LV SEGMENTATION FROM 3D ECHOCARDIOGRAPHY USING FUZZY FEATURES AND A MULTILEVEL FFD MODEL

Author

Yan, Ping ; Sinusas, Albert ; Duncan, James S.

Author_Institution

Yale Univ., New Haven, CT

fYear

2007

fDate

12-15 April 2007

Firstpage

848

Lastpage

851

Abstract

In this paper, we propose a method for 3D left ventricle (LV) segmentation from echocardiography. This method defines an ´active contour´ on a multilevel cubic B-spline formulated grid. Moving the active contour to the boundary is performed by minimizing an energy function, which is defined by fuzzy feature information and multilevel free form deformation model (MFFD). Several echocardiographic LV segmentation problems are considered and experimental results are given to illustrate the robustness of this algorithm against spurious local minima and leaking on the blurred boundary

Keywords

biomechanics; deformation; echocardiography; fuzzy set theory; image segmentation; medical image processing; physiological models; splines (mathematics); surface topography; active contour; blurred boundary leaking; cubic B-spline; energy function; fuzzy features; left ventricle segmentation; multilevel FFD model; multilevel formulated grid; multilevel free form deformation model; spurious local minima; three-dimensional echocardiography; Active contours; Cardiac disease; Echocardiography; Filtering; Humans; Image segmentation; Robustness; Shape; Speckle; Spline;

fLanguage

English

Publisher

ieee

Conference_Titel

Biomedical Imaging: From Nano to Macro, 2007. ISBI 2007. 4th IEEE International Symposium on

Conference_Location

Arlington, VA

Print_ISBN

1-4244-0672-2

Electronic_ISBN

1-4244-0672-2

Type

conf

DOI

10.1109/ISBI.2007.356985

Filename

4193419