Intelligent Data Analysis for the Classification of Body Surface Potential Maps

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Body surface potential maps were investigated to identify a set of optimal recording sites required to discriminate between several diseases. Specifically, recordings captured from subjects exhibiting myocardial infarction or left ventricular hypertrophy, as well as a control group consisting of healthy subjects were investigated. Due to the fact that multi-class problems are inherently difficult to solve we divided the problem into several two-class scenarios. Six datasets were generated from the available 744 records, each viewing the available data differently, to form several two-class problems. A data-driven selection algorithm was applied to each of the generated datasets to produce six classification models, each utilising as features those recording sites offering most to the discrimination task being investigated. Subsequently, a framework was introduced to facilitate the combination of outputs from each classifier. Essentially, the framework used the outputs from half of the classification models to determine which of the remaining models would be employed to form a final decision. A benchmark, in the form of a multi-group classifier, was introduced to evaluate the perceived benefits of the proposed approach. An improvement of approximately 10% upon the benchmark was observed resulting in an overall accuracy of 79.19%.
LanguageEnglish
Pages249-263
JournalInternational Journal of Computational Intelligence and Applications
Volume7
Issue number3
DOIs
Publication statusPublished - 5 Dec 2008

Fingerprint

Surface Potential
Surface potential
Data analysis
Classifiers
Classifier
Benchmark
Hypertrophy
Myocardial Infarction
Output
Multi-class
Data-driven
Discrimination
Model
Scenarios
Evaluate
Class
Framework

Keywords

  • Body surface potential mapping
  • feature subset selection
  • wrapper
  • limited lead sets
  • diagnostic ECG

Cite this

@article{c94d58719a094260af25224d0104500d,
title = "Intelligent Data Analysis for the Classification of Body Surface Potential Maps",
abstract = "Body surface potential maps were investigated to identify a set of optimal recording sites required to discriminate between several diseases. Specifically, recordings captured from subjects exhibiting myocardial infarction or left ventricular hypertrophy, as well as a control group consisting of healthy subjects were investigated. Due to the fact that multi-class problems are inherently difficult to solve we divided the problem into several two-class scenarios. Six datasets were generated from the available 744 records, each viewing the available data differently, to form several two-class problems. A data-driven selection algorithm was applied to each of the generated datasets to produce six classification models, each utilising as features those recording sites offering most to the discrimination task being investigated. Subsequently, a framework was introduced to facilitate the combination of outputs from each classifier. Essentially, the framework used the outputs from half of the classification models to determine which of the remaining models would be employed to form a final decision. A benchmark, in the form of a multi-group classifier, was introduced to evaluate the perceived benefits of the proposed approach. An improvement of approximately 10{\%} upon the benchmark was observed resulting in an overall accuracy of 79.19{\%}.",
keywords = "Body surface potential mapping, feature subset selection, wrapper, limited lead sets, diagnostic ECG",
author = "Mark Donnelly and Christopher Nugent and Dewar Finlay and Norman Black",
year = "2008",
month = "12",
day = "5",
doi = "10.1142/S1469026808002284",
language = "English",
volume = "7",
pages = "249--263",
journal = "International Journal of Computational Intelligence and Applications",
issn = "1469-0268",
number = "3",

}

TY - JOUR

T1 - Intelligent Data Analysis for the Classification of Body Surface Potential Maps

AU - Donnelly, Mark

AU - Nugent, Christopher

AU - Finlay, Dewar

AU - Black, Norman

PY - 2008/12/5

Y1 - 2008/12/5

N2 - Body surface potential maps were investigated to identify a set of optimal recording sites required to discriminate between several diseases. Specifically, recordings captured from subjects exhibiting myocardial infarction or left ventricular hypertrophy, as well as a control group consisting of healthy subjects were investigated. Due to the fact that multi-class problems are inherently difficult to solve we divided the problem into several two-class scenarios. Six datasets were generated from the available 744 records, each viewing the available data differently, to form several two-class problems. A data-driven selection algorithm was applied to each of the generated datasets to produce six classification models, each utilising as features those recording sites offering most to the discrimination task being investigated. Subsequently, a framework was introduced to facilitate the combination of outputs from each classifier. Essentially, the framework used the outputs from half of the classification models to determine which of the remaining models would be employed to form a final decision. A benchmark, in the form of a multi-group classifier, was introduced to evaluate the perceived benefits of the proposed approach. An improvement of approximately 10% upon the benchmark was observed resulting in an overall accuracy of 79.19%.

AB - Body surface potential maps were investigated to identify a set of optimal recording sites required to discriminate between several diseases. Specifically, recordings captured from subjects exhibiting myocardial infarction or left ventricular hypertrophy, as well as a control group consisting of healthy subjects were investigated. Due to the fact that multi-class problems are inherently difficult to solve we divided the problem into several two-class scenarios. Six datasets were generated from the available 744 records, each viewing the available data differently, to form several two-class problems. A data-driven selection algorithm was applied to each of the generated datasets to produce six classification models, each utilising as features those recording sites offering most to the discrimination task being investigated. Subsequently, a framework was introduced to facilitate the combination of outputs from each classifier. Essentially, the framework used the outputs from half of the classification models to determine which of the remaining models would be employed to form a final decision. A benchmark, in the form of a multi-group classifier, was introduced to evaluate the perceived benefits of the proposed approach. An improvement of approximately 10% upon the benchmark was observed resulting in an overall accuracy of 79.19%.

KW - Body surface potential mapping

KW - feature subset selection

KW - wrapper

KW - limited lead sets

KW - diagnostic ECG

U2 - 10.1142/S1469026808002284

DO - 10.1142/S1469026808002284

M3 - Article

VL - 7

SP - 249

EP - 263

JO - International Journal of Computational Intelligence and Applications

T2 - International Journal of Computational Intelligence and Applications

JF - International Journal of Computational Intelligence and Applications

SN - 1469-0268

IS - 3

ER -