Onishi Hideo, Sakai Takayuki, Shiromoto Osamu, Amijima Hizuru
Program in Health and Welfare, Graduate School of Comprehensive Scientific Research, Prefectural University of Hiroshima, Hiroshima, Japan.
Department of Radiology, Kyushu Rosai Hospital, Japan Labor Health and Welfare Organization, Fukuoka, Japan.
Asia Ocean J Nucl Med Biol. 2018 Spring;6(2):139-148. doi: 10.22038/aojnmb.2018.10638.
The partial volume effect (PVE) of single-photon emission computed tomography (SPECT) on corpus striatum imaging is caused by the underestimation of specific binding ratio (SBR). A large ROI (region of interest) set using the Southampton method is independent of PVE for SBR. The present study aimed to determine the optimal ROI size with contrast and SBR for striatum images and validate the Southampton method using a three-dimensional mathematical cylinder (3D-MAC) phantom.
We used ROIs sizes of 27, 36, 44, 51, 61, 68, and 76 mm for targets with diameters 40, 20, and 10 mm on reference and processed images reconstructed using the 3D-MAC phantom. Contrast values and SBR were compared with the theoretical values to obtain the optimal ROI size.
The contrast values in the ROI with diameters of 51 (target: 40 mm in diameter) and 44 (target: 20 mm in diameter) mm matched the theoretical values. However, this value did not correspond with the 10-mm-diameter target. The SBR matched the theoretical value with an ROI of > 44 mm in the 20-mm-diameter target; but, it was under- and overestimated under any other conditions.
These results suggested that an ROI should be 2-4 folds larger than the target size without PVE, and that the Southampton method was remarkably accurate.
单光子发射计算机断层扫描（SPECT）对纹状体成像的部分容积效应（PVE）是由特异性结合率（SBR）的低估引起的。使用南安普敦方法设置的大感兴趣区（ROI）与SBR的PVE无关。本研究旨在确定纹状体图像具有对比度和SBR时的最佳ROI大小，并使用三维数学圆柱体（3D-MAC）体模验证南安普敦方法。
我们对使用3D-MAC体模重建的参考图像和处理后的图像上直径为40、20和10 mm的目标使用了27、36、44、51、61、68和76 mm的ROI大小。将对比度值和SBR与理论值进行比较以获得最佳ROI大小。
直径为51 mm（目标：直径40 mm）和44 mm（目标：直径20 mm）的ROI中的对比度值与理论值匹配。然而，该值与直径10 mm的目标不对应。在直径20 mm的目标中，SBR在ROI大于44 mm时与理论值匹配；但是，在任何其他条件下，它都被低估和高估。
这些结果表明，在无PVE的情况下，ROI应比目标大小大2至4倍，并且南安普敦方法非常准确。
