摘要
目的 探讨肺部肿瘤1 8F FDGPET显像的标准摄取值 (SUV)良恶性分界点与感兴趣区(ROI)大小的关系。方法 12 6例肺部肿瘤患者术前行1 8F FDGPET检查 ,二维采集 ,常规条件重建图像。以SUV最大值的百分比 (10 0 %~ 5 0 % )为等高线值画 6个ROI,得各SUV均值。计算与准确性最大值、Youden’s指数最大值及灵敏度与特异性交点所对应的SUV(典型分界点 )。结果 ①SUV均值随ROI增大而减小 ,良恶性间差异均有显著性 (P <0 0 1) ;②典型分界点值均随ROI增大而减小 ;③对分界点 2和 2 5 ,其灵敏度随ROI增大而减小 ,特异性则相反。对各典型分界点 ,其灵敏度和特异性基本不随ROI大小变化。结论 SUV良恶性分界点与ROI的大小有关 ,不考虑ROI大小而均以SUV =2或 2 5固定值作为良恶性分界点不合适。
Objective To investigate the relationship between the size of the region of interest (ROI) and the standardized uptake value (SUV) cut-off for distinguishing the benign from the malignant lung tumor with 18 F-fluorodeoxyglucose (FDG) PET. Methods 18 F-FDG PET images of 126 patients with various lung tumors were studied retrospectively. Six ROIs were drown from each tumor with contour line at 100%~50% of the maximum SUV, and the average SUVs were calculated from the ROIs. Three typical diagnostic cut-off points were determined based on the maximum accuracy, the maximum Youden's index and the intersections of sensitivity and specificity. Results The SUV increased as the ROI size increased, and there were significant differences between SUVs of benign and malignant lung lesions (P<0.01). The typical diagnostic cut-off values decreased as the ROI size increased. For SUV cut-off of 2 and 2.5, the sensitivity decreased as the ROI size increased, and the specificity fluctuated oppositely to the sensitivity. For the typical diagnostic cut-off points, the sensitivity and specificity basically did not vary with the ROI size. Conclusions The SUV cut-off is correlated with the ROI size. It is incorrect to set 2 or 2.5 as the SUV cut-off without regarding to the ROI size.
出处
《中华核医学杂志》
CAS
CSCD
北大核心
2004年第3期174-176,共3页
Chinese Journal of Nuclear Medicine