Department of Radiological Science, Faculty of Health Science, Junshin Gakuen University, 1-1-1 Chikushigaoka, Minami-ku, Fukuoka, 815-8510, Japan.
Division of Radiology, Department of Medical Technology, Kyushu University Hospital, 3-1-1 Maidashi, Higashi-ku, Fukuoka, 812-8582, Japan.
Ann Nucl Med. 2020 Oct;34(10):725-735. doi: 10.1007/s12149-020-01496-1. Epub 2020 Jul 3.
The aim of this study was to evaluate the image quality and the quantification accuracy of Biograph Vision PET/CT scanner as a SiPM-PET in comparison to the conventional PMT-PET, Biograph mCT PET/CT scanner.
This study consisted of a phantom study and a retrospective clinical analysis where patients underwent F-FDG PET/CT in both PET systems. The body phantom of the NEMA IEC with 10-37 mm diameter spheres were filled with an F-FDG solution. The root mean square error (RMSE) of SUV, the detectability of 10-mm sphere, NEC, the background variability (N) and the contrast-noise-ratio (Q/N) were calculated based on the phantom analysis. We also examined the quality of the acquired clinical images using the NEC, NEC, SNR, SUV and SUV.
In the phantom study on Vision scanner, RMSE was relatively lower when the iteration number was 2, 3 or 4. To satisfy a visual score of 1.5 and the reference range of Q/N, a 60-s or longer acquisition was required. Our clinical findings show that NEC averaged 17.4 ± 1.72 Mcounts/m in mCT and 29.1 ± 2.83 Mcounts/m in Vision. Furthermore, NEC averaged 0.29 ± 0.05 kcounts/cm in mCT and 0.53 ± 0.09 kcounts/cm in Vision, respectively, whereas SNR averaged 14.6 ± 3.77% in mCT and 21.3 ± 1.69% in Vision (P = 0.0156), respectively. Finally, SUV averaged 2.82 ± 0.28 and 2.55 ± 0.30, SUV ranged 1.6-17.6 and 1.9-22.9 in mCT and Vision, respectively.
SiPM-PET/CT provides superior image quality and quantification accuracy compared to PMT-PET/CT.
本研究旨在评估 SiPM-PET/CT 扫描仪（Biograph Vision PET/CT 扫描仪）作为 SiPM-PET 与传统 PMT-PET、Biograph mCT PET/CT 扫描仪相比的图像质量和定量准确性。
本研究包括一项体模研究和一项回顾性临床分析，其中患者在两种 PET 系统中均接受了 F-FDG PET/CT 检查。NEMA IEC 体模用 10-37mm 直径球体填充 F-FDG 溶液。基于体模分析，计算 SUV 的均方根误差（RMSE）、10mm 球体的可探测性、NEC、背景变化（N）和对比噪声比（Q/N）。我们还使用 NEC、NEC、SNR、SUV 和 SUV 检查了获得的临床图像的质量。
在 Vision 扫描仪的体模研究中，当迭代次数为 2、3 或 4 时，RMSE 相对较低。为了满足视觉评分 1.5 和 Q/N 的参考范围，需要进行 60s 或更长时间的采集。我们的临床发现表明，NEC 在 mCT 中平均为 17.4±1.72Mcounts/m，在 Vision 中为 29.1±2.83Mcounts/m。此外，NEC 在 mCT 中平均为 0.29±0.05kcounts/cm，在 Vision 中为 0.53±0.09kcounts/cm，而 SNR 在 mCT 中平均为 14.6±3.77%，在 Vision 中为 21.3±1.69%（P=0.0156）。最后，SUV 在 mCT 中平均为 2.82±0.28，在 Vision 中为 2.55±0.30，SUV 在 mCT 中为 1.6-17.6，在 Vision 中为 1.9-22.9。
与 PMT-PET/CT 相比，SiPM-PET/CT 提供了更高的图像质量和定量准确性。
