Jarry G, DeMarco J J, Beifuss U, Cagnon C H, McNitt-Gray M F
McGill University, Medical Physics Unit, Montreal General Hospital, Montreal, Canada.
Phys Med Biol. 2003 Aug 21;48(16):2645-63. doi: 10.1088/0031-9155/48/16/306.
The purpose of this work is to develop and test a method to estimate the relative and absolute absorbed radiation dose from axial and spiral CT scans using a Monte Carlo approach. Initial testing was done in phantoms and preliminary results were obtained from a standard mathematical anthropomorphic model (MIRD V) and voxelized patient data. To accomplish this we have modified a general purpose Monte Carlo transport code (MCNP4B) to simulate the CT x-ray source and movement, and then to calculate absorbed radiation dose in desired objects. The movement of the source in either axial or spiral modes was modelled explicitly while the CT system components were modelled using published information about x-ray spectra as well as information provided by the manufacturer. Simulations were performed for single axial scans using the head and body computed tomography dose index (CTDI) polymethylmethacrylate phantoms at both central and peripheral positions for all available beam energies and slice thicknesses. For comparison, corresponding physical measurements of CTDI in phantom were made with an ion chamber. To obtain absolute dose values, simulations and measurements were performed in air at the scanner isocentre for each beam energy. To extend the verification, the CT scanner model was applied to the MIRD V model and compared with published results using similar technical factors. After verification of the model, the generalized source was simulated and applied to voxelized models of patient anatomy. The simulated and measured absolute dose data in phantom agreed to within 2% for the head phantom and within 4% for the body phantom at 120 and 140 kVp; this extends to 8% for the head and 9% for the body phantom across all available beam energies and positions. For the head phantom, the simulated and measured absolute dose data agree to within 2% across all slice thicknesses at 120 kVp. Our results in the MIRD phantom agree within 11% of all the different organ dose values published by the UK's ImPACT group for a scan using an equivalent scanner, kVp, collimation, pitch and mAs. The CT source model was shown to calculate both a relative and absolute radiation dose distribution throughout the entire volume in a patient-specific matrix geometry. Results of initial testing are promising and application to patient models was shown to be feasible.
本研究的目的是开发并测试一种使用蒙特卡罗方法估算轴向和螺旋CT扫描相对及绝对吸收辐射剂量的方法。初始测试在体模中进行，并从标准数学人体模型（MIRD V）和体素化患者数据中获得了初步结果。为实现这一目标，我们修改了通用蒙特卡罗传输代码（MCNP4B），以模拟CT X射线源和运动，然后计算目标物体中的吸收辐射剂量。在显式模拟轴向或螺旋模式下源的运动时，使用已发表的关于X射线光谱的信息以及制造商提供的信息对CT系统组件进行建模。针对所有可用束能量和层厚，在中心和周边位置使用头部和体部计算机断层扫描剂量指数（CTDI）聚甲基丙烯酸甲酯体模对单轴向扫描进行了模拟。作为对比，使用电离室对体模中的CTDI进行了相应的物理测量。为获得绝对剂量值，针对每种束能量在扫描器等中心的空气中进行了模拟和测量。为扩展验证范围，将CT扫描仪模型应用于MIRD V模型，并与使用类似技术参数发表的结果进行比较。在对模型进行验证后，对广义源进行了模拟并应用于患者解剖结构的体素化模型。在120和140 kVp时，体模中模拟和测量的绝对剂量数据在头部体模中相差在2%以内，在体部体模中相差在4%以内；在所有可用束能量和位置上，头部体模的差异扩展到8%，体部体模的差异扩展到9%。对于头部体模，在120 kVp时，所有层厚下模拟和测量的绝对剂量数据相差在2%以内。我们在MIRD体模中的结果与英国ImPACT小组发表的使用等效扫描仪、kVp、准直、螺距和mAs进行扫描的所有不同器官剂量值相差在11%以内。CT源模型能够在患者特异性矩阵几何结构中计算整个体积内的相对和绝对辐射剂量分布。初始测试结果很有前景，并且已证明将其应用于患者模型是可行的。
