三种测量方法对肺部不同密度小结节CT体积测量的观察  

CT volume measurement for pulmonary nodules with different densities: observation on three different methods

作　　者：柳学国[1] 章作铨[1] 王颖[1] 李坤炜[1] 高洁冰[1] 李颖勤[1] 李晓黎[1] 

机构地区：[1]中山大学附属第五医院放射科,珠海519000

出　　处：《中华放射学杂志》2005年第1期21-28,共8页Chinese Journal of Radiology

基　　金：广东省医学科技研究项目 (A2 0 0 2 675 ) ;珠海市科技研究项目 (2 0 0 2 5 0 )

摘　　要：目的　评价肺部不同密度小结节CT体积测量 3种方法 (固定阈值法、可变阈值法和部分容积法 )的可重复性。方法　选择 6个 3种不同密度 (实性、部分实性、非实性 )小结节 ,由 5位医生独立采用上述 3种体积测量方法分别测量计算体积 5次 ,ANOVA方差分析医生之间各种方法对不同密度结节体积测量的可重复性。结果　固定阈值法对于实性结节的体积测量可重复性好 (P >0 0 5 ) ,误差小于± 2 %;部分实性结节 (CT值 - 70 0～ 4 0HU)仅识别其中实性部分 ;非实性结节 (CT值- 80 0～ - 4 0 0HU)识别不出来。可变阈值法选择适当的CT阈值范围 ,标记测量的范围与病变重叠良好 ,主要适用于部分实性结节和非实性结节的体积测量 ,但可重复性不好 (P <0 0 5 ) ,误差范围 <±1 0 %(99%的可信区间 - 1 2 %～ 36 %)。通过CT阈值与体积变化关系的实验发现 :单纯实性结节CT值的阈值范围为 - 4 0 0～ 2 0 0HU ,单纯非实性结节为 - 80 0～ 0HU ,部分实性结节为 - 80 0～ 2 0 0HU。部分容积法对于密度均匀的实性结节和非实性结节 ,医生之间有较好的可重复性 (P >0 0 5 ) ,但对于其他结节医生之间的可重复性不佳 (P <0 0 5 ) ,误差范围 - 2 0 %～ 4 0 %(99%的可信区间 - 30 %～ 5 0 %)。各种体积测量方法之间的体积无可比性 ,误?Objective To evaluate the reproducibility of three different CT volume measurement methods including fixed threshold method (FTM), variable threshold method (VTM), and partial volume method (PVM) in pulmonary nodules with different densities.Methods Six nodules (3 different density categories: solid, part solid, and non-solid) were chosen.5 radiologists measured the volume independently with three methods in 5 separate times.One-way ANOVA test was used to analyze the reproducibility of doctors in each volume measurement method for different density nodules.Results FTM had very good reproducibility in solid nodule volume measurement (P>0.05), and the error was less than ±2%.The reproducibility of FTM was very poor for volume measurement in part solid nodule and non-solid nodule, it could only evaluate solid part of part solid nodule and could not evaluate the non-solid part of nodule ( CT value -800 to -400 HU).VTM could mark different density nodule very well through adjusting CT value threshold, but the reproducibility was very poor between doctors (P<0.05), and the error was less than ±10% (99% CI: -12% to 36%).In order to choose suitable CT value threshold for volume measurement, a test was made to draw a CT value threshold and volume correlation curve of different density nodules.The results showed that CT value threshold was -400 to 200 HU for solid nodule, -800 to 200 HU for part solid nodule, and -800 to 0 HU for non-solid nodule, respectively.PVM had good reproducibility in solid and homogeneous non-solid nodule volume measurement only (P<0.05), but the error range to part solid nodule or inhomogenous non-solid nodule was from -20% to 40% (99% CI: -30% to 50%).The volume of the same nodule in different methods were not comparable, the error range could be from 8% to 10 000%. Conclusion FTM is a good method for small solid nodule volume measurement.VTM and PVM are not reproducible in clinical nodule volume measurement.The improvement direction might be as follows: enlargement of the CT value

关 键 词：结节 可重复性 体积测量 临床应用 CT 医生 肺部 阈值法 均匀 域值 

分 类 号：R816.4[医药卫生—放射医学]