雷昕锟,王镜添,郭小瑞,郝潞平,武晓伟,尚华.点扩散函数校正迭代次数对于体模18F-FDG标准摄取值的影响[J].中国医学影像技术,2026,42(3):439~443
点扩散函数校正迭代次数对于体模18F-FDG标准摄取值的影响
Impact of iteration number of point spread function correction on 18F-FDG standard uptake value of phantom
投稿时间:2025-04-11  修订日期:2025-09-29
DOI:10.13929/j.issn.1003-3289.2026.03.024
中文关键词:  正电子发射断层扫描与计算机断层扫描  氟脱氧葡萄糖F18  迭代次数
英文关键词:positron-emission tomography and computed tomography  fluorodeoxyglucose F18  iteration ordinal number
基金项目:
作者单位E-mail
雷昕锟 河北医科大学第二医院核医学科, 河北 石家庄 050061
河北省核医学专业质量管理与控制中心, 河北 石家庄 050061 
 
王镜添 河北医科大学第二医院核医学科, 河北 石家庄 050061
河北省核医学专业质量管理与控制中心, 河北 石家庄 050061 
 
郭小瑞 河北医科大学第二医院核医学科, 河北 石家庄 050061
河北省核医学专业质量管理与控制中心, 河北 石家庄 050061 
 
郝潞平 河北医科大学第二医院核医学科, 河北 石家庄 050061
河北省核医学专业质量管理与控制中心, 河北 石家庄 050061 
 
武晓伟 河北医科大学第二医院核医学科, 河北 石家庄 050061
河北省核医学专业质量管理与控制中心, 河北 石家庄 050061 
 
尚华 河北医科大学第二医院核医学科, 河北 石家庄 050061
河北省核医学专业质量管理与控制中心, 河北 石家庄 050061 
27700107@hcbmu.edu.cn 
摘要点击次数: 71
全文下载次数: 50
中文摘要:
      目的 观察点扩散函数(PSF)校正迭代次数对于体模18F-FDG标准摄取值(SUV)的影响。方法 对国际电工[LM]委员会体模中直径分别为10、13、17、22、28及37 mm的6个球体灌注浓度20 kBq/ml的18F-FDG溶液后,利用Vereos PET/CT系统对其进行PET/CT扫描,并以有序子集最大期望值法(OSEM)重建PET图像(参数固定为迭代次数2,8个子集),且行PSF校正,其正则化次数为10,迭代次数分别为1、2、3、4、5、6、7、8、9、10、13、15、20、25、35及50,共获得16组相应PET数据,将PSF(10,1)组记为PSF_1iter组,依此类推。测量体模中各球体的最大SUV(SUVmax)及平均SUV(SUVmean),计算SUV恢复系数(RC)及其误差值;比较连续组间SUV差异,观察PSF迭代次数及球体大小对SUV RC的影响。结果 基于PET所测连续组间球体平均SUVmax差异均有统计学意义(P均<0.05),SUVmax随迭代次数增加而增加;PSF_1iter组至PSF_7iter组连续组间及PSF_25iter组与PSF_35iter组间球体平均SUVmean差异均有统计学意义(P均<0.05),表现为SUVmean随迭代次数增加而增加,其余连续组间差异均无统计学意义(P均>0.05)。随PSF校正迭代次数增加,同一球体SUVmax RC及SUVmean RC误差值均呈增大趋势,而SUVmean RC比SUVmax RC更接近真实值。随PSF校正迭代次数增加,直径10、13及17 mm球体的SUVmax RC和SUVmean RC增幅均大于其他球体。结论 体模SUVmax随PSF校正迭代次数增加而增加,而SUVmean在迭代次数为1~7时表现为逐渐增加、迭代次数>7时则具有收敛性。
英文摘要:
      Objective To investigate the impact of iteration number of point spread function (PSF) on 18F-FDG standardized uptake values (SUV) of phantom. Methods Six spheres with diameter of 10, 13, 17, 22, 28 and 37 mm in International Electrotechnical Commission body phantom were filled with 18F-FDG solution of 20 kBq/ml concentration. PET/CT scanning was performed using Vereos PET/CT system, and PET images were reconstructed using ordered subsets expectation maximization (OSEM) (with fixed parameters of 2 iterations and 8 subsets) and PSF correction (regularization number was set to be 10, while PSF iteration number was 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 13, 15, 20, 25, 35 and 50, respectively). Totally 16 sets of PET data were obtained, PSF (10, 1) group was denoted as PSF_1iter group, and so on. The maximum SUV (SUVmax) and mean SUV (SUVmean) of each sphere in the phantom were measured, while SUV recovery coefficient (RC) and its error value were calculated. SUV between consecutive groups were compared, and the impact of PSF iteration number and sphere size on SUV RC were observed. Results The differences of average SUVmax of the spheres measured in PET between consecutive groups were all significant (all P<0.05), indicating that SUVmax increased with the increase of iteration number. There were significant differences of average SUVmean of the spheres between consecutive groups of PSF_1iter to PSF_7iter groups, also between PSF_25iter and PSF_35iter groups (all P<0.05), manifested as SUVmean increased with the increase of iteration number, while the average SUVmean of the spheres of the other consecutive groups were not different (all P>0.05). The error values of SUVmax RC and SUVmean RC for the same sphere both showed an increasing trend when PSF correction iteration number increased. Compared with SUVmax RC, SUVmean RC of the spheres were closer to the true values. When PSF correction iteration number increased, the increased value of SUVmax RC and SUVmean RC in 10, 13 and 17 mm diameter spheres were all greater than those of other spheres. Conclusion With the increase of PSF correction iteration number, SUVmax of the phantom increased. SUVmean of the phantom showed a gradually increase when then iteration number ranged from 1 to 7, but became convergent when the iteration number was greater than 7.
查看全文  查看/发表评论  下载PDF阅读器