A hybrid method for 3D segmentation of MRI brain images

Segmentation of some complex images, especially in brain MRI, is often difficult to perform satisfactorily using only a single image segmentation approach. An approach towards the integration of several techniques seems to be the best solution. A new hybrid method for 3D segmentation of the whole brain based on fuzzy region growing, edge detection and mathematical morphology is introduced. The threshold controlling the process of region growing is determined by a fuzzy technique. By combining methods of edge detection and region growing, and post-processing using mathematical morphology, the whole brain region is obtained. Experimental results indicate that the proposed method for 3D segmentation of MRI brain images provides much better results than the traditional method using a single technique in the segmentation of a human brain MRI data set.