Title :
Detection of elliptic shells using fuzzy clustering: application to MRI images
Author :
Gath, Isak ; Hoory, Dan
Author_Institution :
Dept. of Biomed. Eng., Technion-Israel Inst. of Technol., Haifa, Israel
Abstract :
The study describes a new fuzzy clustering algorithm, aimed at detection of disjoint and/or superimposed elliptic ring-shaped clusters, as well as combinations of compact clusters and circular and elliptic rings. It is based on minimization of an objective function in which the parameters of the ellipses are included in the optimization scheme. The fuzzy k-ellipses (FKE) algorithm has been tested on a large number of synthetic examples and was found to be efficient in detection of multiple disjoint elliptic ring-shaped clusters, superimposed ellipses, and combinations of both. The FKE algorithm has been applied to 2-D MRI images of cross sections of the heart´s left ventricle. The inner and outer contours of the left ventricle have been characterized as nearly concentric elliptic rings. From multiple such sections, taken at different points in time during the cardiac cycle (systole-diastole) a dynamic 3-D image of the left ventricle has been reconstructed
Keywords :
biomedical NMR; 2D MRI images; MRI images; cardiac cycle; cardiology; cross sections; elliptic shell detection; fuzzy clustering; fuzzy k-ellipses algorithm; heart; left ventricle; multiple disjoint elliptic ring-shaped clusters; objective function minimization; superimposed elliptic ring-shaped clusters; Biomedical engineering; Clustering algorithms; Image reconstruction; Iterative algorithms; Magnetic resonance imaging; Multidimensional systems; Nonlinear equations; Partitioning algorithms; Prototypes; Testing;
Conference_Titel :
Pattern Recognition, 1994. Vol. 2 - Conference B: Computer Vision & Image Processing., Proceedings of the 12th IAPR International. Conference on
Conference_Location :
Jerusalem
Print_ISBN :
0-8186-6270-0
DOI :
10.1109/ICPR.1994.576914
