Authors:
Antonio Bravo
1
;
José Clemente
1
;
Miguel Vera
2
;
José Avila
2
and
Rubén Medina
2
Affiliations:
1
Universidad Nacional Experimental del Táchira, Venezuela
;
2
Universidad de Los Andes, Venezuela
Keyword(s):
Segmentation, Generalized Hough transform, Mathematical morphology, Unsupervised clustering, Cardiac images, Left ventricle.
Related
Ontology
Subjects/Areas/Topics:
Computer Vision, Visualization and Computer Graphics
;
Image and Video Analysis
;
Segmentation and Grouping
Abstract:
An automatic approach based on the generalized Hough transform (GHT) and unsupervised clustering technique to obtain the endocardial surface is proposed. The approach is applied to multi slice computerized tomography (MSCT) images of the heart. The first step is the initialization, where a GHT–based segmentation algorithm is used to detect the edocardial contour in one MSCT slice. The centroid of this contour is used as a seed point for initializing a clustering algorithm. A two stage segmentation algorithm is used for segmenting the three–dimensional MSCT database. First, the complete database is filtered using mathematical morphology operators in order to improve the left ventricle cavity information in these images. The second stage is based on a region growing method. A seed point located inside the cardiac cavity is used as input for the clustering algorithm. This seed point is propagated along the image sequence to obtain the left ventricle surfaces for all instants of the card
iac cycle. The method is validated by comparing the estimated surfaces with respect to left ventricle shapes drawn by a cardiologist. The average error obtained was 1.52 mm.
(More)