TY - JOUR
T1 - Segmentation of liver vasculature from contrast enhanced CT images using context-based voting
AU - Chi, Yanling
AU - Liu, Jimin
AU - Venkatesh, Sudhakar K.
AU - Huang, Su
AU - Zhou, Jiayin
AU - Tian, Qi
AU - Nowinski, Wieslaw L.
N1 - Funding Information:
Manuscript received June 15, 2010; revised August 29, 2010 and November 2, 2010; accepted November 3, 2010. Date of publication November 18, 2010; date of current version July 20, 2011. This work was supported by the Agency for Science, Technology and Research, Singapore, under Grant JCOAG03_FG05_2009. Asterisk indicates corresponding author. *Y. Chi is with the Biomedical Imaging Laboratory, A∗STAR, Singapore 138671 (e-mail: chi_yanling@sbic.a-star.edu.sg).
PY - 2011/8
Y1 - 2011/8
N2 - A novel vessel context-based voting is proposed for automatic liver vasculature segmentation in CT images. It is able to conduct full vessel segmentation and recognition of multiple vasculatures effectively. The vessel context describes context information of a voxel related to vessel properties, such as intensity, saliency, direction, and connectivity. Voxels are grouped to liver vasculatures hierarchically based on vessel context. They are first grouped locally into vessel branches with the advantage of a vessel junction measurement and then grouped globally into vasculatures, which is implemented using a multiple feature point voting mechanism. The proposed method has been evaluated on ten clinical CT datasets. Segmentation of third-order vessel trees from CT images (0.76×0.76 2.0 mm) of the portal venous phase takes less than 3 min on a PC with 2.0 GHz dual core processor and the average segmentation accuracy is up to 98.
AB - A novel vessel context-based voting is proposed for automatic liver vasculature segmentation in CT images. It is able to conduct full vessel segmentation and recognition of multiple vasculatures effectively. The vessel context describes context information of a voxel related to vessel properties, such as intensity, saliency, direction, and connectivity. Voxels are grouped to liver vasculatures hierarchically based on vessel context. They are first grouped locally into vessel branches with the advantage of a vessel junction measurement and then grouped globally into vasculatures, which is implemented using a multiple feature point voting mechanism. The proposed method has been evaluated on ten clinical CT datasets. Segmentation of third-order vessel trees from CT images (0.76×0.76 2.0 mm) of the portal venous phase takes less than 3 min on a PC with 2.0 GHz dual core processor and the average segmentation accuracy is up to 98.
KW - Liver vasculature segmentation
KW - multiple feature point voting
KW - vessel context
KW - vessel junction measure
UR - http://www.scopus.com/inward/record.url?scp=79960733564&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=79960733564&partnerID=8YFLogxK
U2 - 10.1109/TBME.2010.2093523
DO - 10.1109/TBME.2010.2093523
M3 - Article
C2 - 21095856
AN - SCOPUS:79960733564
VL - 58
SP - 2144
EP - 2153
JO - IRE transactions on medical electronics
JF - IRE transactions on medical electronics
SN - 0018-9294
IS - 8
M1 - 5639035
ER -