Brain MRI Segmentation Using Fuzzy Clustering Algorithms

Authors : Fatemeh JAMALİNABİJAN, Ayda BAHERİ ESLAMİ, Gulcihan OZDEMİR

Pages : 112-118

View : 40 | Download : 14

Publication Date : 2021-12-30

Article Type : Research Paper

Abstract :On the authority of human realization, the process of partitioning an image into non-overlapping and meaningful parts is called image segmentation. One of the traditional and conventional implementations of image segmentation is Brain MRI segmentation. In most cases, the MRI segmentation procedures are based on clustering approaches and according to the literature studies FCM based algorithms are more noticeable among other methods. Due to some drawbacks of FCM algorithm, like its weak function in the presence of noise, random initial values and easily falling into local optimal solution research have been trying to make some improvements on FCM algorithm. There are plenty of novel FCM based algorithms and In this work, we have implemented two FCM based algorithms insert ignore into journalissuearticles values(ARKFCM, SFCM2D); with different types of brain MRI images and compared them with conventional FCM to see which one has the better performance on the images with and without noise. Results are shown in the form of segmented images, and they demonstrate that ARK-FCM shows a better performance in keeping the details and being more resistant in working on noisy MRI images.
Keywords : Fuzzy, Image segmentation, clustering, Medical images

ORIGINAL ARTICLE URL