基于混合模型的醫(yī)學圖像融合研究

12700字
自己原創(chuàng)的畢業(yè)論文，已經(jīng)通過校內(nèi)系統(tǒng)檢測，重復率低，僅在本站獨家出售，大家放心下載使用

摘 要 在科學技術快速發(fā)展的今天，醫(yī)學成像技術已經(jīng)進入了前所未有的快速發(fā)展中，醫(yī)學圖像成為了臨床診斷治療的重要輔助手段。然而由于成像原理的不同，得到的醫(yī)學圖像特點也就不一樣，而通過單一模態(tài)的圖像所包含的信息在臨床上很難讓醫(yī)生進行全面綜合的診斷?；趩我荒B(tài)的醫(yī)學圖像效果不理想，且它們之間往往存在著很大的互補性和冗余性，因此醫(yī)學圖像融合技術就顯得尤為重要，通過它能減少圖像間的冗余信息，增加互補信息已獲得更為可靠、全面的融合結果，可以全面反映病人的生理狀況，為醫(yī)生做出準確的診斷或制訂出合理的治療方案大好了基礎。它對醫(yī)學的發(fā)展有著重要的推動作用，醫(yī)學圖像融合技術也是在這樣的環(huán)境背景下發(fā)展并逐漸趨向成熟。
在閱讀大量國內(nèi)外優(yōu)秀有關醫(yī)學圖像融合文獻的基礎上，本文從介紹醫(yī)學圖像融合的研究意義和發(fā)展現(xiàn)狀出發(fā)，分析了醫(yī)學圖像融合的基本原理和具體的實現(xiàn)過程，和常用的圖像融合方法，并提出了基于混合模型的醫(yī)學圖像融合研究的方法。先用高斯混合模型對醫(yī)學圖像建模，在通過EM算法對所建的高斯模型進行參數(shù)估計，最終進行圖像重構得到融合結果，且通過實驗仿真驗證了算法的有效性。本文還通過對加權平均算法融合方法和基于混合模型兩種方法進行了比較分析，深入研究了基于高斯混合模型對圖像融合的具體影響，通過實驗結果表明基于高斯混合模型的圖像融合算法效果比較好。
關鍵詞：醫(yī)學圖像融合 高斯混合模型 EM算法

Medical Image Fusion Based On Mixed Model
Abstract In the rapid development of science and technology today, medical imaging technology has entered a rapid development of hitherto unknown, the medical image has become an important adjunct to clinical diagnosis and treatment. However, because of the different imaging principle, image features are also not the same, and image by a single mode of the information contained in clinically difficult to let the doctor to diagnose the comprehensive. Based on the medical image effect of a single mode is not ideal, and among them there are complementary and redundancy of big, so it is very important for medical image fusion technique, which can reduce the redundant information between images, adding complementary information has been more fusion results are reliable, comprehensive, can fully reflect the physiological status of the patient the doctor, to make an accurate diagnosis or make good foundation reasonable treatment scheme. It is the development of medicine plays an important role in promoting, and medical image fusion technology is also in the context of such an environment development and gradually mature.
On the basis of reading a lot of domestic and foreign excellent medical image fusion literature, this paper introduces the research significance and the development present situation of medical image fusion, analyzes the basic principles of medical image fusion and the concrete implementation process, analysis of the common image fusion method, and puts forward the research of medical image fusion hybrid model based on the. First with the Gauss mixture model of medical image modeling, through the EM algorithm to estimate the model parameters of Gauss, the final image reconstructed to obtain the fusion results, and the results of simulation show the effectiveness of the algorithm. The fusion method of weighted average algorithm and hybrid models of two kinds of methods based on a comparative analysis, in-depth study of the Gauss mixture model of concrete based on image fusion, the experimental results show that the image of Gauss mixture model fusion algorithm based on the effect is better.
Key words Medical image fusion Gaussian mixture model EM algorithm

目錄
摘 要 I
Abstract II
第一章 緒論 1
1.1醫(yī)學成像技術的發(fā)展 1
1.2課題背景及意義 2
1.3醫(yī)學圖像融合的研究現(xiàn)狀 2
1.4論文主要內(nèi)容和章節(jié)安排 5
第二章 醫(yī)學圖像融合和預處理技術 6
2.1醫(yī)學圖像融合概念 6
2.2醫(yī)學圖像融合方法分類 6
2.2.1像素級融合 6
2.2.2特征級融合 8
2.2.3決策級融合 9
2.3 醫(yī)學圖像融合的評價標準 9
2.3.1主觀評價 9
2.3.2客觀評價 10
2.4 醫(yī)學圖像融合前準備階段介紹 12
2.4.1醫(yī)學圖像預處理 13
2.4.2醫(yī)學圖像配準 13
2.5小結 14
第三章 相關介紹 16
3.1 高斯混合模型 16
3.2 EM算法 16
3.3 多分辨率分析 19
第四章 基于高斯混合模型的醫(yī)學圖像融合 20
4.1 圖像信息模型 20
4.2 基于多分辨的高斯混合醫(yī)學圖像融合 21
4.3 實驗與結果 22
4.4 本章小結 24
結 論 25
致 謝 26
參考文獻 27