• DocumentCode
    1956632
  • Title

    Quantitative evaluation of correlation-based 3D vs. 2D speckle tracking using finite element cardiac mechanical model and in-vitro phantom

  • Author

    Tripathy, S. ; Simon, M.A. ; Kim, K.

  • Author_Institution
    Center for Ultrasound Mol. Imaging & Therapeutics, Univ. of Pittsburgh, Pittsburgh, PA, USA
  • fYear
    2010
  • fDate
    11-14 Oct. 2010
  • Firstpage
    579
  • Lastpage
    582
  • Abstract
    Two-dimensional (2D) speckle tracking (ST) has been used extensively over the years in many different ultrasound (US) elasticity imaging applications with some limitations. 2D ST is limited especially for cardiac applications because of the 3D and complex motion of the heart. Out-of-plane motion is a major source of decorrelation for 2D ST. Recently 3D ST based on full 3D kernel search has been developed to overcome these limitations. In this study, the improvement of 3D ST performance compared to 2D ST is quantitatively analyzed via both a finite element (FE) based cardiac mechanical model and water tank experiments using a tissue phantom. Finite element based analysis suggested an improvement of about two folds in the errors. Considerable improvement in the quality of the displacement estimate images was observed in the experiments.
  • Keywords
    biomechanics; biomedical ultrasonics; cardiology; finite element analysis; phantoms; speckle; 3D kernel search; correlation-based 2D speckle tracking; correlation-based 3D speckle tracking; decorrelation; finite element cardiac mechanical model; phantom; ultrasound elasticity imaging; Elasticity; Phantoms; Speckle; Strain; Three dimensional displays; Ultrasonic imaging; 3D speckle tracking; Ultrasound elasticity imaging; echochardiography;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Ultrasonics Symposium (IUS), 2010 IEEE
  • Conference_Location
    San Diego, CA
  • ISSN
    1948-5719
  • Print_ISBN
    978-1-4577-0382-9
  • Type

    conf

  • DOI
    10.1109/ULTSYM.2010.5935680
  • Filename
    5935680