基于一种新的非局部二次MRF先验模型的Bayesian图像重建  

Bayesian Image Reconstruction Algorithm Based on a Novel Nonlocal Quadratic MRF Prior Model

作　　者：陈阳[1] 严勇[2] 吴昊[2] 罗立民[1] 陈武凡[3] 

机构地区：[1]东南大学计算机科学与工程学院影像科学与技术实验室,江苏南京210096 [2]中国人民解放军总医院医学工程保障中心,北京100853 [3]南方医科大学生物医学工程学院,广东广州510515

出　　处：《电子学报》2009年第4期744-749,共6页Acta Electronica Sinica

基　　金：国家973重点基础研究发展规划(No.2003CB716102)

摘　　要：Bayesian方法或最大后验(maximuma posteriori,MAP)法被认为是解决图像重建中的病态问题的有效方法.根据Bayesian理论,目标图像的先验信息被加诸于图像重建中来抑制噪声.然而,大部分的先验模型提供的先验信息来自于一个较小的局部邻域内灰度值的简单加权差,只能对Bayesian重建提供有限的先验信息.本文提出一个新的非局部的且具有二次的先验能量方程的马尔可夫随机场(Markov RandomFields,MRF)先验,该先验通过选择较大邻域和新的加权方式来充分利用图像的全局信息.文章同时还提出使用该非局部先验的Bayesian重建的迭代算法.最后给出了该先验在PET(正电子发射成像)重建中的应用.实验结果以及同其他先验的比较证明该先验在降低噪声效果和保持边缘方面具有很好的表现.Bayesian approach, or MAP （maximum a posteriori） method, has been accepted as an effective solution to overcome the ill-posed problem of image reconstruction. Based on Bayesian theory, a prior information of the objective image is imposed on image reconstruction to suppress noise. However, the information from most prior model is from a simple weighted differences of the pixel intensities in a small local neighborhood, thus can only provide limit prior information for Bayesian Reconstruction. In this article, a novel nortlocal MRF （Markov Random Fields） prior with convex prior energy function, which is able to exploit global information in image by choosing large neighborhood and a new weighting strategy. Iterative algorithm for Bayesian reconstruction using the proposed nonlocal prior is also given. Experiments about the proposed prior＇ s application in PET （Positron Emission Tomography） reconstruction are inustrated. Experimental results and comparisons with other priors proved the proposed prior＇ s good performance in both lowering noise effect and preserving edges.

关 键 词：Bayesian重建 正电子发射成像(PET) 马尔可夫随机场(Markov Random Fields MRF) 

分 类 号：R318[医药卫生—生物医学工程]