APPLICATION OF A 3D LEVEL SET METHOD IN MRI SURFACE SEGMENTATION
Yu Hou · OhioLink ETD Center (Ohio Library and Information Network) · 2005
Level Set Method is a powerful numerical technique of tracking the evolution of interfaces moving under a variety of complex motions.The method has been widely used in the area of image segmentation.In these years, although there are a lot of literatures relating to Level Set Methods, it is still hard to find a comprehensive investigation for the application of this method, especially in three-dimensional situations.This thesis developed a software package to implement 3D MRI surface segmentations using Level Set Method.Very detailed discussion for its application on the brain MRI segmentation problem is presented.A new efficient speed function is used and evaluated in this 3D method.We did a large number of 3D brain tissues segmentation experiments on brain MRI data.For these tissues, we put the emphasis on analyzing and discussing the segmentation of hippocampus, because of it is significantly a more difficult and important segmentation problem.Using the evaluation of the experiment results with the ground truth, the settings for all of parameters are analyzed and concluded accurately.The software package is developed using Microsoft Visual C++ 6.0.