18F-FDG PET/CT不同重建方法对肺内病灶SUV及图像质量的影响

Effect of different reconstruction methods of 18F-FDG PET/CT on standardized uptake value and image quality of intrapulmonary lesions

  • 摘要:
    目的 研究18F-氟脱氧葡萄糖(FDG) PET/CT不同重建方法对肺内病灶标准摄取值(SUV)及图像质量的影响。
    方法 回顾性收集2023年3月至7月于江苏省中医院行PET/CT全身显像发现有肺内占位性病变且病灶摄取升高的44例恶性肿瘤患者的影像资料,其中男性33例、女性11例,年龄(68.3±11.0)岁。采用西门子Biograph vision 450 型PET/CT内置的不同重建方法进行图像质量分析。主要使用的重建方法包括反投影法(BP)、BP+时间飞行(TOF)、迭代(IT)、IT+TOF,分别对4组图像进行图像质量评估及定量分析。采用5级评分法进行图像视觉质量评估;分别对各组图像病灶、肝血池以及纵隔血池的定量参数最大标准摄取值(SUVmax)、平均标准摄取值(SUVmean)、SUV的标准差(SUVSD)、信噪比(SNR)以及相对SNR,以评估加入TOF后的重建方法对图像质量的影响。视觉评分的一致性分析采用Kappa检验进行评估。各组图像视觉评分构成的比较采用χ2检验。不符合正态分布的计量资料的组间比较采用Mann-Whitney U检验(秩和检验)。
    结果 2位医师对44例患者BP、BP+TOF、IT、IT+TOF 4组图像的视觉评分结果具有高度一致性(Kappa=0.862,95%CI:0.745~0.979,P<0.001)。BP与BP+TOF组间的视觉评分构成差异有统计学意义(χ2=28.037,P<0.001); IT与IT+TOF组间的视觉评分构成差异有统计学意义(χ2=47.143,P<0.001)。BP与IT加入TOF后均能降低噪声、提高SNR。与BP组相比,BP+TOF组病灶SNR由22.470(11.414,50.835)提升至28.835(15.101,68.391),Z=−2.684,P=0.007;与IT组相比,IT+TOF组病灶SNR由76.880(29.933,156.604)提升至126.593(40.364,250.589),Z=−3.583,P<0.001。IT+TOF组相对SNR>1的比例为97.7%(43/44),BP+TOF组相对SNR>1的比例为81.8%(36/44)。
    结论 使用BP与IT进行肺内病灶18F-FDG PET/CT显像时可加入TOF来提高图像质量。

     

    Abstract:
    Objective  To investigate the influence of different reconstruction methods of 18F-fluorodeoxyglucose (FDG) PET/CT on the standardized uptake value (SUV) and image quality of intrapulmonary lesions.
    Methods  A retrospective analysis was performed for 44 patients (33 males, 11 females, age (68.3±11.0) years) with malignant tumors who underwent PET/CT whole-body imaging in Jiangsu Province Hospital of Chinese Medicine from March to July 2023. Different reconstruction methods built into Siemens Biograph vision 450 PET/CT were used for image quality analysis. The main reconstruction methods used were back projection (BP), BP+time of flight (TOF), iterative (IT) and IT+TOF. Image quality was evaluated, and the four groups of images were quantitatively analyzed. A five-level scoring method was used to evaluate the visual quality of images. To evaluate the effect of the reconstruction method after adding TOF on image quality, the quantitative parameters of the lesions, liver blood pool, and mediastinal blood pool were measured in each group, including the maximum standardized uptake value (SUVmax), mean standardized uptake value (SUVmean), standard deviation of SUV (SUVSD), signal-to-noise ratio (SNR), and relative SNR. The consistency of visual scores was assessed using Kappa test. The composition of image visual scores among each group was compared using χ2 test. Non-normally distributed continuous data between groups were compared using the Mann-Whitney U test (rank-sum test).
    Results  The visual score results of the two physicians on the BP, BP+TOF, IT, and IT+TOF images of 44 patients were highly consistent (Kappa=0.862, 95%CI: 0.745–0.979, P<0.001). There was a statistically significant difference in the composition of visual score between BP and BP+TOF groups (χ2=28.037, P<0.001). A statistically significant difference in visual score composition was observed between the IT and IT+TOF groups (χ2=47.143, P<0.001). Quantitative analysis of BP and IT can reduce noise and improve SNR after adding TOF. Compared with the BP group, the SNR of lesions in the BP+TOF group increased from 22.470 (11.414, 50.835) to 28.835 (15.101, 68.391), Z=−2.684, P=0.007. Compared with the IT group, the SNR of lesions in the IT+TOF group increased from 76.880 (29.933, 156.604) to 126.593 (40.364, 250.589), Z=−3.583, P<0.001. The proportion of relative SNR>1 in the IT+TOF group was 97.7% (43/44), and the proportion of relative SNR>1 in the BP+TOF group was 81.8% (36/44).
    Conclusion TOF can be added to improve the image quality of 18F-FDG PET/CT imaging of intrapulmonary lesions using BP and IT.

     

/

返回文章
返回