Department of Medical Physics, Medical School, University of Patras, 265 00 Patras, Greece.
Comput Methods Programs Biomed. 2010 Sep;99(3):219-29. doi: 10.1016/j.cmpb.2009.11.011. Epub 2010 Jan 18.
We have studied the properties of the pixel updating coefficients in the 2D ordered subsets expectation maximization (OSEM) algorithm for iterative image reconstruction in positron emission tomography, in order to address the problem of image quality degradation-a known property of the technique after a number of iterations. The behavior of the updating coefficients has been extensively analyzed on synthetic coincidence data, using the necessary software tools. The experiments showed that the statistical properties of these coefficients can be correlated with the quality of the reconstructed images as a function of the activity distribution in the source and the number of subsets used. Considering the fact that these properties can be quantified during the reconstruction process of data from real scans where the activity distribution in the source is unknown the results of this study might be useful for the development of a stopping criterion for the OSEM algorithm.
我们研究了二维有序子集期望最大化（OSEM）算法中像素更新系数的性质，以便解决在正电子发射断层成像中迭代图像重建时图像质量下降的问题，这是该技术在经过多次迭代后的一个已知问题。使用必要的软件工具，我们在合成符合数据上对更新系数的行为进行了广泛分析。实验表明，这些系数的统计性质可以与重建图像的质量相关联，其函数是源中的放射性活度分布和使用的子集数量。考虑到在对源中放射性活度分布未知的实际扫描数据进行重建过程中可以对这些性质进行量化的事实，这项研究的结果可能有助于为 OSEM 算法开发一个停止准则。
