目的 定量研究不同扫描参数组合导致的医科达XVI锥形束CT辐射剂量变化,为评估影像引导放疗中成像剂量的参数依赖性提供数学模型。方法 基于Versa HD加速器XVI,利用PTW 30 009千伏电离室和UNIDOS webline静电计,在PTW标准CT剂量指数(CTDI)体部模体中,测量标准扫描参数及多种扫描电压(kVp)、管电流(mA)组合下的模体内各点比释动能,并计算加权CTDIw。利用SigmaPlot 10.0软件将测量结果拟合为以管电流和/或扫描电压为变量的模型。结果 标准扫描参数下,瓦里安OBI锥形束CT的CTDIw值仅为医科达XVI的11.23%(胸部参数)和9.15%(盆腔参数)。在标准及其余4个扫描电压条件下,模体中心和外周各点比释动能与管电流均呈现线性正比关系,但斜率a值差异较大(0.479-6.679),主要受扫描电压值、模体测量位置、剂量描述方法等因素影响。模体内各点剂量和CTDIw值均可拟合为以扫描电压为变量的非线性经验公式(R2〉0.997),各系数差异有统计学意义(P〈0.05)。同时改变管电流和扫描电压对模体中心点剂量的影响可以表述为mGy=(5.917-0.197×kVp+0.002×kVp2-5.063×10-6×kVp3)×mA。结论 医科达XVI锥形束CT剂量显著依赖于扫描参数,数学模型可用于快速准确描述其变化特征。
Objective To quantify the correlations between Elekta XVI cone beam CT dose and various scanning protocols, providing mathematical models to assess the protocol-dependency of imaging dose during image guided radiotherapy. Methods Based on standard protocols and various combinations of kVp and mA on an XVI mounted on an Elekta Versa HD accelerator, the air KERMA was measured at various positions in a standard PTW CTDI body phantom using calibrated PTW 30009 kV chamber and UNIDOS webline electrometer. Weighted CT dose index (CTDIw) was computed thereafter. SigmaPlot 10.0 was used to fit the measurements against mA and/or kVp yielding empirical functions. Results Under standard protocols, the CTDIw of Varian OBI was only 11.23% (chest) and 9.15% (pelvis) of Elekta XVI. Using the default and other 4 investigated kVp values, the central and peripheral KERMA were both proportional to mA, and yet the slope value a varied dramatically from 0.479 to 6.679. Major affecting factors included kVp settings, measurement locations, and dosimetric metrics, etc. None linear regressions were used to fit kVp against KERMA at various locations and CTDIw (R2〉0.997). The differences between all coefficients were statistically significant (P〈0.05). The impact of changing both mA and kVp on the dose to phantom center can be described as mGy=(5.917 -0.197×kVp+0.002×kVp2-5.063×10-6×kVp3)×mA. Conclusions Imaging dose of Elekta XVI is strongly dependent on scanning parameters. The proposed mathematical models can be used as efficient and robust indicators of such dependency.