Abstract

A Novel approach for the detection of tumor in brain MR images and its classification via Independent Component Analysis and Kernel Support Vector Machine

Author(s): Sandhya. G, Giri Babu.Kande, Satya Savithri.T

Automatic and exact detection and classification of tumors in brain MR images is very important for the medical analysis and interpretation. Tumors which are detected and treated in the early stage gives better long-term survival than those detected lately. The best classification algorithm helps to take appropriate decision and provides the best treatment. This paper proposes a novel approach for the accurate segmentation and classification of the brain tumor from MR images. Initially, the tumor image is pre-processed with the anisotropic diffusion filter then a region based active contour is used to detect the tumor. Active Contour Model (ACM) will provide smooth and close contours and can achieve high accuracy.In the classification process, various features are extracted from the tumor images using the 2-D Daubechies DWT. The feature vector dimensions are reduced using Independent Component Analysis (ICA). A trained Support Vector Machine with different kernels which can be treated as KSVM is used for the classification of the tumor. The suggested method can identify the benign and malignant type of tumors. The recommended method is also compared with the other two existing methods in terms of its effectiveness in segmentation as well as in classification. Results proved that designed method is effective and rapid