, University of Maryland, School of Pharmacy, Baltimore, MD, USA.
7
Department of Surgery and Cancer, Imperial College London, London, UK. anthony.gordon@imperial.ac.uk.
8
Department of Bioengineering, Imperial College London, London, UK. a.faisal@imperial.ac.uk.
9
Department of Computer Science, Imperial College London, London, UK. a.faisal@imperial.ac.uk.
10
Medical Research Council London Institute of Medical Sciences, London, UK. a.faisal@imperial.ac.uk.
11
Behaviour Analytics Lab, Data Science Institute, London, UK. a.faisal@imperial.ac.uk.

Abstract

Sepsis is the third leading cause of death worldwide and the main cause of mortality in hospitals1-3, but the best treatment strategy remains uncertain. In particular, evidence suggests that current practices in the administration of intravenous fluids and vasopressors are suboptimal and likely induce harm in a proportion of patients1,4-6. To tackle this sequential decision-making problem, we developed a reinforcement learning agent, the Artificial Intelligence (AI) Clinician, which extracted implicit knowledge from an amount of patient data that exceeds by many-fold the life-time experience of human clinicians and learned optimal treatment by analyzing a myriad of (mostly suboptimal) treatment decisions. We demonstrate that the value of the AI Clinician's selected treatment is on average reliably higher than human clinicians. In a large validation cohort independent of the training data, mortality was lowest in patients for whom clinicians' actual doses matched the AI decisions. Our model provides individualized and clinically interpretable treatment decisions for sepsis that could improve patient outcomes.

PMID:
30349085
DOI:
10.1038/s41591-018-0213-5

Supplemental Content

Full text links

Icon for Nature Publishing Group
Loading ...
Support Center