Osawa Atsushi, Miwa Kenta, Wagatsuma Kei, Takiguchi Tomohiro, Tamura Shintaro, Akimoto Kenta
Department of Radiology Nuclear Medicine, Cancer Institute Hospital of Japanese Foundation for Cancer Research.
Nihon Hoshasen Gijutsu Gakkai Zasshi. 2012;68(12):1600-7. doi: 10.6009/jjrt.2012_jsrt_68.12.1600.
The image quality in (18)FDG PET/CT often degrades as the body size increases. The purpose of this study was to evaluate the relationship between image quality and the body size using original phantoms of variable cross-sectional areas in PET/CT. We produced five water phantoms with different cross-sectional areas. The long axis of phantom was 925 mm, and the cross-sectional area was from 324 to 1189 cm(2). These phantoms with the sphere (diameter 10 mm) were filled with (18)F-FDG solution. The radioactivity concentration of background in the phantom was 1.37, 2.73, 4.09 and 5.46 kBq/mL. The scanning duration was 30 min in list mode acquisition for each measurement. Background variability (N(10 mm)), noise equivalent count rates (NECR(phantom)), hot sphere contrast (Q(H,10 mm)) as physical evaluation and visual score of sphere detection were measured, respectively. The relationship between image quality and the various cross-sectional areas was also analyzed under the above-mentioned conditions. As cross-sectional area increased, NECR(phantom) progressively decreased. Furthermore, as cross-sectional area increased, N(10 mm) increased and Q(H,10 mm) decreased. Image quality became degraded as body weight increased because noise and contrast contributed to image quality. The visual score of sphere detection deteriorated in high background radioactivity concentration because a false positive detection in cross-sectional area of the phantom increased. However, additional increases in scanning periods could improve the visual score. We assessed tendencies in the relationship between image quality and body size in PET/CT. Our results showed that time adjustment was more effective than dose adjustment for stable image quality of heavier patients in terms of the large cross-sectional area.
在（18）FDG PET/CT中，图像质量常常会随着体型增大而下降。本研究的目的是使用PET/CT中具有不同横截面积的原始体模来评估图像质量与体型之间的关系。我们制作了五个具有不同横截面积的水体模。体模的长轴为925毫米，横截面积从324至1189平方厘米。这些带有球体（直径10毫米）的体模被注入（18）F-FDG溶液。体模中背景的放射性浓度分别为1.37、2.73、4.09和5.46 kBq/mL。每次测量在列表模式采集中扫描持续时间为30分钟。分别测量了作为物理评估的背景变异性（N（10毫米））、噪声等效计数率（NECR（体模））、热球对比度（Q（H，10毫米））以及球体检测的视觉评分。在上述条件下还分析了图像质量与各种横截面积之间的关系。随着横截面积增加，NECR（体模）逐渐降低。此外，随着横截面积增加，N（10毫米）增加而Q（H，10毫米）降低。由于噪声和对比度对图像质量有影响，图像质量随着体重增加而下降。在高背景放射性浓度下，球体检测的视觉评分变差，因为体模横截面积中的假阳性检测增加。然而，扫描时间的额外增加可以改善视觉评分。我们评估了PET/CT中图像质量与体型之间关系的趋势。我们的结果表明，就大横截面积而言，对于体型较重的患者，时间调整比剂量调整对稳定图像质量更有效。
