Improvement of MR brain images segmentation based on interval type-2 fuzzy C-Means

Magnetic resonance imaging (MRI) is a powerful tool for clinical diagnosis because it allows to distinguish different tissues and allows multiple modalities (T1, T2, ...) each having particular properties. This work is an improvement of an existing method which is Fuzzy C-Means (FCM) to separate the different tissues of MR Brain images-type 2 Fuzzy C-Means-. First, membership function defined by Hamid R Tizhoosh is used to measure the image fuzziness. Second, we propose a new membership function is proposed. The evaluation of adopted approaches was compared using the validity functions: partition coefficient Vpc and Vpe partition entropy. The experimental results on MR brain images prove that the proposed approaches are more accurate and robust than the standard FCM approach.