Brain MRI Segmentation Using a Fast Fuzzy c-means Algorithm

In this paper, we introduced a fast version of Fuzzy C-Means (FCM) algorithm for the task of brain MRI segmentation. In the proposed algorithm FaFCM (Fast Fuzzy c means), the calculation of the membership matrix and the cluster centers depend on histogram bins instead of individual pixels values to decrease execution time of standard FCM. To evaluate the performance of the proposed algorithm, FaFCM has been tested over a set of MR brain images corrupted by different levels of noise and intensity non-uniformity. The experimental results proved that the proposed algorithm FaFCM converges much faster than the conventional FCM.