比较三种勾画18F-FDG PET/CT图像中肺癌放疗靶区的方法

目的比较三种勾画肺癌18F-FDG PET/CT放疗定位图像肺原发病灶大体肿瘤体积(GTV)边界的方法,寻找适合临床应用的肺癌GTV勾画方法。方法选择14例在18 F-FDG PET/CT的CT图像上边界清晰的肺癌病例,在18 F-FDG PET图像上分别用40%阈值法(GTV40)、SUV=2.5勾画法(GTV2.5)、公式法(GTV公式)对肺癌原发病灶GTV靶区进行勾画,然后与CT肺窗体积(GTVCT)进行比较。结果 14例肺癌的GTV公式与GTVCT的平均差值小于GTV40与GTVCT体积的平均差值,但差异无统计学意义(P=0.185);GTV公式与GTVCT的平均差值明显小于GTV2.5与GTVCT的平均差值(P=0.0289)。14例肺癌中,12例GTV40小于GTV公式,但二者平均GTV差异无统计学意义(P=0.466)。结论公式法及40%阈值法均可用于勾画18F-FDG PET/CT放疗定位图像中肺原发病灶的GTV边界,而以SUV=2.5为边界阈值不适合SUV较低的病灶。

Comparison of three methods for delineation gross tumor volume of primary lung cancer on 18F-FDG PET/CT

Objective To compare the gross tumor volumes (GTVs) resulting from 3 methods used for defining the gross tumor volume on PET/CT in primary lung cancer, in order to find the suitable method for target volume delineation of lung cancer. Methods Fourteen patients with lung cancer whose PET/CT images were clear were selected. Sets of 3 GTVs were generated for each case by 3 methods: Applying a threshold of 40% of the maximum standardized uptake value (GTV₄₀), using an isocontour of SUV=2.5 around the tumor (GTV_2.5), and GTV_function was obtained from phantom measurements. GTV₄₀, GTV_2.5 and GTV_function were compared with GTV_CT, which was determined on CT images of the lung. Results The average difference between GTV_function and GTV_CT was smaller than that between GTV₄₀ and GTV_CT, but there was no statistical difference (P=0.185). The average difference between GTV_function and GTV_CT was much smaller than that between GTV_2.5 and GTV_CT (P=0.0289). In 14 lung cancer patients, GTV₄₀ was smaller than GTV_function in 12 patients, but there was no significant difference between them (P=0.466). Conclusion Both the regressive function and the threshold 40% of maximum intensity methods can be well applied in lung cancer patients. Taking SUV=2.5 as the threshold is not fit to define the lesions with low SUV.