skip to main content
DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: A minimal set of physiomarkers in continuous high frequency data streams predict adult sepsis onset earlier

Abstract

Purpose: We report that sepsis is a life-threatening condition with high mortality rates and expensive treatment costs. To improve short- and long-term outcomes, it is critical to detect at-risk sepsis patients at an early stage.Methods: A data-set consisting of high-frequency physiological data from 1161 critically ill patients was analyzed. 377 patients had developed sepsis, and had data at least 3 h prior to the onset of sepsis. A random forest classifier was trained to discriminate between sepsis and non-sepsis patients in real-time using a total of 132 features extracted from a moving time-window. The model was trained on 80% of the patients and was tested on the remaining 20% of the patients, for two observational periods of lengths 3 and 6 h prior to onset.Results: The model that used continuous physiological data alone resulted in sensitivity and F1 score of up to 80% and 67% one hour before sepsis onset. On average, these models were able to predict sepsis 294.19 ± 6.50 min (5 h) before the onset.Conclusions: Lastly, the use of machine learning algorithms on continuous streams of physiological data can allow for early identification of at-risk patients in real-time with high accuracy.

Authors:
 [1]; ORCiD logo [1];  [2];  [3];  [2]; ORCiD logo [2]
  1. Univ. of Tennessee, Knoxville, TN (United States)
  2. University of Tennessee Health, Memphis, TN (United States)
  3. Univ. of Tennessee, Knoxville, TN (United States); Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States)
Publication Date:
Research Org.:
Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
1502531
Grant/Contract Number:  
AC05-00OR22725
Resource Type:
Accepted Manuscript
Journal Name:
International Journal of Medical Informatics
Additional Journal Information:
Journal Volume: 122; Journal Issue: C; Journal ID: ISSN 1386-5056
Publisher:
Elsevier
Country of Publication:
United States
Language:
English
Subject:
60 APPLIED LIFE SCIENCES; 97 MATHEMATICS AND COMPUTING; 59 BASIC BIOLOGICAL SCIENCES

Citation Formats

van Wyk, Franco, Khojandi, Anahita, Mohammed, Akram, Begoli, Edmon, Davis, Robert L., and Kamaleswaran, Rishikesan. A minimal set of physiomarkers in continuous high frequency data streams predict adult sepsis onset earlier. United States: N. p., 2018. Web. https://doi.org/10.1016/j.ijmedinf.2018.12.002.
van Wyk, Franco, Khojandi, Anahita, Mohammed, Akram, Begoli, Edmon, Davis, Robert L., & Kamaleswaran, Rishikesan. A minimal set of physiomarkers in continuous high frequency data streams predict adult sepsis onset earlier. United States. https://doi.org/10.1016/j.ijmedinf.2018.12.002
van Wyk, Franco, Khojandi, Anahita, Mohammed, Akram, Begoli, Edmon, Davis, Robert L., and Kamaleswaran, Rishikesan. Mon . "A minimal set of physiomarkers in continuous high frequency data streams predict adult sepsis onset earlier". United States. https://doi.org/10.1016/j.ijmedinf.2018.12.002. https://www.osti.gov/servlets/purl/1502531.
@article{osti_1502531,
title = {A minimal set of physiomarkers in continuous high frequency data streams predict adult sepsis onset earlier},
author = {van Wyk, Franco and Khojandi, Anahita and Mohammed, Akram and Begoli, Edmon and Davis, Robert L. and Kamaleswaran, Rishikesan},
abstractNote = {Purpose: We report that sepsis is a life-threatening condition with high mortality rates and expensive treatment costs. To improve short- and long-term outcomes, it is critical to detect at-risk sepsis patients at an early stage.Methods: A data-set consisting of high-frequency physiological data from 1161 critically ill patients was analyzed. 377 patients had developed sepsis, and had data at least 3 h prior to the onset of sepsis. A random forest classifier was trained to discriminate between sepsis and non-sepsis patients in real-time using a total of 132 features extracted from a moving time-window. The model was trained on 80% of the patients and was tested on the remaining 20% of the patients, for two observational periods of lengths 3 and 6 h prior to onset.Results: The model that used continuous physiological data alone resulted in sensitivity and F1 score of up to 80% and 67% one hour before sepsis onset. On average, these models were able to predict sepsis 294.19 ± 6.50 min (5 h) before the onset.Conclusions: Lastly, the use of machine learning algorithms on continuous streams of physiological data can allow for early identification of at-risk patients in real-time with high accuracy.},
doi = {10.1016/j.ijmedinf.2018.12.002},
journal = {International Journal of Medical Informatics},
number = C,
volume = 122,
place = {United States},
year = {2018},
month = {12}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record

Citation Metrics:
Cited by: 7 works
Citation information provided by
Web of Science

Save / Share:

Works referenced in this record:

A New Initiative on Precision Medicine
journal, February 2015

  • Collins, Francis S.; Varmus, Harold
  • New England Journal of Medicine, Vol. 372, Issue 9
  • DOI: 10.1056/NEJMp1500523

Comparing deep neural network and other machine learning algorithms for stroke prediction in a large-scale population-based electronic medical claims database
conference, July 2017

  • Hung, Chen-Ying; Chen, Wei-Chen; Lai, Po-Tsun
  • 2017 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)
  • DOI: 10.1109/EMBC.2017.8037515

Precision Medicine—Personalized, Problematic, and Promising
journal, January 2015


Big data analytics in healthcare: promise and potential
journal, February 2014

  • Raghupathi, Wullianallur; Raghupathi, Viju
  • Health Information Science and Systems, Vol. 2, Issue 1
  • DOI: 10.1186/2047-2501-2-3

Big Data in Healthcare
journal, January 2016


Specific Etiologies Associated With the Multiple Organ Dysfunction Syndrome in Children: Part 1
journal, January 2017


Clinical Mimics: An Emergency Medicine–Focused Review of Sepsis Mimics
journal, January 2017


Assessing available information on the burden of sepsis: global estimates of incidence, prevalence and mortality
journal, June 2012

  • Jawad, Issrah; Lukšić, Ivana; Rafnsson, Snorri Bjorn
  • Journal of Global Health, Vol. 2, Issue 1
  • DOI: 10.7189/jogh.01.010404

Early sepsis detection in critical care patients using multiscale blood pressure and heart rate dynamics
journal, November 2017


Reduction of false arrhythmia alarms using signal selection and machine learning
journal, July 2016

  • Eerikäinen, Linda M.; Vanschoren, Joaquin; Rooijakkers, Michael J.
  • Physiological Measurement, Vol. 37, Issue 8
  • DOI: 10.1088/0967-3334/37/8/1204

Predicting disease risks from highly imbalanced data using random forest
journal, July 2011

  • Khalilia, Mohammed; Chakraborty, Sounak; Popescu, Mihail
  • BMC Medical Informatics and Decision Making, Vol. 11, Issue 1
  • DOI: 10.1186/1472-6947-11-51

A survey on deep learning in medical image analysis
journal, December 2017


Mitosis detection in breast cancer pathology images by combining handcrafted and convolutional neural network features
journal, October 2014

  • Wang, Haibo; Cruz-Roa, Angel; Basavanhally, Ajay
  • Journal of Medical Imaging, Vol. 1, Issue 3
  • DOI: 10.1117/1.JMI.1.3.034003

A Multilayer Perceptron Based Smart Pathological Brain Detection System by Fractional Fourier Entropy
journal, June 2016


Learning representations for the early detection of sepsis with deep neural networks
journal, October 2017


A robust deep convolutional neural network for the classification of abnormal cardiac rhythm using single lead electrocardiograms of variable length
journal, March 2018

  • Kamaleswaran, Rishikesan; Mahajan, Ruhi; Akbilgic, Oguz
  • Physiological Measurement, Vol. 39, Issue 3
  • DOI: 10.1088/1361-6579/aaaa9d

How much data should we collect? A case study in sepsis detection using deep learning
conference, November 2017

  • van Wyk, Franco; Khojandi, Anahita; Kamaleswaran, Rishikesan
  • 2017 IEEE Healthcare Innovations and Point-of-Care Technologies (HI-POCT), 2017 IEEE Healthcare Innovations and Point of Care Technologies (HI-POCT)
  • DOI: 10.1109/HIC.2017.8227596

An Interpretable Machine Learning Model for Accurate Prediction of Sepsis in the ICU
journal, January 2018


A targeted real-time early warning score (TREWScore) for septic shock
journal, August 2015

  • Henry, Katharine E.; Hager, David N.; Pronovost, Peter J.
  • Science Translational Medicine, Vol. 7, Issue 299
  • DOI: 10.1126/scitranslmed.aab3719

Abnormal Heart Rate Characteristics Preceding Neonatal Sepsis and Sepsis-Like Illness
journal, June 2003


Continuous Multi-Parameter Heart Rate Variability Analysis Heralds Onset of Sepsis in Adults
journal, August 2009


Applying Artificial Intelligence to Identify Physiomarkers Predicting Severe Sepsis in the PICU*
journal, January 2018

  • Kamaleswaran, Rishikesan; Akbilgic, Oguz; Hallman, Madhura A.
  • Pediatric Critical Care Medicine, Vol. 19, Issue 10
  • DOI: 10.1097/PCC.0000000000001666

Raising concerns about the Sepsis-3 definitions
journal, January 2018

  • Sartelli, Massimo; Kluger, Yoram; Ansaloni, Luca
  • World Journal of Emergency Surgery, Vol. 13, Issue 1
  • DOI: 10.1186/s13017-018-0165-6

Comparison of WBC, ESR, CRP and PCT serum levels in septic and non-septic burn cases
journal, September 2008


Random Forests
journal, January 2001


Classification and Regression Trees.
journal, September 1984

  • Gordon, A. D.; Breiman, L.; Friedman, J. H.
  • Biometrics, Vol. 40, Issue 3
  • DOI: 10.2307/2530946

Bagging predictors
journal, August 1996


Deep learning
journal, May 2015

  • LeCun, Yann; Bengio, Yoshua; Hinton, Geoffrey
  • Nature, Vol. 521, Issue 7553
  • DOI: 10.1038/nature14539

Python for Scientific Computing
journal, January 2007


    Works referencing / citing this record:

    Differential gene expression analysis reveals novel genes and pathways in pediatric septic shock patients
    journal, August 2019


    Computational Health Engineering Applied to Model Infectious Diseases and Antimicrobial Resistance Spread
    journal, June 2019

    • Cartelle Gestal, Mónica; Dedloff, Margaret R.; Torres-Sangiao, Eva
    • Applied Sciences, Vol. 9, Issue 12
    • DOI: 10.3390/app9122486

    Machine learning for the prediction of sepsis: a systematic review and meta-analysis of diagnostic test accuracy
    journal, January 2020

    • Fleuren, Lucas M.; Klausch, Thomas L. T.; Zwager, Charlotte L.
    • Intensive Care Medicine, Vol. 46, Issue 3
    • DOI: 10.1007/s00134-019-05872-y

    Using Machine Learning to Predict Early Onset Acute Organ Failure in Critically Ill Intensive Care Unit Patients With Sickle Cell Disease: Retrospective Study
    journal, January 2020

    • Mohammed, Akram; Podila, Pradeep S. B.; Davis, Robert L.
    • Journal of Medical Internet Research, Vol. 22, Issue 5
    • DOI: 10.2196/14693