Brain Tumor Segmentation Using clustering Approach

Segmentation is main step at Brain tumor detection in the medical field. Although there are number of methods have been available for brain tumor segmentation, but segmenting the tumor is a difficult and challenging task. Brain tumor’ segmentation must be done with accuracy in the clinical practices. The objective of this paper presents a comparison of brain MRI tumor segmentation methods. In this paper, various techniques like OTSU, K-MEANS, FCM, KWFLICM is compared. Comparative analysis of this segmentation methods results has been evaluated. In Existing methods Color, Cluster and Morphological Based Approach is applied to Segment the Part of Brain Tumor. In Color and morphological based method affected by noise Where Cluster based method needs A priori specification of no. of clusters. From above problems a proposed method can be made by Adaptive KWFCM method to segment accurate and fast brain tumor. Final results is compared and the analysis of pre-processing and segmentation methods is shows that combination of fast local laplacian filter and KWFLICM is gives accurate and faster comparing to FCM.