Title :
Automatic vessel tree structure extraction by growing minimal paths and a mask
Author :
Da Chen ; Cohen, Laurent D.
Author_Institution :
CEREMADE, Univ. Paris Dauphine, Paris, France
fDate :
April 29 2014-May 2 2014
Abstract :
In this paper, we propose a completely automatic method to extract the vessel tree structure including its centerlines and radius using geodesic paths technology. Our main goal is to find a set of key points located in the vessel centerlines, link each pair of key points by minimal paths for finding the vessels between them and stop this process automatically. This work adapts the growing minimal paths method to find the set of key points. The main drawback of growing minimal paths is when to stop the processing. To solve this problem we propose an automatic stopping criteria. Additionally, we use a cake wavelet to compute the vessel measurement consisting of both radius and orientation to develop the classical growing minimal paths model which cannot guarantee the extracted tree corresponds to the centerlines of the vessel tree.
Keywords :
blood vessels; differential geometry; feature extraction; masks; medical image processing; wavelet transforms; automatic stopping criteria; automatic vessel tree structure extraction; cake wavelet; geodesic paths technology; mask; Computational modeling; Equations; Frequency modulation; Joining processes; Minimization; Noise; Three-dimensional displays; Vessel extraction; growing minimal paths; isotropic fast marching; key points;
Conference_Titel :
Biomedical Imaging (ISBI), 2014 IEEE 11th International Symposium on
Conference_Location :
Beijing
DOI :
10.1109/ISBI.2014.6867992