Using artificial neural networks for classifying ICU patient states

Mark van Gils, Holger Jansen, Kari Nieminen, Ron Summers, Peter Weller

Research output: Contribution to journalArticleScientificpeer-review

15 Citations (Scopus)

Abstract

The rapid accurate diagnosis of critical disorders is an essential component of intensive care. Traditional diagnostic techniques have relied on physician experience, which is based on a data set chosen from his or her personal preferences, rather than from scientific merit. In this article, we show that there are alternative methods of selecting clinical variables on which to base a diagnosis. We suggest that a model-based technique utilizing artificial neural networks (ANNs) can be used to investigate alternative, objectively chosen data input sets. Traditionally, ANNs have been used for diagnosis or prediction tasks; however, this article introduces a novel method of exploring the inner structure of suitably trained ANNs to determine a set of key variables for each clinical state defined. Two different ANN techniques are proposed: self-organizing maps and backpropagation networks. We do not claim that these techniques provide the optimal data set for decision making, but we do show that other combinations of data exist that may be an improvement over the physician methods currently used.
Original languageEnglish
Pages (from-to)41 - 47
Number of pages7
JournalIEEE Engineering in Medicine and Biology Magazine
Volume16
Issue number6
DOIs
Publication statusPublished - 1997
MoE publication typeA1 Journal article-refereed

Fingerprint

Intensive care units
Neural networks
Self organizing maps
Backpropagation
Decision making

Cite this

van Gils, Mark ; Jansen, Holger ; Nieminen, Kari ; Summers, Ron ; Weller, Peter. / Using artificial neural networks for classifying ICU patient states. In: IEEE Engineering in Medicine and Biology Magazine. 1997 ; Vol. 16, No. 6. pp. 41 - 47.
@article{926abcf67fa948c3bfc1900cc43c602d,
title = "Using artificial neural networks for classifying ICU patient states",
abstract = "The rapid accurate diagnosis of critical disorders is an essential component of intensive care. Traditional diagnostic techniques have relied on physician experience, which is based on a data set chosen from his or her personal preferences, rather than from scientific merit. In this article, we show that there are alternative methods of selecting clinical variables on which to base a diagnosis. We suggest that a model-based technique utilizing artificial neural networks (ANNs) can be used to investigate alternative, objectively chosen data input sets. Traditionally, ANNs have been used for diagnosis or prediction tasks; however, this article introduces a novel method of exploring the inner structure of suitably trained ANNs to determine a set of key variables for each clinical state defined. Two different ANN techniques are proposed: self-organizing maps and backpropagation networks. We do not claim that these techniques provide the optimal data set for decision making, but we do show that other combinations of data exist that may be an improvement over the physician methods currently used.",
author = "{van Gils}, Mark and Holger Jansen and Kari Nieminen and Ron Summers and Peter Weller",
note = "Project code: T5SU000254",
year = "1997",
doi = "10.1109/51.637116",
language = "English",
volume = "16",
pages = "41 -- 47",
journal = "IEEE Pulse",
issn = "2154-2287",
publisher = "Institute of Electrical and Electronic Engineers IEEE",
number = "6",

}

Using artificial neural networks for classifying ICU patient states. / van Gils, Mark; Jansen, Holger; Nieminen, Kari; Summers, Ron; Weller, Peter.

In: IEEE Engineering in Medicine and Biology Magazine, Vol. 16, No. 6, 1997, p. 41 - 47.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Using artificial neural networks for classifying ICU patient states

AU - van Gils, Mark

AU - Jansen, Holger

AU - Nieminen, Kari

AU - Summers, Ron

AU - Weller, Peter

N1 - Project code: T5SU000254

PY - 1997

Y1 - 1997

N2 - The rapid accurate diagnosis of critical disorders is an essential component of intensive care. Traditional diagnostic techniques have relied on physician experience, which is based on a data set chosen from his or her personal preferences, rather than from scientific merit. In this article, we show that there are alternative methods of selecting clinical variables on which to base a diagnosis. We suggest that a model-based technique utilizing artificial neural networks (ANNs) can be used to investigate alternative, objectively chosen data input sets. Traditionally, ANNs have been used for diagnosis or prediction tasks; however, this article introduces a novel method of exploring the inner structure of suitably trained ANNs to determine a set of key variables for each clinical state defined. Two different ANN techniques are proposed: self-organizing maps and backpropagation networks. We do not claim that these techniques provide the optimal data set for decision making, but we do show that other combinations of data exist that may be an improvement over the physician methods currently used.

AB - The rapid accurate diagnosis of critical disorders is an essential component of intensive care. Traditional diagnostic techniques have relied on physician experience, which is based on a data set chosen from his or her personal preferences, rather than from scientific merit. In this article, we show that there are alternative methods of selecting clinical variables on which to base a diagnosis. We suggest that a model-based technique utilizing artificial neural networks (ANNs) can be used to investigate alternative, objectively chosen data input sets. Traditionally, ANNs have been used for diagnosis or prediction tasks; however, this article introduces a novel method of exploring the inner structure of suitably trained ANNs to determine a set of key variables for each clinical state defined. Two different ANN techniques are proposed: self-organizing maps and backpropagation networks. We do not claim that these techniques provide the optimal data set for decision making, but we do show that other combinations of data exist that may be an improvement over the physician methods currently used.

U2 - 10.1109/51.637116

DO - 10.1109/51.637116

M3 - Article

VL - 16

SP - 41

EP - 47

JO - IEEE Pulse

JF - IEEE Pulse

SN - 2154-2287

IS - 6

ER -