基于有效直径和水当量直径计算头颅CT扫描体型特异性剂量估算值的对比研究

Comparison of SSDE values based on effective diameter and water equivalent diameter in head CT examination of adult patients

目的 比较利用扫描范围中心层面有效直径（effective diameter， deff）和水当量直径（water equivalent diameter， dw）计算成人头颅CT扫描体型特异性剂量估算值（size specific dose estimate， SSDE）的差异。方法 连续收集111例头颅CT检查的成人患者。记录每例患者的容积CT剂量指数（volume CT dose index， CTDIvol）。将111例患者的头颅CT SSDE数据进行 SSDEdeff 和 SSDEdw 两种方法计算，记为A组和B组，分别以中心层面deff或dw的体型转换系数（size dependent specific conversion factor，f）与标称CTDIvol乘积计算SSDE。计算A、B两组的体径、f和SSDE。以扫描容积体型特异性剂量估算值（SSDEgross）为参照，评估两种方法计算SSDE的准确度。结果 A、B两组的体径、f差异均有统计学意义（t=47.587，z=-9.242，P〈0.05）。A、B两组的SSDE分别为 （56.20±2.66）和（53.49±2.48） mGy，差异亦有统计学意义（t=-46.687，P〈0.05）。A、B两组中deff与dw比较（r=0.873，R2=0.761）、SSDEdeff与SSDEdw比较（r=0.974，R2=0.949）均为正相关（P〈0.05）。SSDEdeff 与SSDEgross比较（r=0.900，R2=0.809）、SSDEdw与SSDEgross比较（r=0.904，R2=0.817），亦为正相关（P〈0.05）。以SSDEgross为参考，SSDEdeff、SSDEdw估算辐射剂量的平均绝对误差分别为2.34和0.78 mGy，平均相对误差分别为4.38%和1.40%，均方根误差分别为1.17 mGy （2.17%）和1.06 mGy （1.91%）。SSDEdw与SSDEgross 估算剂量数值分布相近，SSDEdeff 和SSDEdw的四分位间距、全距的剂量分布分别为3.22和2.39 mGy、13.65和12.48 mGy。结论 基于中心层面SSDEdw与SSDEgross的一致性，且优于SSDEdeff，因此，SSDEdw可替代SSDEgross作为头颅CT扫描体型特异性剂量估算值的简便指标。

Objective To compare and quantify the differences in size-specific dose estimates （SSDE） obtained by effective diameter and water-equivalent diameter from the central slice of the scan range in head CT examination. Methods A total of 111 consecutive adult patients who underwent head CT examination were enrolled in this study. All of CTDIvol values in the dose report were documented. The dataset was assigned into group A and group B, based on the individual size-dependent conversion factors （f） of effective diameter （deff） and water-equivalent diameter （dw） at the central slice multiplied by normalized volume computed tomography dose index （CTDIvol） respectively. Body size, f and SSDE were calculated. With SSDEgross served as the reference level, the performance of SSDEdeff and SSDEdw was evaluated. Results Statistically significant differences were found in body size （t=47.587,P〈0.05） and f（z=-9.242,P〈0.05） between group A and group B. Statistically significant difference also existed in SSDE （t=-46.687,P〈0.05）, （56.20±2.66） and （53.49±2.48） mGy for group A and group B respectively. Strongly positive correlation was shown in body size （r=0.873, R2=0.761） and SSDE （r=0.974, R2=0.949） between group A and group B（all P〈0.05）. Positive correlation was also found between SSDEdeff and SSDEgross（r=0.900, R2=0.809）, SSDEdw and SSDEgross （r=0.904, R2=0.817, all P〈0.05）. Mean absolute difference was 2.34 and 0.78 mGy, for SSDEdeff vs. SSDEgross and SSDEdw vs. SSDEgross respectively; mean absolute relative difference was 4.38%, 1.40%; root mean square difference was 1.17 mGy （2.17%）, 1.06 mGy （1.91%）. Interquartile range and full range of SSDEdeff and SSDEdw were 3.22 vs. 2.39 mGy, 13.65 vs. 12.48mGy, respectively. A less degree of variation was observed in SSDEdw than that in SSDEdeff. Conclusions SSDEdw values based on the water-equivalent diameter at the central slice of the scan range got better agreement with those derived from all slices, which could serve as a simpler and more valid indicator to represent the average value of size-specific dose estimates of the whole scan range in head CT examination.