Extraction of brain tumor from MR images using one-class support vector machine

J Zhou; K L Chan; V F Chong; S M Krishnan

doi:10.1109/IEMBS.2005.1615965

Extraction of brain tumor from MR images using one-class support vector machine

Conf Proc IEEE Eng Med Biol Soc. 2005:2005:6411-4. doi: 10.1109/IEMBS.2005.1615965.

Authors

J Zhou¹, K L Chan, V F Chong, S M Krishnan

Affiliation

¹ School of Chemical and Biomedical Engineering, Nanyang Technological University, Singapore.

PMID: 17281735
DOI: 10.1109/IEMBS.2005.1615965

Abstract

A novel image segmentation approach by exploring one-class support vector machine (SVM) has been developed for the extraction of brain tumor from magnetic resonance (MR) images. Based on one-class SVM, the proposed method has the ability of learning the nonlinear distribution of the image data without prior knowledge, via the automatic procedure of SVM parameters training and an implicit learning kernel. After the learning process, the segmentation task is performed. The proposed technique is applied to 24 clinical MR images of brain tumor for both visual and quantitative evaluations. Experimental results suggest that the proposed query-based approach provides an effective and promising method for brain tumor extraction from MR images with high accuracy.