Segmentation of breast MRI using effective Fuzzy C-Means method based on Support Vector Machine

Abstract

In this paper, we introduce improved segmentation method for segmenting breast MRI images using Kernelized Fuzzy C-Means and Support Vector Machine. Firstly, the new modified Kernelized FCM is constructed in this paper with inclusion of spatial neighborhood term. Then, the data are labeled by new FCM and the input vector for SVM classifier is generated by membership function of proposed new FCM. The robust SVM is proposed for providing effective segmentation result. In both KFCM and SVM, to enlighten the performance of dealing non linearity, the new Quadratic kernel function is used. In order to show the effectiveness of the proposed method, the experimental work is executed on artificially generated random data, and real dynamic contrast enhanced breast MRIs. The superiority of proposed method is proven by comparative analysis of result of proposed and existed methods. To work up the comparative analysis, Fuzzy Partition coefficient, Fuzzy Entropy and Silhouette method are utilized as a cluster validity measure in this paper. Finally, the experimental study shows our proposed method is promising method for segmenting medical images and complex data analysis.