Extraction of Grayscale Brain Tumor in Magnetic Resonance Image

Abstract

Magnetic resonance imaging (MRI) has great contribution in diagnosis and treatment of brain tumor. Segmentation of grayscale tumor or low intensity tumor is the most important and challenging task. In this paper, we propose a method to extract grayscale tumor from MRI image. Due to low intensity profile of input image, global contrast enhancement method is applied as a pre-processing step. This enhanced grayscale image is converted into binary image. Further, algorithm segments largest connected region using mathematical morphological and segmentation methods. At last, tumor region is extracted using resultant image and input MRI image. Experimental results show the proposed method effectively extracts a grayscale brain tumor from MRI images.

An Approach for Enhancement of MR Images of Brain Tumor

“…Birla and Shantaiya [21] employed a combination of blind convolution techniques, median filter, and Wiener filter. Sharma and Meghrajani [22] suggested a method for enhancing the contrast of low-intensity greyscale MR images with histogram equalization. Benson and Lajish [23] proposed an algorithm for skull stripping of the brain via mathematical morphology.…”

Automated segmentation of multiple sclerosis lesions using statistical approach

“…In our previous work, we have used mathematical morphological reconstruction for segmenting single higher intensity brain tumor of largest size [6]. For segmenting grayscale brain tumor with low intensity, Bauer [7] used Markov random field based tumor growth model, non-rigid registration based on atlas based segmentation and in our previous work, we have used mathematical morphological and segmentation methods for grayscale tumor segmentation [8]. For detecting multiple lesions from MRI image, Goldberg-Zimring [10] applied artificial neural networks as an automatic algorithm for contouring of multiple lesions in brain MR images.…”