HeartCast: Predicting acute hypotensive episodes in intensive care units

Sun Hee Kim, Lei Li, Christos Faloutsos, Hyung Jeong Yang, Seong Whan Lee

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Acute hypotensive episodes (AHEs) are serious clinical events in intensive care units (ICUs), and require immediate treatment to prevent patient injury. Reducing the risks associated with an AHE requires effective and efficient mining of data generated from multiple physiological time series. We propose HeartCast, a model that extracts essential features from such data to effectively predict AHE. HeartCast combines a non-linear support vector machine with best-feature extraction via analysis of the baseline threshold, quartile parameters, and window size of the physiological signals. Our approach has the following benefits: (a) it extracts the most relevant features; (b) it provides the best results for identification of an AHE event; (c) it is fast and scales with linear complexity over the length of the window; and (d) it can manage missing values and noise/outliers by using a best-feature extraction method. We performed experiments on data continuously captured from physiological time series of ICU patients (roughly 3 GB of processed data). HeartCast was found to outperform other state-of-the-art methods found in the literature with a 13.7% improvement in classification accuracy.

Original languageEnglish
Pages (from-to)1-13
Number of pages13
JournalStatistical Methodology
Volume33
DOIs
Publication statusPublished - 2016 Dec 1

Fingerprint

Acute
Unit
Feature Extraction
Time series
Quartile
Linear Complexity
Missing Values
Outlier
Mining
Baseline
Support Vector Machine
Predict
Experiment
Model

Keywords

  • Acute hypotensive episodes
  • Feature selection
  • Physiological signal analysis
  • Prediction
  • Quartile parameters

ASJC Scopus subject areas

  • Statistics and Probability

Cite this

HeartCast : Predicting acute hypotensive episodes in intensive care units. / Kim, Sun Hee; Li, Lei; Faloutsos, Christos; Yang, Hyung Jeong; Lee, Seong Whan.

In: Statistical Methodology, Vol. 33, 01.12.2016, p. 1-13.

Research output: Contribution to journalArticle

Kim, Sun Hee ; Li, Lei ; Faloutsos, Christos ; Yang, Hyung Jeong ; Lee, Seong Whan. / HeartCast : Predicting acute hypotensive episodes in intensive care units. In: Statistical Methodology. 2016 ; Vol. 33. pp. 1-13.
@article{ebf276f1e47b4cdaa3a6db5af727f69f,
title = "HeartCast: Predicting acute hypotensive episodes in intensive care units",
abstract = "Acute hypotensive episodes (AHEs) are serious clinical events in intensive care units (ICUs), and require immediate treatment to prevent patient injury. Reducing the risks associated with an AHE requires effective and efficient mining of data generated from multiple physiological time series. We propose HeartCast, a model that extracts essential features from such data to effectively predict AHE. HeartCast combines a non-linear support vector machine with best-feature extraction via analysis of the baseline threshold, quartile parameters, and window size of the physiological signals. Our approach has the following benefits: (a) it extracts the most relevant features; (b) it provides the best results for identification of an AHE event; (c) it is fast and scales with linear complexity over the length of the window; and (d) it can manage missing values and noise/outliers by using a best-feature extraction method. We performed experiments on data continuously captured from physiological time series of ICU patients (roughly 3 GB of processed data). HeartCast was found to outperform other state-of-the-art methods found in the literature with a 13.7{\%} improvement in classification accuracy.",
keywords = "Acute hypotensive episodes, Feature selection, Physiological signal analysis, Prediction, Quartile parameters",
author = "Kim, {Sun Hee} and Lei Li and Christos Faloutsos and Yang, {Hyung Jeong} and Lee, {Seong Whan}",
year = "2016",
month = "12",
day = "1",
doi = "10.1016/j.stamet.2016.07.001",
language = "English",
volume = "33",
pages = "1--13",
journal = "Statistical Methodology",
issn = "1572-3127",
publisher = "Elsevier",

}

TY - JOUR

T1 - HeartCast

T2 - Predicting acute hypotensive episodes in intensive care units

AU - Kim, Sun Hee

AU - Li, Lei

AU - Faloutsos, Christos

AU - Yang, Hyung Jeong

AU - Lee, Seong Whan

PY - 2016/12/1

Y1 - 2016/12/1

N2 - Acute hypotensive episodes (AHEs) are serious clinical events in intensive care units (ICUs), and require immediate treatment to prevent patient injury. Reducing the risks associated with an AHE requires effective and efficient mining of data generated from multiple physiological time series. We propose HeartCast, a model that extracts essential features from such data to effectively predict AHE. HeartCast combines a non-linear support vector machine with best-feature extraction via analysis of the baseline threshold, quartile parameters, and window size of the physiological signals. Our approach has the following benefits: (a) it extracts the most relevant features; (b) it provides the best results for identification of an AHE event; (c) it is fast and scales with linear complexity over the length of the window; and (d) it can manage missing values and noise/outliers by using a best-feature extraction method. We performed experiments on data continuously captured from physiological time series of ICU patients (roughly 3 GB of processed data). HeartCast was found to outperform other state-of-the-art methods found in the literature with a 13.7% improvement in classification accuracy.

AB - Acute hypotensive episodes (AHEs) are serious clinical events in intensive care units (ICUs), and require immediate treatment to prevent patient injury. Reducing the risks associated with an AHE requires effective and efficient mining of data generated from multiple physiological time series. We propose HeartCast, a model that extracts essential features from such data to effectively predict AHE. HeartCast combines a non-linear support vector machine with best-feature extraction via analysis of the baseline threshold, quartile parameters, and window size of the physiological signals. Our approach has the following benefits: (a) it extracts the most relevant features; (b) it provides the best results for identification of an AHE event; (c) it is fast and scales with linear complexity over the length of the window; and (d) it can manage missing values and noise/outliers by using a best-feature extraction method. We performed experiments on data continuously captured from physiological time series of ICU patients (roughly 3 GB of processed data). HeartCast was found to outperform other state-of-the-art methods found in the literature with a 13.7% improvement in classification accuracy.

KW - Acute hypotensive episodes

KW - Feature selection

KW - Physiological signal analysis

KW - Prediction

KW - Quartile parameters

UR - http://www.scopus.com/inward/record.url?scp=84979743504&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84979743504&partnerID=8YFLogxK

U2 - 10.1016/j.stamet.2016.07.001

DO - 10.1016/j.stamet.2016.07.001

M3 - Article

AN - SCOPUS:84979743504

VL - 33

SP - 1

EP - 13

JO - Statistical Methodology

JF - Statistical Methodology

SN - 1572-3127

ER -