The PET Image Reconstruction Based on Combined Bases and Reweighted Sparse Regularization
Positron Emission Tomography (PET) is an important technology for clinical diagnosis.This article studied the PET image reconstruction techniques with Poisson noise interference.Firstly we established an image reconstruction model using combined bases and reweighted minimization sparse regularization under the framework of Poisson likelihood function,and used the negative log-likelihood function as the objective function,the reweighted minimization of Wavelet Transform and Discrete Cosine Transform (DCT) were used as regularization constraint.Then the objective function was solved using split Bregman method and it was transferred into two sub-problems,the sparse regularization of Gaussian model and Poisson denoising.We solved each sub-problem and got the iterative reconstruction algorithms.In the end the numerical simulations with Zubal model were carried out and it had shown the effectiveness of the proposed model and algorithms.
PET Image Reconstruction Combined Bases Reweighted Minimization Sparse Regularization Split Bregman
Ji-Jun TONG Qin-Guang LIN Qiao-Zhi QI Qiang CAI
School of Information Science & Technology,Zhejiang Sci-Tech University,Hangzhou,310018,China Yangtze Delta Region Institute of Tsinghua University,Jiaxing,314000,China
国内会议
杭州
英文
1-6
2014-10-18（万方平台首次上网日期，不代表论文的发表时间）