Publication: Mri brain image segmentation
Date
2010-04-01
Authors
Chiah, Hock Chuan
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
In the research on magnetic resonance (MR) brain image, there are a lot of researchers using various segmentation techniques ranging from simple image segmentation technique to complex image segmentation technique in order to extract the tissues of research interest. The complex technique such as clustering, Markov Random Field (MRF) and deformable model are usually being used. The calculation of tissues volume will help recognize diseases. Recently, the support vector machine (SVM) has been widely used in pattern recognition application. It serves as a good generalization and classifier tool. SVM is one of the supervised learning techniques of neural network. There has not been much research using the SVM in image processing field especially when it comes to the MR brain images. This project presents SVM image segmentation method on magnetic resonance image of brain to segment the White Matter (WM), Gray Matter (GM) and Cerebrospinal Fluid (CSF). Firstly, a fast and robust algorithm also has been developed to strip the skull out of the MR brain image. This is an important pre-processing step. The algorithm involves combination of morphology and image segmentation techniques. Next, the SVM will be used to classify the skull stripped MR image into three types of tissues: WM, GM and CSF. This project takes a set of MR brain images of 3 different patients. From the outcome of this project, the SVM has been proven to be a good segmentation tool in MRI brain image segmentation.