Divergence-based segmental profile of myocardial motion for the detection of cardiac abnormality

Slamet Riyadi, Mohd. Marzuki Mustafa, Aini Hussain, Oteh Maskon, I. F M Nor

Research output: Contribution to journalArticle

Abstract

The quantification of myocardial motion is an important aid to cardiologists in diagnosing cardiac abnormalities. Currently, displacement and vector angle are the main motion features extracted from the computed flow vectors. These features are sensitive to the accuracy of flow vector computation. In this paper, we propose a new features extraction technique based on a gradient approach to provide a segmental profile of the myocardial boundary motion to localize cardiac abnormalities. Once the flow vectors along the endocardial boundary are computed, we apply a divergence operator to every frame to obtain a segmental divergence profile from end-diastole to end-systole. This profile provides the direction of vectors with respect to the cavity center and the ratio of displacement on one graph for every segment; hence, it can be used to provide preliminary detection of an abnormal segment. For validation, the proposed technique was compared to manual scoring of the cardiac motion performed by a cardiologist. This technique has been applied in a number of healthy and abnormal cardiac studies and has been shown to be successful in providing a simpler, more concise and robust cardiac profile than those based on the displacement and angle.

Original languageEnglish
Pages (from-to)5167-5176
Number of pages10
JournalJournal of Computational Information Systems
Volume7
Issue number14
Publication statusPublished - Dec 2011

Fingerprint

Feature extraction

Keywords

  • Cardiac abnormality
  • Divergence
  • Myocardial motion
  • Optical flow
  • Regional profile

ASJC Scopus subject areas

  • Computer Science Applications
  • Information Systems

Cite this

@article{7da5fd84facc4d039bdf33578d12272e,
title = "Divergence-based segmental profile of myocardial motion for the detection of cardiac abnormality",
abstract = "The quantification of myocardial motion is an important aid to cardiologists in diagnosing cardiac abnormalities. Currently, displacement and vector angle are the main motion features extracted from the computed flow vectors. These features are sensitive to the accuracy of flow vector computation. In this paper, we propose a new features extraction technique based on a gradient approach to provide a segmental profile of the myocardial boundary motion to localize cardiac abnormalities. Once the flow vectors along the endocardial boundary are computed, we apply a divergence operator to every frame to obtain a segmental divergence profile from end-diastole to end-systole. This profile provides the direction of vectors with respect to the cavity center and the ratio of displacement on one graph for every segment; hence, it can be used to provide preliminary detection of an abnormal segment. For validation, the proposed technique was compared to manual scoring of the cardiac motion performed by a cardiologist. This technique has been applied in a number of healthy and abnormal cardiac studies and has been shown to be successful in providing a simpler, more concise and robust cardiac profile than those based on the displacement and angle.",
keywords = "Cardiac abnormality, Divergence, Myocardial motion, Optical flow, Regional profile",
author = "Slamet Riyadi and Mustafa, {Mohd. Marzuki} and Aini Hussain and Oteh Maskon and Nor, {I. F M}",
year = "2011",
month = "12",
language = "English",
volume = "7",
pages = "5167--5176",
journal = "Journal of Computational Information Systems",
issn = "1553-9105",
publisher = "Binary Information Press",
number = "14",

}

TY - JOUR

T1 - Divergence-based segmental profile of myocardial motion for the detection of cardiac abnormality

AU - Riyadi, Slamet

AU - Mustafa, Mohd. Marzuki

AU - Hussain, Aini

AU - Maskon, Oteh

AU - Nor, I. F M

PY - 2011/12

Y1 - 2011/12

N2 - The quantification of myocardial motion is an important aid to cardiologists in diagnosing cardiac abnormalities. Currently, displacement and vector angle are the main motion features extracted from the computed flow vectors. These features are sensitive to the accuracy of flow vector computation. In this paper, we propose a new features extraction technique based on a gradient approach to provide a segmental profile of the myocardial boundary motion to localize cardiac abnormalities. Once the flow vectors along the endocardial boundary are computed, we apply a divergence operator to every frame to obtain a segmental divergence profile from end-diastole to end-systole. This profile provides the direction of vectors with respect to the cavity center and the ratio of displacement on one graph for every segment; hence, it can be used to provide preliminary detection of an abnormal segment. For validation, the proposed technique was compared to manual scoring of the cardiac motion performed by a cardiologist. This technique has been applied in a number of healthy and abnormal cardiac studies and has been shown to be successful in providing a simpler, more concise and robust cardiac profile than those based on the displacement and angle.

AB - The quantification of myocardial motion is an important aid to cardiologists in diagnosing cardiac abnormalities. Currently, displacement and vector angle are the main motion features extracted from the computed flow vectors. These features are sensitive to the accuracy of flow vector computation. In this paper, we propose a new features extraction technique based on a gradient approach to provide a segmental profile of the myocardial boundary motion to localize cardiac abnormalities. Once the flow vectors along the endocardial boundary are computed, we apply a divergence operator to every frame to obtain a segmental divergence profile from end-diastole to end-systole. This profile provides the direction of vectors with respect to the cavity center and the ratio of displacement on one graph for every segment; hence, it can be used to provide preliminary detection of an abnormal segment. For validation, the proposed technique was compared to manual scoring of the cardiac motion performed by a cardiologist. This technique has been applied in a number of healthy and abnormal cardiac studies and has been shown to be successful in providing a simpler, more concise and robust cardiac profile than those based on the displacement and angle.

KW - Cardiac abnormality

KW - Divergence

KW - Myocardial motion

KW - Optical flow

KW - Regional profile

UR - http://www.scopus.com/inward/record.url?scp=83255192368&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=83255192368&partnerID=8YFLogxK

M3 - Article

VL - 7

SP - 5167

EP - 5176

JO - Journal of Computational Information Systems

JF - Journal of Computational Information Systems

SN - 1553-9105

IS - 14

ER -