Title :
Fast Edge-Preserving Filtering for 3D+t Echocardiographic Volume Rendering
Author :
Yanez-Suarez, Oscar ; Lerallut, Jean Francois
Author_Institution :
Dept. of Electr. Eng., Univ. Autonoma Metropolitana, Mexico City
fDate :
Aug. 30 2006-Sept. 3 2006
Abstract :
A novel edge-preserving filtering method based on unsupervised clustering within the joint four-dimensional location-intensity space of volume echocardiographic data is presented. A vector codebook, formed by vectors that lie close to the modes of the joint data distribution is derived through an iterative vector quantization procedure. Distribution estimates are non-parametric, kernel based, and thus the quality of the quantization can be controlled by adequate choice of the kernel parameters
Keywords :
data visualisation; echocardiography; image segmentation; iterative methods; medical image processing; pattern clustering; rendering (computer graphics); smoothing methods; statistical analysis; unsupervised learning; vector quantisation; 3D+t visualization; distribution estimates; echocardiographic volume rendering; edge-preserving filtering method; image segmentation; iterative vector quantization; joint data distribution; joint four-dimensional location-intensity space; kernel parameters; nonparametric estimates; smoothing methods; unsupervised clustering; vector codebook; Filtering; Image processing; Image segmentation; Information geometry; Kernel; Laboratories; Rendering (computer graphics); Smoothing methods; Ultrasonic imaging; Vector quantization;
Conference_Titel :
Engineering in Medicine and Biology Society, 2006. EMBS '06. 28th Annual International Conference of the IEEE
Conference_Location :
New York, NY
Print_ISBN :
1-4244-0032-5
Electronic_ISBN :
1557-170X
DOI :
10.1109/IEMBS.2006.260687
