Share Email Print

Proceedings Paper

Object-tracking segmentation method: vertebra and rib segmentation in CT images
Author(s): Dongsung Kim; Hanyoung Kim; Heung Sik Kang
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

This paper proposes a method that segments an object in 3D image slices by tracking not only the target objects to segment but also adjacent non-target objects. The tracking methodology enables the proposed method to detect and prevent a leakage, which is one of the most challenging problems in medical image segmentation. The method also provides an efficient way of modification of segmentation results by informing only suspicious image slices because of a leakage. The proposed method is applied to segment vertebrae in angio image slices in which blood vessels are enhanced with contrast media, and produced promising results in terms of both accuracy and efficiency.

Paper Details

Date Published: 9 May 2002
PDF: 10 pages
Proc. SPIE 4684, Medical Imaging 2002: Image Processing, (9 May 2002); doi: 10.1117/12.467136
Show Author Affiliations
Dongsung Kim, Soongsil Univ. (South Korea)
Hanyoung Kim, Soongsil Univ. (South Korea)
Heung Sik Kang, Seoul National Univ. College of Medicine (South Korea)

Published in SPIE Proceedings Vol. 4684:
Medical Imaging 2002: Image Processing
Milan Sonka; J. Michael Fitzpatrick, Editor(s)

© SPIE. Terms of Use
Back to Top