部分容积效应校正技术在PET/CT诊断单发肺结节中的应用  

Improving diagnostic performance of PET/CT in single lung nodule with partial volume effect correction

作　　者：于丽娟[1] 吴文凯[2] 赵周社[3] 

机构地区：[1]哈尔滨医科大学附属肿瘤医院PET/CT中心,150081 [2]中国医学科学院,中国协和医科大学肿瘤医院核医学科 [3]GE医疗系统集团

出　　处：《中华核医学杂志》2007年第5期306-308,共3页Chinese Journal of Nuclear Medicine

摘　　要：目的采用部分容积效应(PVE)校正技术提高^(18)F-FDG PET/CT 对单发肺结节诊断的准确性。方法采用 Jaszczak 模型,在模型内放置直径分别为24.8,19.7,15.6,12.4,9.85,7.82和4.9 mm 的"热区"小球,每个小球内含^(18)F-FDG 24 kBq/ml,进行 PET/CT 模型扫描,获得恢复系数(RC)后,选择 PET/CT 显像阳性,病灶直径<25 mm 且有病理检查结果的24例单发肺结节的患者资料进行分析。以最大 SUV(SUV_(max)≥2.5为标准,与病理检查结果进行对比研究。结果二维(2D)扫描模式的 RC 值大于三维(3D)扫描模式的 RC 值,表明同等条件下2D 扫描模式 PVE 比3D 扫描模式小。PET/CT 的SUV_(校正)降低了对单发肺结节诊断的假阴性,使其灵敏度和准确性分别由原来的62.5%,62.5%提高为100.0%,87.5%。结论 SUV_(校正)的应用明显提高了 PET/CT 对肺结节诊断的灵敏度和准确性,但是对肺结节诊断的特异性并无变化。采用 PVE 校正技术明显提高了 PET/CT 诊断直径<25 mm 的单发肺结节的准确性。Objective The aim of this prospective study was to improve diagnosis veracity of PET/ CT in single lung nodule with partial volume effect （PVE） correction. Methods Jaszczak model with several small ＂hot＂ balls of different diameter （24.8, 19.7, 15.6, 12.4, 9.85, 7.82 and 4.9 mm） filling with ISF-FDG （24 kBq/ml） were set and scanned on PET/CT to derive recover coefficient （RC）. Based on RC obtained from phantom study, maximum SUV （SUVmax） without PVE correction and SUVmax with PVE correction from both 2 dimension （D） and 3D mode PET scan on 24 cases with single lung nodule （ 〈 25 mm and confirmed pathologically） were calculated and analyzed. Results The RC value at 2D scan was larger than that at 3D mode, suggesting less PVE in 2D scan mode. In 24 patients with single lung nodule, the diagnostic sensitivity, specificity and accuracy using SUVmax without PVE correction and SUVmax with PVE correction were 62.5%, 5/8, 62.5% and 100.0%, 5/8, 87.5% respectively. Conclusion The application of PVE correction could improve the sensitivity and accuracy, but not the specificity, of PET/CT in single lung nodule （ 〈25 mm） diagnosis.

关 键 词：肺肿瘤 体层摄影术 发射型计算机 体层摄影术 X线计算机 部分容积效应 

分 类 号：R817.4[医药卫生—影像医学与核医学]