Automatic tracking of local myocardial motion by correlation weighted velocity method

Wataru Ohyama, Tetsushi Wakabayashi, Fumitaka Kimura, Shinji Tsuruoka, Kiyotsugu Sekioka

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

In this paper, we propose a new method for automatically tracking the motion of local region in left ventricular myocardium by means of ultrasonic pulsed Doppler signal. This method consists of a velocity detection procedure based on correlation weighted mean instantaneous velocity and a motion tracking procedure employing a myocardial elastic model. Most of ultrasonic pulsed Doppler signals observed in clinical diagnosis contain considerable amount of speckle noise, which causes detection error of velocity. The detection error is accumulated in the motion tracking procedure and yields obviously incorrect motion trajectory. The procedure of correlation weighted mean velocity is aimed to reduce the velocity detection error, and the myocardial elastic model is used to avoid the accumulation of the error to keep track of the motion of the myocardium in reasonable accuracy. The result of evaluation test shows that this method is able to improve the accuracy of tracking approximately 40 % relative to a conventional CLS approach.

Original languageEnglish
Title of host publicationProceedings - 16th International Conference on Pattern Recognition, ICPR 2002
EditorsR. Kasturi, R. Woodham, Y. Shirai, D. Laurendeau, H. Bunke, G. Sanniti di Baja, N. Ayache, K. Boyer, M. Kunt, H. Christensen, L. Shapiro, C. Suen
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages711-714
Number of pages4
ISBN (Electronic)0769516963
DOIs
Publication statusPublished - Jan 1 2002
Event16th International Conference on Pattern Recognition, ICPR 2002 - Quebec City, Canada
Duration: Aug 11 2002Aug 15 2002

Publication series

NameProceedings - International Conference on Pattern Recognition
Volume1
ISSN (Print)1051-4651

Other

Other16th International Conference on Pattern Recognition, ICPR 2002
Country/TerritoryCanada
CityQuebec City
Period8/11/028/15/02

All Science Journal Classification (ASJC) codes

  • Computer Vision and Pattern Recognition

Fingerprint

Dive into the research topics of 'Automatic tracking of local myocardial motion by correlation weighted velocity method'. Together they form a unique fingerprint.

Cite this