Wang Changyin, Zhao Yanfen, Shen Ying
Department of Nuclear Medicine, Zhongnan Hospital of Wuhan University, Hubei, China.
Department of Applied Mathematics, School of Mathematics and Statistics of Wuhan University, Hubei, China.
Asia Ocean J Nucl Med Biol. 2019 Spring;7(2):160-171. doi: 10.22038/AOJNMB.2019.12734.
The aim of this study was to explore the accuracy of thyroid to background uptake ratio (UR) in the evaluation of Tc-pertechnetate thyroid uptake (TcTU) and establishment of an improved algorithm.
This study was conducted on the thyroid images of 322 patients with thyroid diseases and 67 controls. For the purpose of the study, URs of the images were calculated, and then corrected by standardized thyroid area size to establish a corrected uptake ratio (CUR). Subsequently, the accuracy between UR and CUR was compared.
The results of linear regression using weighted least squares (using TcTU as a dependent variable and CUR, UR, or thyroid area size as independent variables) showed that CUR (t=105.5, P=0.000), UR (t=31.9, P=0.000), and thyroid area size (t=15.9, P=0.000) are influential factors of TcTU. Furthermore, the standardized coefficient of CUR (β=0.983) was obviously higher than those of UR (β=0.851) and thyroid area size (β=0.629). The linear goodness-of-fit between CUR and TcTU (R=0.983) was better than that between UR and TcTU (R=0.851). In addition, the total concordance rate between CUR and TcTU (96.7%) was significantly higher than that between UR and TcTU (83.0%; χ=42.9, P=0.000). Discordance rates of CUR in large thyroid area (1.4% vs. 13.4%, χ=17.0, P=0.000) and small thyroid area (3.3% vs. 42.2 %, χ=44.3, P=0.000), were significantly lower than that of UR. In the abnormal thyroid areas, the discordance rates of UR obviously increased as compared to those of CUR. The UR overestimated the thyroid uptake in small thyroid areas and underestimated it in large thyroid areas.
Based on the findings, CUR is more accurate than UR in measuring TcO- thyroid uptake; accordingly, it is more significant in the diagnosis of thyroid disease.
本研究旨在探讨甲状腺与本底摄取率（UR）在评估高锝酸盐甲状腺摄取（TcTU）中的准确性，并建立一种改进算法。
本研究对322例甲状腺疾病患者和67例对照者的甲状腺图像进行分析。为进行本研究，计算图像的UR，然后通过标准化甲状腺面积大小进行校正，以建立校正摄取率（CUR）。随后，比较UR和CUR之间的准确性。
使用加权最小二乘法进行线性回归的结果（以TcTU作为因变量，CUR、UR或甲状腺面积大小作为自变量）显示，CUR（t = 105.5，P = 0.000）、UR（t = 31.9，P = 0.000）和甲状腺面积大小（t = 15.9，P = 0.000）是TcTU的影响因素。此外，CUR的标准化系数（β = 0.983）明显高于UR（β = 0.851）和甲状腺面积大小（β = 0.629）。CUR与TcTU之间的线性拟合优度（R = 0.983）优于UR与TcTU之间的线性拟合优度（R = 0.851）。此外，CUR与TcTU之间的总符合率（96.7%）显著高于UR与TcTU之间的总符合率（83.0%；χ = 42.9，P = 0.000）。大甲状腺面积中CUR的不一致率（1.4%对13.4%，χ = 17.0，P = 0.000）和小甲状腺面积中CUR的不一致率（3.3%对42.2%，χ = 44.3，P = 0.000）均显著低于UR。在甲状腺异常区域，UR的不一致率与CUR相比明显增加。UR在小甲状腺面积中高估了甲状腺摄取，而在大甲状腺面积中低估了甲状腺摄取。
基于这些发现，CUR在测量高锝酸盐甲状腺摄取方面比UR更准确；因此，它在甲状腺疾病的诊断中更具意义。
