CHOSUN

Extraction of Brain Tumor Using Level Set Method with Automatic Selective Local Statistics and SPF in MR Images

Metadata Downloads
Author(s)
Kiran Thapaliya
Issued Date
2012
Abstract
Level set 방식은 영상의 추출을 위한 훌륭한 도구로 사용되어질 수 있다. 본 연구의 목적은 MR 영상으로부터 뇌종양 영상의 추출 방법을 제안하는 것이다. 이 논문에서는 무뎌진 에지 혹은 약한 윤곽선을 멈추도록 새로운 SPF 방법을 소개한다. 다른 객체 부분의 지역적 통계치는 계산되어 진다. MR영상에서 지역 통계에 의해 종양객체는 다른 객체 중 동일시된다. Level set 방식에 있어 파라미터의 계산은 도전적인 역할이다. 영상의 다른 종류를 위한 다른 파라미터의 계산은 자동적이다. 기본적인 임계치는 다른 MR 영상을 위한 자동적으로 저장되고 업데이트 된다. 이러한 임계치는 제안한 알고리듬에 있어 다른 파라미터로 계산되어진다. 제안한 알고리듬은 종양추출을 위한 뇌의 MR 영상을 시험한다. 또한 시각적으로나 양적으로 평가한다. 뇌 종양 영상들의 다양한 실험은 제안한 방법의 효율성과 강인성을 입증한다.|The level set approach can be used as a powerful tool for the segmentation of the images. The purpose of this study is to propose a method of a segmentation of brain tumor images from MR images. In this paper, we introduce a new signed pressure function (SPF) which can efficiently stop the contours at weak or blurred edges. The local statistics of the different objects present in the MR images are calculated. With the help of the local statistics, tumor objects are identified among different objects. In this level set method, the calculation of parameters is a challenging job. The calculations of different parameters for different kind of images are automatic. The basic thresholding value is updated and adjusted automatically for different MR images. This thresholding values are used to calculate the different parameters in the proposed algorithm. The proposed algorithm has been tested on magnetic resonance images of the brain for tumor segmentation and its performance evaluated visually and quantitatively. Numerical experiments on some brain tumor images have demonstrated the efficiency and robustness of our method.
Alternative Title
MR 영상에서 SPF와 자동 선택적 지역 통계에 의한 Level set 방식을 이용한 뇌 종양 추출
Alternative Author(s)
타팔리야기란
Affiliation
Chosun University, Graduate School of Chosun University
Department
일반대학원 정보통신공학과
Advisor
Goo-Rak Kwon
Awarded Date
2012-08
Table Of Contents
Table of Contents I
List of Tables iii
List of Figures iv
Acronyms vii
Abstract viii

Ⅰ. Introduction 1
A. Thesis Motivation and Overview 2
B. Research Objectives 3
C. Thesis Contribution 3
D. Thesis Organization 4
II. Background 5
A. Brain Tumor 5
B. Classification of Brain Tumor 5
1. Classification of Brain Tumor by WHO 6
2. Classification of Brain Tumors Based on their Location 7
3. Classification of Brain Tumor Based on their Radiologic
Appearance 8
4. Classification of Brain Tumor Based on their Alterations 9
C. Overview of Related Works 10
1. The GAC and C-V Models 14
a. The GAC Model 14
b. C-V Model 15
Ⅲ. The Proposed Method 18
A. Level Set Model and SPF 18
B. Implementation 19
C. Threshold and Parameter Calculation 21
D. Extraction of Tumor from Multiple Objects 24
Ⅳ. Performance Evaluation 26
A. Subjective Quality 26
B. Segmentation Validation and Quantitative Analysis 48
C. Complexity Analysis 54
Ⅴ. Conclusion 55
References 56
List of Publications 60
Degree
Master
Publisher
Chosun University, Graduate School of Chosun University
Citation
Kiran Thapaliya. (2012). Extraction of Brain Tumor Using Level Set Method with Automatic Selective Local Statistics and SPF in MR Images.
Type
Dissertation
URI
https://oak.chosun.ac.kr/handle/2020.oak/9515
http://chosun.dcollection.net/common/orgView/200000263309
Appears in Collections:
General Graduate School > 3. Theses(Master)
Authorize & License
  • AuthorizeOpen
  • Embargo2012-08-09
Files in This Item:

Items in Repository are protected by copyright, with all rights reserved, unless otherwise indicated.