Motion estimation along the myocardial boundary using boundary extraction and optical flow

Adhi Harmoko Saputro, Mohd Marzuki Mustafa, Aini Hussain, Oteh Maskon, Ika Faizura Mohd Noh

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

4 Citations (Scopus)

Abstract

Myocardial motion can be helpful in diagnosing the heart abnormalities because it relates to cardiac vascular supply. Analyzing the motion of all segments of the myocardial is a difficult task because of the random noise of echocardiographic images. Noise in the echocardiographic images that is produced by the image acquisition is not consistent in each frame. It causes error in motion computation especially in optical flow computation that relies on brightness constancy. To increase the accuracy of motion estimation along the myocardial boundary, we proposed a method that combines boundary detection and optical flow to compute myocardial motion at the myocardial boundary. In this method, computation of optical flow from 2 consecutive images is done after myocardial boundary is detected in each frame. The result shows that the motion estimation algorithm along the myocardial boundary yields better result compared to without using boundary extraction methods.

Original languageEnglish
Title of host publicationWCE 2010 - World Congress on Engineering 2010
Pages677-680
Number of pages4
Publication statusPublished - 2010
EventWorld Congress on Engineering 2010, WCE 2010 - London, United Kingdom
Duration: 30 Jun 20102 Jul 2010

Publication series

NameWCE 2010 - World Congress on Engineering 2010
Volume1

Conference

ConferenceWorld Congress on Engineering 2010, WCE 2010
Country/TerritoryUnited Kingdom
CityLondon
Period30/06/102/07/10

Keywords

  • Boundary estimation
  • Echocardiographic
  • Myocardial
  • Optical flow

Fingerprint

Dive into the research topics of 'Motion estimation along the myocardial boundary using boundary extraction and optical flow'. Together they form a unique fingerprint.

Cite this