基于不同重建算法的PET重建参数优化研究

Optimization of PET Reconstruction Parameters based on Different Reconstruction Algorithms

目的:应用PET/CT模型评估不同重建算法条件下PET重建参数优化方案。方法:参照国际电工协会(IEC)标准中PET/CT测试模型的要求,采用聚甲基丙烯酸甲酯材料制作模型,内部配有直径大小不一的圆柱体用于模拟病灶。向小圆柱体和模型分别注入不同浓度的18F-FDG溶液,模拟放射性药物在人体内的分布情况。采用不同的重建算法及重建参数对PET/CT原始图像进行重建,通过对比分析不同重建条件下图像的放射性计数、对比度、背景变化率和信噪比进行图像质量评估。结果:当重建算法相同时,16子集3次迭代的条件下具有最为均衡的图像质量。相比于有序子集期望值最大化(OSEM)算法组,True-X组的背景变化率较低,并具有更高的对比度百分数和信噪比,差异具有统计学意义(均P<0.05)。结论:自制PET/CT模型可以作为临床工作中重建参数优化以及图像质量控制的有效研究工具。PET/CT重建参数可优先选用16子集3次迭代和True-X高清重建算法。

Purpose:To optimize PET reconstruction parameters under different reconstruction algorithms based on PET/CT phantom study.Methods:According to the International Electrotechnical Commission(IEC)standard,the phantom was made of polymethyl methacrylate.Cylinders of different diameters were made to simulate different lesions.The cylinders and the phantom were filled with 18F-FDG solution to simulate the distribution of radiopharmaceuticals in vivo.The counts in regions of interest,percent contrast,background variability,and signalto-noise ratios were evaluated under different reconstruction algorithms and parameters.Results:With same reconstruction algorithms,the optimal reconstruction condition was 16 sub-sets with three times of iterations.True-X method produced lower background variability,higher percent contrast values,and signal-to-noise ratios(all P<0.05).Conclusion:Self-manufacturing PET/CT phantom can be used for evaluating the quantification accuracy of different PET/CT reconstruction algorithms and parameters and PET/CT quality control.Sixteen sub-sets with three times of iterations and True-X algorithm are recommended.