以類神經網路模式輔助外科重症病人預測死亡機率

Translated title of the contribution: Predicting Mortality in Surgical Intensive Care Unit Patients Using an Artificial Neural Network

Hui-Ju Chen, Hung-Wen Chiu, Wen-Jinn Liaw

Research output: Contribution to journalArticle

Abstract

Background: The aim of this study was to investigate predictors of mortality inpatients admitted to a surgical Intensive Care Unit (SICU). Methods: An artificial neural network model was constructed on 588 consecutive patients treated in an ICU of a medical center during Jan to May, 2012. The prognostic factors, retrieved from IntelliVue Clinical Information Portfolio (ICIP), included days of ICU stay, causes of transfer to ICU, APACHE II score at 24hrs stay and 48hrs stay in ICU, and the day of discharge from ICU. The data was randomly divided into a 441 patient training dataset and a 147 patient testing dataset. The best predicting model was established using MLP (MLP 59-19-2); the output variable was death (1, yes; 2, no). Results: The accuracy of mortality prediction was 96.9%; the area under the ROC curve was 0.975; sensitivity was 78.8%, and specificity was 98.7%. Conclusion: Our results show disease severity is highly correlated with mortality. The accuracy, sensitivity, and specificity are all also very high. In addition, this study demonstrates patients with nonintracranial hemorrhage disclosed less unpredictable mortality, but doctors can easier predict poor outcome in patients with intracranial hemorrhage and thus recommend hospice care to the family.
Original languageTraditional Chinese
Pages (from-to)1-7
Number of pages7
Journal重症醫學雜誌
Volume14
Issue number1
Publication statusPublished - 2013

Fingerprint

Critical Care
Intensive Care Units
Mortality
Hospice Care
APACHE
Neural Networks (Computer)
Intracranial Hemorrhages
ROC Curve
Area Under Curve
Inpatients
Hemorrhage
Sensitivity and Specificity
Datasets

Cite this

以類神經網路模式輔助外科重症病人預測死亡機率. / Chen, Hui-Ju; Chiu, Hung-Wen; Liaw, Wen-Jinn .

In: 重症醫學雜誌, Vol. 14, No. 1, 2013, p. 1-7.

Research output: Contribution to journalArticle

Chen, Hui-Ju ; Chiu, Hung-Wen ; Liaw, Wen-Jinn . / 以類神經網路模式輔助外科重症病人預測死亡機率. In: 重症醫學雜誌. 2013 ; Vol. 14, No. 1. pp. 1-7.
@article{ddbcb2c08ccc4b1ebbc371d2b936ae8c,
title = "以類神經網路模式輔助外科重症病人預測死亡機率",
abstract = "背景:本文以類神經網路(Artificial Neural Network, ANN)資料探勘(Datamining)演算法分析2012年1-5月某醫學中心外科加護病房轉入病人(n=588)之基本人口學資料、入住加護病房日數、轉入加護病房原因及住院24小時、48小時及轉出(或出院當日)之疾病嚴重度等7個變項建構預測模式,以探討影響死亡預測因子之相關數值。方法:資料來源為2012年1-5月轉入外科加護病房之病人共588人,其中死亡人數共54人,死亡率為9.1{\%}。以STATISTICA軟體隨機選取441位病人作為訓練組,佔全體病人的75{\%};147位病人作為測試組,佔全體病人的25{\%}。利用多層次類神經網路(MLP)方式建立最佳化的預測模型為MLP 59-19-2,輸出變數:病人出院時死亡、病危自動出院為1,存活為2。結果:疾病嚴重度分數高確實有較高之死亡機會,死亡預測準確率(Accuracy)達96.9{\%},訓練組ROC曲線下面積為0.975,靈敏度(Sensitivity)為78.8{\%},特異性(Specificity)為98.7{\%}。結論:本研究結果證實疾病嚴重度確實與較高之死亡相關,其準確率、靈敏度與特異性皆非常高。此外,本研究結果顯示,非顱內出血相關疾病之病人,較少發生非預期之死亡;而顱內出血之病人,則醫師較能提前預知疾病預後不佳,並進而對家屬推薦重症安寧照顧理念。",
keywords = "加護病房, 預測, 死亡, 類神經網路, 資料探勘, Intensive care unit, Prediction, Mortality, Artificial neural network, Datamining",
author = "Hui-Ju Chen and Hung-Wen Chiu and Wen-Jinn Liaw",
year = "2013",
language = "繁體中文",
volume = "14",
pages = "1--7",
journal = "重症醫學雜誌",
issn = "1563-356x",
publisher = "中華民國重症醫學會",
number = "1",

}

TY - JOUR

T1 - 以類神經網路模式輔助外科重症病人預測死亡機率

AU - Chen, Hui-Ju

AU - Chiu, Hung-Wen

AU - Liaw, Wen-Jinn

PY - 2013

Y1 - 2013

N2 - 背景:本文以類神經網路(Artificial Neural Network, ANN)資料探勘(Datamining)演算法分析2012年1-5月某醫學中心外科加護病房轉入病人(n=588)之基本人口學資料、入住加護病房日數、轉入加護病房原因及住院24小時、48小時及轉出(或出院當日)之疾病嚴重度等7個變項建構預測模式,以探討影響死亡預測因子之相關數值。方法:資料來源為2012年1-5月轉入外科加護病房之病人共588人,其中死亡人數共54人,死亡率為9.1%。以STATISTICA軟體隨機選取441位病人作為訓練組,佔全體病人的75%;147位病人作為測試組,佔全體病人的25%。利用多層次類神經網路(MLP)方式建立最佳化的預測模型為MLP 59-19-2,輸出變數:病人出院時死亡、病危自動出院為1,存活為2。結果:疾病嚴重度分數高確實有較高之死亡機會,死亡預測準確率(Accuracy)達96.9%,訓練組ROC曲線下面積為0.975,靈敏度(Sensitivity)為78.8%,特異性(Specificity)為98.7%。結論:本研究結果證實疾病嚴重度確實與較高之死亡相關,其準確率、靈敏度與特異性皆非常高。此外,本研究結果顯示,非顱內出血相關疾病之病人,較少發生非預期之死亡;而顱內出血之病人,則醫師較能提前預知疾病預後不佳,並進而對家屬推薦重症安寧照顧理念。

AB - 背景:本文以類神經網路(Artificial Neural Network, ANN)資料探勘(Datamining)演算法分析2012年1-5月某醫學中心外科加護病房轉入病人(n=588)之基本人口學資料、入住加護病房日數、轉入加護病房原因及住院24小時、48小時及轉出(或出院當日)之疾病嚴重度等7個變項建構預測模式,以探討影響死亡預測因子之相關數值。方法:資料來源為2012年1-5月轉入外科加護病房之病人共588人,其中死亡人數共54人,死亡率為9.1%。以STATISTICA軟體隨機選取441位病人作為訓練組,佔全體病人的75%;147位病人作為測試組,佔全體病人的25%。利用多層次類神經網路(MLP)方式建立最佳化的預測模型為MLP 59-19-2,輸出變數:病人出院時死亡、病危自動出院為1,存活為2。結果:疾病嚴重度分數高確實有較高之死亡機會,死亡預測準確率(Accuracy)達96.9%,訓練組ROC曲線下面積為0.975,靈敏度(Sensitivity)為78.8%,特異性(Specificity)為98.7%。結論:本研究結果證實疾病嚴重度確實與較高之死亡相關,其準確率、靈敏度與特異性皆非常高。此外,本研究結果顯示,非顱內出血相關疾病之病人,較少發生非預期之死亡;而顱內出血之病人,則醫師較能提前預知疾病預後不佳,並進而對家屬推薦重症安寧照顧理念。

KW - 加護病房

KW - 預測

KW - 死亡

KW - 類神經網路

KW - 資料探勘

KW - Intensive care unit

KW - Prediction

KW - Mortality

KW - Artificial neural network

KW - Datamining

M3 - 文章

VL - 14

SP - 1

EP - 7

JO - 重症醫學雜誌

JF - 重症醫學雜誌

SN - 1563-356x

IS - 1

ER -