Optimized segmentation for MRI brain images

Magnetic Resonance Imaging plays a pivotal role in pre-surgical evaluation of patients with intractable epilepsy. Research on optimal segmentation process from MRI brain images was carried out. The methods such as K-Means and Fuzzy C Means were done for segmentation of White Matter, Gray Matter and Cerebro-Spinal Fluid tissues. Clustering technique is a neighborhood attraction, which is dependent on the relative location and features of neighboring pixels. FCM is more effective to the fuzzy boundary region segment, but the major drawback is that no better way to determine the centroid value of clustering and the initial cluster centers, essentially. FCM is a local search optimization algorithm, it will converge to the local minimum point and this clustering effect would have a greater impact if the initial centroid values are not properly. To overcome this limitation, Genetic Algorithm is integrated with FCM for improving the segmentation performance with higher accuracy rate.