Efficient 3-class Fuzzy C-Means Clustering algorithm with Thresholding for Effective Medical Image Segmentation

Medical image segmentation is a method of extracting the desired parts and features from the input medical image data. The conventional FCM algorithm is an efficient clustering algorithm that is used in medical image segmentation. But FCM is extremely susceptible to noise since it uses intensity values for clustering the image. This paper aims to develop 3-class FCM algorithm with thresholding which is noise efficient. The proposed 3-Class FCM method improves efficiency of image by modifying the initialization of fuzzy petition matrix this method uses normally distributed pseudorandom numbers generator with Gaussian distribution for initial estimates of petition matrix. The thresholding with FCM is used to generate the logical segmented images. As another improvement in this paper, colour segments are generated by utilizing the results of FCM. And the results of the proposed FCM method are tested upon the variety of medical images and compared with the widely used Global thresholding and Otsu’s method. The performance is also evaluated for comparison of FCM with different distance masers. Keywords—Color segment’s, Medical Image segmentation, Fuzzy C-mean Clustering, Thresholding, Entropy.