Many methods have been proposed for MRI brain tissue segmentation. Since the boundaries of different tissues in MRI brain images are indistinct and the intensities of the white and gray matter are very close, the edge-based segmentation methods may not provide satisfactory results. Thus the knowledge-based methods, e.g. clustering algorithms for brain tissue classification, integrated edge and region detection, and deformable template-based methods are being investigated for automatic and accurate image segmentation. Knowledge-based methods include several pattern recognition procedures needing a priori anatomical knowledge of the region-of-interest and the human training data set. Clustering methods such as Kmeans and Fuzzy C-means do not always arrive at a meaningful segmentation, and often require long computational time. Effective segmentation is required. For MRI image segmentation, it is important to identify anatomical areas of interest for diagnosis. In this context, we are more concerned with the anatomical region or structure segmentation. We know that the intensity of the MRI image of human tissue is homogeneous and the structure of each tissue is connected, but it is difficult to separate the adjacent tissues due to the small intensity changes and smoothed boundaries between the tissues. The intensity-based segmentation using global thresholding can not segment MRI images properly because of the non-uniform nature of the MRI. 
    Combining both spatial and intensity information in the image, this design based on an MRI brain image segmentation approach based on multi-resolution edge detection, region selection, and intensity threshold methods. The detection of white matter structure in the brain is emphasized in this project. First, a multi-resolution brain image representation and segmentation procedure based on a multi-scale image filtering method are presented. Given the nature of the structural connectivity and intensity homogeneity of brain tissues, region-based methods such as region growing and subtraction to segment the brain tissue structure from the multi-resolution images are utilized. From the segmented structure, the region-of-interest (ROI) image in the structure region is derived, and then a modified segmentation of the ROI based on an automatic threshold method using our threshold selection criterion is presented. Examples on both T1 and T2 weighted MRI brain image segmentation is presented, showing finer brain tissue structures.

Reference Paper: MRI brain image segmentation by multi-resolution edge detection and region selection
Author’s Name: H. Tang, E.X. Wua, Q.Y. Ma, D. Gallagher, G.M. Perera, and T. Zhuang
Source: Elsevier
Year:2004

Request source code for academic purpose, fill REQUEST FORM or contact +91 7904568456 by WhatsApp or sales@verilogcourseteam.com, fee applicable.