选择一体化PET/MR小病灶图像重建方案

目的 观察不同重建条件组合对不同大小病灶图像质量的影响,提出一体化PET/MR对于小病灶的最佳图像重建方案。方法 采用GE SIGNA TOF PET/MR系统对符合国际电工协会（IEC）推荐标准的PET图像质量（IQ）发射体模进行扫描,并对列表模式PET原始数据按照以下条件分别组合进行重建：有序子集最大期望值（迭代子集：28;迭代次数：1~9次）;重建矩阵（128×128,192×192,256×256）;高斯低通滤波器,半高宽（FWHM）（1~6 mm）;单独使用飞行时间（TOF）技术;单独使用点扩散函数（PSF）;同时使用TOF和PSF（TOF+PSF）;不使用TOF和PSF（NOTOF+NOPSF）。分析不同重建条件下图像对比度恢复（CR）、背景变化率（BV）和信噪比（SNR）,评估图像质量。结果 10 mm和13 mm小球在OSEM（3次迭代,28个子集）、192×192矩阵、2 mm高斯滤波半高宽、TOF+PSF重建组合下信噪比最高,分别为13.31%、21.73%,13.31%、21.73%,25.74%、35.80%和26.25%、46.01%。结论 3次迭代、28个子集、192×192矩阵、TOF+PSF为本研究条件下一体化PET/MR对于模拟小病灶的最佳图像重建方案。

effects of different reconstruction combinations on image quality of different sizes of lesions

Objective To explore the impacts of different reconstruction combinations on image quality of different sized lesions, so as to propose the optimal reconstruction scheme for smaller lesions on integrated PET/MR. Methods SIGNA TOF PET/MR system was used for scanning of PET image quality (IQ) emission phantom designed taken standards of the International Electrotechnical Commission (IEC) recommendations. Then the list mode raw data of PET were reconstructed using time of flight(TOF)-ordered subsets expectation maximization(OSEM)under different conditions, including the maximum expected value of the ordered subset (the iterative subset:28, the number of iterations:1-9 times),the reconstruction matrix (128×128, 192×192, 256×256), Gaussian low-pass filter full width at half maximum (FWHM) (1-6 mm), time-of-flight (TOF) technology alone (TOF+NOTOF), separate point spread function (PSF) (PSF+NOPSF), both TOF and PSF (TOF+PSF), as well as neither TOF nor PSF (NOTOF+NOPSF). Imaging quality was assessed under different reconstruction conditions using parameters including contrast recovery (CR), background variability (BV) and signal-to-noise ratio (SNR). Results Balls with 10 and 13 mm diameters simulating small lesions in the OSEM (3 iterations, 28 subsets), 192×192 matrix, 2 mm Gaussian low-pass filter full width at half maximum and TOF+PSF reconstruction combination achieved the highest SNR, which was 13.31% and 21.73%, 13.31% and 21.73%, 25.74% and 35.8% as well as 26.25% and 46.01%, respectively. Conclusion Using reconstruction combination as 3 iterations, 28 subsets, 192×192 matrix and TOF+PSF, the best image quality of small lesions can be obtained under the above mentioned conditions.