Classification of acute myocardial ischemia by artificial neural network using echocardiographic strain waveforms

Eileen M. McMahon, Josef Korinek, Shiro Yoshifuku, Partho P. Sengupta, Armando Manduca, Marek Belohlavek

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

Echocardiographic strain waveforms are highly variable, so their interpretation is experience-dependent and subjective. We tested whether an artificial neural network (ANN) can distinguish between strain waveforms obtained at baseline and during experimentally induced acute ischemia. An open-chest model of coronary occlusion and acute ischemia was used in 14 adult pigs. Strain waveforms were obtained using a GE Vivid 7 ultrasound system. An ANN design was implemented in MATLAB®, and backpropagation and "leave-one-out" processes were used to train and test it. Specificity of 86% and sensitivity of 87% suggest that ANNs could aid in diagnostic prescreening of echocardiographic strain waveforms.

Original languageEnglish (US)
Pages (from-to)416-424
Number of pages9
JournalComputers in Biology and Medicine
Volume38
Issue number4
DOIs
StatePublished - Apr 2008

Fingerprint

Myocardial Ischemia
Ischemia
Neural networks
Coronary Occlusion
Swine
Thorax
Sensitivity and Specificity
Backpropagation
MATLAB
Ultrasonics

Keywords

  • Acute myocardial ischemia
  • Artificial neural network
  • Strain echocardiography

ASJC Scopus subject areas

  • Computer Science Applications

Cite this

Classification of acute myocardial ischemia by artificial neural network using echocardiographic strain waveforms. / McMahon, Eileen M.; Korinek, Josef; Yoshifuku, Shiro; Sengupta, Partho P.; Manduca, Armando; Belohlavek, Marek.

In: Computers in Biology and Medicine, Vol. 38, No. 4, 04.2008, p. 416-424.

Research output: Contribution to journalArticle

@article{15b9da97d0ac4c50bcec68cb594258c6,
title = "Classification of acute myocardial ischemia by artificial neural network using echocardiographic strain waveforms",
abstract = "Echocardiographic strain waveforms are highly variable, so their interpretation is experience-dependent and subjective. We tested whether an artificial neural network (ANN) can distinguish between strain waveforms obtained at baseline and during experimentally induced acute ischemia. An open-chest model of coronary occlusion and acute ischemia was used in 14 adult pigs. Strain waveforms were obtained using a GE Vivid 7 ultrasound system. An ANN design was implemented in MATLAB{\circledR}, and backpropagation and {"}leave-one-out{"} processes were used to train and test it. Specificity of 86{\%} and sensitivity of 87{\%} suggest that ANNs could aid in diagnostic prescreening of echocardiographic strain waveforms.",
keywords = "Acute myocardial ischemia, Artificial neural network, Strain echocardiography",
author = "McMahon, {Eileen M.} and Josef Korinek and Shiro Yoshifuku and Sengupta, {Partho P.} and Armando Manduca and Marek Belohlavek",
year = "2008",
month = "4",
doi = "10.1016/j.compbiomed.2008.01.003",
language = "English (US)",
volume = "38",
pages = "416--424",
journal = "Computers in Biology and Medicine",
issn = "0010-4825",
publisher = "Elsevier Limited",
number = "4",

}

TY - JOUR

T1 - Classification of acute myocardial ischemia by artificial neural network using echocardiographic strain waveforms

AU - McMahon, Eileen M.

AU - Korinek, Josef

AU - Yoshifuku, Shiro

AU - Sengupta, Partho P.

AU - Manduca, Armando

AU - Belohlavek, Marek

PY - 2008/4

Y1 - 2008/4

N2 - Echocardiographic strain waveforms are highly variable, so their interpretation is experience-dependent and subjective. We tested whether an artificial neural network (ANN) can distinguish between strain waveforms obtained at baseline and during experimentally induced acute ischemia. An open-chest model of coronary occlusion and acute ischemia was used in 14 adult pigs. Strain waveforms were obtained using a GE Vivid 7 ultrasound system. An ANN design was implemented in MATLAB®, and backpropagation and "leave-one-out" processes were used to train and test it. Specificity of 86% and sensitivity of 87% suggest that ANNs could aid in diagnostic prescreening of echocardiographic strain waveforms.

AB - Echocardiographic strain waveforms are highly variable, so their interpretation is experience-dependent and subjective. We tested whether an artificial neural network (ANN) can distinguish between strain waveforms obtained at baseline and during experimentally induced acute ischemia. An open-chest model of coronary occlusion and acute ischemia was used in 14 adult pigs. Strain waveforms were obtained using a GE Vivid 7 ultrasound system. An ANN design was implemented in MATLAB®, and backpropagation and "leave-one-out" processes were used to train and test it. Specificity of 86% and sensitivity of 87% suggest that ANNs could aid in diagnostic prescreening of echocardiographic strain waveforms.

KW - Acute myocardial ischemia

KW - Artificial neural network

KW - Strain echocardiography

UR - http://www.scopus.com/inward/record.url?scp=40949095558&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=40949095558&partnerID=8YFLogxK

U2 - 10.1016/j.compbiomed.2008.01.003

DO - 10.1016/j.compbiomed.2008.01.003

M3 - Article

VL - 38

SP - 416

EP - 424

JO - Computers in Biology and Medicine

JF - Computers in Biology and Medicine

SN - 0010-4825

IS - 4

ER -