基于全变分α散度最小化的PET优质重建

摘要

为了获得优质的PET成像,本文提出一种基于全变分阿尔法散度最小化的PET重建新方法.新方法通过引入阿尔法散度度量投影数据和估计值之间的偏差;通过增加全变分正则化修正阿尔法散度最小化解的一致性.针对新构建的PET重建目标函数的求解,本文提出一种基于次梯度理论的交替式迭代策略,期间运用自适应非单调线性搜索来保证算法的收敛性.仿真和临床PET数据实验表明,本文方法在噪声抑制和边缘保持方面均优于传统的PET重建方法.%To achieve high diagnostic PET imaging, we propose a novel total variation (TV) based alpha-divergence minimization reconstruction algorithm. The presented cost function uses the alpha-divergence to measure the discrepancy between the measured and the estimated emission projection data and utilizes the TV regularization to regularize the consistency of solution. A semi-implicit iteration scheme is used in the proposed algorithm by adapting the subgradient theory; and then an adaptive nonmono-tone line search scheme is taken to guarantee the algorithm convergence.The experiments from the simulated phantom data and the real emission data show that the presented algorithm performs better than the other classical PET reconstruction methods in the noise suppressing and the edge details preserving.