Authors: MOHAMED ZAKI ABDERREZAK, MOUATEZ BILLAH CHIBANE, KARIM MANSOUR
Abstract: The aim of this work is to present a new method for cerebral MRI image segmentation based on modification of the fuzzy c-means (FCM) algorithm. We used local and nonlocal information distance in the initial function of the robust FCM model. The obtained results of the classification of MRI images showed the effectiveness of the suggested model. Calculation of the similarity index confirms that our method is well adapted to MRI images even in the presence of noise.
Keywords: Segmentation, magnetic resonance imaging, fuzzy c-means algorithm
Full Text: PDF