Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Artificial neural networks improve and simplify intensive care mortality prognostication: a national cohort study of 217,289 first-time intensive care unit admissions

Holmgren, Gustav LU ; Andersson, Peder LU orcid ; Jakobsson, Andreas LU orcid and Frigyesi, Attila LU (2019) In Journal of Intensive Care 7.
Abstract
Purpose
We investigated if early intensive care unit (ICU) scoring with the Simplified Acute Physiology Score (SAPS 3) could be improved using artificial neural networks (ANNs). Methods All first-time adult intensive care admissions in Sweden during 2009-2017 were included. A test set was set aside for validation. We trained ANNs with two hidden layers with random hyper-parameters and retained the best ANN, determined using cross-validation. The ANNs were constructed using the same parameters as in the SAPS 3 model. The performance was assessed with the area under the receiver operating characteristic curve (AUC) and Brier score.
Results
A total of 217,289 admissions were included. The developed ANN (AUC 0.89 and Brier score... (More)
Purpose
We investigated if early intensive care unit (ICU) scoring with the Simplified Acute Physiology Score (SAPS 3) could be improved using artificial neural networks (ANNs). Methods All first-time adult intensive care admissions in Sweden during 2009-2017 were included. A test set was set aside for validation. We trained ANNs with two hidden layers with random hyper-parameters and retained the best ANN, determined using cross-validation. The ANNs were constructed using the same parameters as in the SAPS 3 model. The performance was assessed with the area under the receiver operating characteristic curve (AUC) and Brier score.
Results
A total of 217,289 admissions were included. The developed ANN (AUC 0.89 and Brier score 0.096) was found to be superior (p textless10-15 for AUC and p textless10-5 for Brier score) in early prediction of 30-day mortality for intensive care patients when compared with SAPS 3 (AUC 0.85 and Brier score 0.109). In addition, a simple, eight-parameter ANN model was found to perform just as well as SAPS 3, but with better calibration (AUC 0.85 and and Brier score 0.106, p textless10-5). Furthermore, the ANN model was superior in correcting mortality for age.
Conclusion
ANNs can outperform the SAPS 3 model for early prediction of 30-day mortality for intensive care patients. (Less)
Please use this url to cite or link to this publication:
author
; ; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
artificial intelligence, artificial neural networks, critical care, intensive care, machine learning, mortality, prediction, survival
in
Journal of Intensive Care
volume
7
article number
44
pages
8 pages
publisher
BioMed Central (BMC)
external identifiers
  • scopus:85070980594
  • pmid:31428430
ISSN
2052-0492
DOI
10.1186/s40560-019-0393-1
language
English
LU publication?
yes
id
486bda86-9c1e-4a44-ac83-b55528f96d7a
date added to LUP
2019-09-10 10:43:49
date last changed
2022-04-26 05:11:44
@article{486bda86-9c1e-4a44-ac83-b55528f96d7a,
  abstract     = {{Purpose <br/>We investigated if early intensive care unit (ICU) scoring with the Simplified Acute Physiology Score (SAPS 3) could be improved using artificial neural networks (ANNs). Methods All first-time adult intensive care admissions in Sweden during 2009-2017 were included. A test set was set aside for validation. We trained ANNs with two hidden layers with random hyper-parameters and retained the best ANN, determined using cross-validation. The ANNs were constructed using the same parameters as in the SAPS 3 model. The performance was assessed with the area under the receiver operating characteristic curve (AUC) and Brier score. <br/>Results <br/>A total of 217,289 admissions were included. The developed ANN (AUC 0.89 and Brier score 0.096) was found to be superior (p textless10-15 for AUC and p textless10-5 for Brier score) in early prediction of 30-day mortality for intensive care patients when compared with SAPS 3 (AUC 0.85 and Brier score 0.109). In addition, a simple, eight-parameter ANN model was found to perform just as well as SAPS 3, but with better calibration (AUC 0.85 and and Brier score 0.106, p textless10-5). Furthermore, the ANN model was superior in correcting mortality for age. <br/>Conclusion <br/>ANNs can outperform the SAPS 3 model for early prediction of 30-day mortality for intensive care patients.}},
  author       = {{Holmgren, Gustav and Andersson, Peder and Jakobsson, Andreas and Frigyesi, Attila}},
  issn         = {{2052-0492}},
  keywords     = {{artificial intelligence; artificial neural networks; critical care; intensive care; machine learning; mortality; prediction; survival}},
  language     = {{eng}},
  publisher    = {{BioMed Central (BMC)}},
  series       = {{Journal of Intensive Care}},
  title        = {{Artificial neural networks improve and simplify intensive care mortality prognostication: a national cohort study of 217,289 first-time intensive care unit admissions}},
  url          = {{http://dx.doi.org/10.1186/s40560-019-0393-1}},
  doi          = {{10.1186/s40560-019-0393-1}},
  volume       = {{7}},
  year         = {{2019}},
}