Sobel Operator with BayesShrink Wavelet De-Noising for Segmentation of Neurosarcoidosis in Brain MRI Images

Neurosarcoidosis is a disorder caused due to an unknown situation which gives raise to complication of sarcoidosis. It leads to inflammation in nervous system like brain and spinal cord of human body. This paper proposes a new method for segmenting neurosarcoidosis in brain MRI image using Sobel edge detection over BayesShrink wavelet thresholding. The BayesShrink threshold wavelet acts a boosting technique by mitigating Gaussian white noise appearing the image. The performance of proposed algorithm is evaluated by comparing the segmentation results with the standard Sobel method, Prewitt method, Laplacian of Gaussian method and Canny method of image segmentation techniques. It is clearly observed that the proposed algorithm efficiently segments part of the MRI brain image affected by neurosarcoidosis.