Segmentation of white matter, gray matter, and csf from MR brain images and extraction of vertebrae from MR spinal images [electronic resource] /

by Peng, Zhinang; Theses and, OhioLINK Electronic

Abstract (Summary)
In this dissertation, we address two kinds of the biomedical images/volumes segmentation problems: 1). Segmentation of white matter, gray matter, and cerebral spinal fluid from MR brain images/volumes; 2). Extraction of the vertebrae from MR spinal images. We propose a statistical decision model under maximum a posterior probability (MAP) estimation and Markov random field (MRF) framework to segment MR brain images, where the spatial-varying Gaussian mixture (SVGM) is used to represent the intensity probability distribution of each of the three brain tissues, and MRF is used to estimate the prior probability. Three methods, the supervised method, the automatic (unsupervised) method, and the 3D method, are proposed to achieve the final segmentation. The supervised method is a 2D method proposed to effectively learn the expert's segmentation and pursue the tissue labeling on 3D MR brain images with severe intensity non-uniformity. The fully automatic or unsupervised 2D method is presented to accurately segment WM, GM, and CSF. The parameters of SVGM are estimated from the reference images using either the expectation maximization (EM) algorithm or the MKM algorithm, and the final tissue labeling are obtained by using the ICM algorithm. Furthermore we impose a criterion that minimizes the intensity means difference within the same segmentation tissue to improve the accuracy of the parameters of SVGM. Due to the nature three dimensional characteristics of MR brain volumes, we extend the 2D SVGM-MRF method to 3D and employ the 3D spatial and intensity information for a more accurate conditional probability representation. To reduce the large computation time and memory requirements for 3D implementation, the algorithms using a local window instead of the whole volume are proposed to perform the necessary parameter estimations and achieve the tissue labeling. For MR spine images segmentation problem, we propose a two-step algorithm that can detect and segment the vertebrae simultaneously. In the first step, an intensity profile on a polynomial function for all these disk clues on the best slice is used to pursue the disk searching process. Sequentially, vertebra centers are detected, and initial boundaries are extracted in the second step.
Bibliographical Information:


School:University of Cincinnati

School Location:USA - Ohio

Source Type:Master's Thesis

Keywords:university of cincinnati


Date of Publication:

© 2009 All Rights Reserved.