CT冠状动脉成像体型特异性剂量估算值的对比研究

Comparative analysis of size-specific dose estimates in coronary computed tomography angiography

目的探讨CT冠状动脉成像基于有效直径(effective diameter,deff)和水等效直径(water equivalent diameter,dw)的体型特异性剂量估算值(size-specific dose estimate, SSDE)差异,并分析原因。方法回顾性收集90例因可疑或确诊冠状动脉粥样硬化性心脏病行CT冠状动脉成像的患者。以deff和dw对应的体型转换系数(size-dependent conversion factor, f)与容积CT剂量指数(volume CT index,CTDIvol)的乘积计算SSDEdeff和SSDEdw,分别记为A组和B组。计算SSDEdeff和SSDEdw的平均绝对相对误差(mean absolute relative difference, MARD),并以多元逐步线性回归分析扫描区域不同组织结构对MARD的影响。结果deff和dw均与体质量指数(body mass index, BMI)呈正相关(r=0.869、0.823,P<0.05)。SSDEdeff、SSDEdw的中位数(四分位数间距)分别为12.34(11.75, 12.98) mGy、13.78(13.02,15.04) mGy,SSDEdeff较SSDEdw约低10.45%,差异具有统计学意义(Z=-8.186,P<0.05)。两者均与BMI和dw呈负相关(r=-0.765、-0.680,-0.701、-0.840, P<0.05)。SSDEdeff与SSDEdw总体平均MARD为11.39%。MARD与BMI相关性无统计学意义(P>0.05),但与deff呈正相关(r=0.251,P<0.05),与dw呈负相关(r=-0.379,P<0.05)。影响MARD的因素中,骨骼和强化心腔血管的CT值(SIhigh)、脂肪面积(Areafat)和CT值(SIfat)、软组织的CT值(SIsoft)未进入回归方程,MARD与含气肺组织的面积(Arealow)和软组织的面积(Areasoft)独立正相关(β=0.634、0.102,P<0.05),与高衰减结构面积(Areahigh)和含气肺组织的CT值(SIlow)呈独立负相关(β=-0.234、-0.343,P<0.05)。结论CT冠状动脉成像的SSDEdeff较SSDEdw约低10.45%,低衰减结构面积占比是影响估算剂量差异的最重要的独立因素。

Objective To investigate the difference of size-specific dose estimates (SSDEs) based on effective diameter (deff) and water equivalent diameter (dw) in coronary computed tomography angiography (CCTA) and explore the causes. Methods A total of 99 patients undergoing CCTA were enrolled in this retrospective study. SSDEs (SSDEdeff and SSDEdw) were calculated in two approaches using deff and dw, respectively. Mean absolute relative difference (MARD) was computed as an index to quantify the consistency of SSDEdeff and SSDEdw. Multivariate stepwise regression analysis was performed to study the factors influencing MARD. Results The values deff and dw were positively correlated with body mass index (BMI)(r=0.869, 0.823, P<0.05). The median (interquartile range) of SSDEdeff and SSDEdw were 12.34(11.75, 12.98) mGy, 13.78(13.02, 15.04) mGy, respectively. SSDEdeff was lower by 10.45% than SSDEdw (Z=-8.186, P<0.05). Both SSDEdeff and SSDEdw were negatively correlated with BMI and dw (r=-0.765,-0.680,-0.701,-0.840, P<0.05). MARD of SSDEdeff and SSDEdw was generally at 11.39%. No statistical significance was found in the correlation of MARD with BMI (r=0.031, P>0.05), however, positive correlation was shown between MARD and deff (r=0.251, P<0.05), but negative correlation for MARD and dw (r=-0.379, P<0.05). With respect to the factors influencing MARD, four variables were included into the regression equation. MARD was positively correlated with the area of both air-filled lungs (Arealow) and soft tissues (Areasoft)(β=0.634, 0.102, P<0.05), and negatively correlated with the area of bone, enhanced cardiac chambers and aorta (Areahigh) and the CT value of air-filled lungs (SIlow)(β=-0.234,-0.343, P<0.05). Conclusions SSDEdeff was approximately 10.45% lower than SSDEdw, which was predominantly influenced by the area of air-filled lungs due to the characteristics of low X-ray attenuation in CCTA.