Cerebral arteries extraction using level set segmentation and adaptive tracing for CT angiography

Yong Zhang, Geoff Young, Xiaobo Zhou, Ranga Srinivasan, Stephen T.C. Wong

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

1 Scopus citations

Abstract

We propose an approach for extracting cerebral arteries from partial Computed Tomography Angiography (CTA). The challenges of extracting cerebral arteries from CTA come from the fact that arteries are usually surrounded by bones and veins in the lower portion of a CTA volume. There exists strong intensity-value overlap between vessels and surrounding objects. Besides, it is inappropriate to assume the 2D cross sections of arteries are circle or ellipse, especially for abnormal vessels. The navigation of the arteries could change suddenly in the 3D space. In this paper, a method based on level set segmentation is proposed to target this challenging problem. For the lower portion of a CTA volume, we use geodesic active contour method to detect cross section of arteries in the 2D space. The medial axis of the artery is obtained by adaptively tracking along its navigation path. This is done by finding the minimal cross section from cutting the arteries under different angles in the 3D spherical space. This method is highly automated, with minimum user input of providing only the starting point and initial navigation direction of the arteries of interests.

Original languageEnglish (US)
Title of host publicationComputational Models For Life Sciences (CMLS '07) - 2007 International Symposium
Pages57-65
Number of pages9
DOIs
StatePublished - 2007
Event2007 International Symposium on Computational Models for Life Sciences, CMLS '07 - Gold Coast, QLD, Australia
Duration: Dec 17 2007Dec 19 2007

Publication series

NameAIP Conference Proceedings
Volume952
ISSN (Print)0094-243X
ISSN (Electronic)1551-7616

Other

Other2007 International Symposium on Computational Models for Life Sciences, CMLS '07
Country/TerritoryAustralia
CityGold Coast, QLD
Period12/17/0712/19/07

Keywords

  • Computed tomography angiography (CTA)
  • Geodesic active contour
  • Vessel tracking

ASJC Scopus subject areas

  • General Physics and Astronomy

Fingerprint

Dive into the research topics of 'Cerebral arteries extraction using level set segmentation and adaptive tracing for CT angiography'. Together they form a unique fingerprint.

Cite this