Skull removal of noisy magnetic resonance brain images using Contourlet transform and morphological operations

Efficient segmentation of noisy Magnetic Resonance (MR) brain images is a challenging task, as pre and post surgery decisions are required to make accurately in achieving better medical practices while treating brain disorders. This paper presents an automatic segmentation technique to remove non brain tissue (skull, fat, skin, muscle) of noisy MR brain images and to extract brain tissue (cortex and cerebellum). Here, Contourlet transform is applied to denoise a noisy MR brain image and threshold based morphological operations are applied to extract brain region on denoised images. Hence a comparative study is developed on skull removed MR brain images with and without denoising based on similarity index and segmentation error. The experimental results prove that the proposed method yields consistent results irrespective of noise levels.