TY - JOUR
T1 - Deep Interpretable Early Warning System for the Detection of Clinical Deterioration
AU - Shamout, Farah E.
AU - Zhu, Tingting
AU - Sharma, Pulkit
AU - Watkinson, Peter J.
AU - Clifton, David A.
N1 - Funding Information:
Manuscript received February 20, 2019; revised June 7, 2019 and July 28, 2019; accepted August 16, 2019. Date of publication September 19, 2019; date of current version February 6, 2020. The work of F. E. Shamout was supported by the Rhodes Trust and the P. J. Watkin-son was supported by the NIHR Biomedical Research Centre, Oxford. Disclaimer: This publication presents independent research commissioned by the Health Innovation Challenge Fund (HICF-R9-524; WT-103703/Z/14/Z), a parallel funding partnership between the Department of Health and Wellcome Trust. The views expressed in this publication are those of the author(s) and not necessarily those of the Department of Health or Wellcome Trust. (Corresponding author: Farah E. Shamout.) F. E. Shamout, T. Zhu, P. Sharma, and D. A. Clifton are with the Institute of Biomedical Engineering, Department of Engineering Science, University of Oxford, OX1 2JD Oxford, U.K. (e-mail: farah.shamout@ eng; [email protected]; [email protected]; david. [email protected]).
Publisher Copyright:
© 2013 IEEE.
PY - 2020/2
Y1 - 2020/2
N2 - Assessment of physiological instability preceding adverse events on hospital wards has been previously investigated through clinical early warning score systems. Early warning scores are simple to use yet they consider data as independent and identically distributed random variables. Deep learning applications are able to learn from sequential data, however they lack interpretability and are thus difficult to deploy in clinical settings. We propose the 'Deep Early Warning System' (DEWS), an interpretable end-to-end deep learning model that interpolates temporal data and predicts the probability of an adverse event, defined as the composite outcome of cardiac arrest, mortality or unplanned ICU admission. The model was developed and validated using routinely collected vital signs of patients admitted to the the Oxford University Hospitals between 21st March 2014 and 31st March 2018. We extracted 45 314 vital-sign measurements as a balanced training set and 359 481 vital-sign measurements as an imbalanced testing set to mimic a real-life setting of emergency admissions. DEWS achieved superior accuracy than the state-of-the-art that is currently implemented in clinical settings, the National Early Warning Score, in terms of the overall area under the receiver operating characteristic curve (AUROC) (0.880 vs. 0.866) and when evaluated independently for each of the three outcomes. Our attention-based architecture was able to recognize 'historical' trends in the data that are most correlated with the predicted probability. With high sensitivity, improved clinical utility and increased interpretability, our model can be easily deployed in clinical settings to supplement existing EWS systems.
AB - Assessment of physiological instability preceding adverse events on hospital wards has been previously investigated through clinical early warning score systems. Early warning scores are simple to use yet they consider data as independent and identically distributed random variables. Deep learning applications are able to learn from sequential data, however they lack interpretability and are thus difficult to deploy in clinical settings. We propose the 'Deep Early Warning System' (DEWS), an interpretable end-to-end deep learning model that interpolates temporal data and predicts the probability of an adverse event, defined as the composite outcome of cardiac arrest, mortality or unplanned ICU admission. The model was developed and validated using routinely collected vital signs of patients admitted to the the Oxford University Hospitals between 21st March 2014 and 31st March 2018. We extracted 45 314 vital-sign measurements as a balanced training set and 359 481 vital-sign measurements as an imbalanced testing set to mimic a real-life setting of emergency admissions. DEWS achieved superior accuracy than the state-of-the-art that is currently implemented in clinical settings, the National Early Warning Score, in terms of the overall area under the receiver operating characteristic curve (AUROC) (0.880 vs. 0.866) and when evaluated independently for each of the three outcomes. Our attention-based architecture was able to recognize 'historical' trends in the data that are most correlated with the predicted probability. With high sensitivity, improved clinical utility and increased interpretability, our model can be easily deployed in clinical settings to supplement existing EWS systems.
KW - Early warning system
KW - data interpolation
KW - deep learning
KW - supervised learning
KW - time-series data
UR - http://www.scopus.com/inward/record.url?scp=85079104660&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85079104660&partnerID=8YFLogxK
U2 - 10.1109/JBHI.2019.2937803
DO - 10.1109/JBHI.2019.2937803
M3 - Article
C2 - 31545746
AN - SCOPUS:85079104660
SN - 2168-2194
VL - 24
SP - 437
EP - 446
JO - IEEE Journal of Biomedical and Health Informatics
JF - IEEE Journal of Biomedical and Health Informatics
IS - 2
M1 - 8844833
ER -