Title :
The generation of feature map in high dimensional feature space
Author :
He, Renjie ; Narayana, Ponnada A.
Author_Institution :
Medical Sch., Texas Univ., Houston, TX, USA
Abstract :
A method for generating feature maps in high dimensional (>4) feature space for tissue segmentation based on K-nearest neighbor (KNN) classification is presented. This technique considerably reduces the computational and memory complexity that are associated with the analysis of high dimensional feature space. This method has been successfully applied for segmenting MR images, based on four echoes, of multiple sclerosis brains.
Keywords :
biomedical MRI; brain; image segmentation; medical image processing; K-nearest neighbor classification; MR images; echoes; feature map generation; high dimensional feature space; sclerosis brains; tissue segmentation; Biomedical imaging; Computational complexity; Helium; Hypercubes; Image segmentation; Magnetic resonance imaging; Multiple sclerosis; Partitioning algorithms; Prototypes; Solids;
Conference_Titel :
Engineering in Medicine and Biology Society, 2003. Proceedings of the 25th Annual International Conference of the IEEE
Print_ISBN :
0-7803-7789-3
DOI :
10.1109/IEMBS.2003.1279842
