Lee Choonsik, Lee Choonik, Lee Jai-Ki
Department of Nuclear and Radiological Engineering, University of Florida, Gainesville, FL, USA.
Phys Med Biol. 2006 Nov 7;51(21):N393-402. doi: 10.1088/0031-9155/51/21/N03. Epub 2006 Oct 17.
Distributions of radiation absorbed dose within human anatomy have been estimated through Monte Carlo radiation transport techniques implemented for two different classes of computational anthropomorphic phantoms: (1) mathematical equation-based stylized phantoms and (2) tomographic image-based voxel phantoms. Voxel phantoms constructed from tomographic images of real human anatomy have been actively developed since the late 1980s to overcome the anatomical approximations necessary with stylized phantoms, which themselves have been utilized since the mid 1960s. However, revisions of stylized phantoms have also been pursued in parallel to the development of voxel phantoms since voxel phantoms (1) are initially restricted to the individual-specific anatomy of the person originally imaged, (2) must be restructured on an organ-by-organ basis to conform to reference individual anatomy and (3) cannot easily represent very fine anatomical structures and tissue layers that are thinner than the voxel dimensions of the overall phantom. Although efforts have been made to improve the anatomic realism of stylized phantoms, most of these efforts have been limited to attempts to alter internal organ structures. Aside from the internal organs, the exterior shapes, and especially the arm structures, of stylized phantoms are also far from realistic descriptions of human anatomy, and may cause dosimetry errors in the calculation of organ-absorbed doses for external irradiation scenarios. The present study was intended to highlight the need to revise the existing arm structure within stylized phantoms by comparing organ doses of stylized adult phantoms with those from three adult voxel phantoms in the lateral photon irradiation geometry. The representative stylized phantom, the adult phantom of the Oak Ridge National Laboratory (ORNL) series and two adult male voxel phantoms, KTMAN-2 and VOXTISS8, were employed for Monte Carlo dose calculation, and data from another voxel phantom, VIP-Man, were obtained from literature sources. The absorbed doses for lungs, oesophagus, liver and kidneys that could be affected by arm structures in the lateral irradiation geometry were obtained for both classes of phantoms in lateral monoenergetic photon irradiation geometries. As expected, those organs in the ORNL phantoms received apparently higher absorbed doses than those in the voxel phantoms. The overestimation is mainly attributed to the relatively poor representation of the arm structure in the ORNL phantom in which the arm bones are embedded within the regions describing the phantom's torso. The results of this study suggest that the overestimation of organ doses, due to unrealistic arm representation, should be taken into account when stylized phantoms are employed for equivalent or effective dose estimates, especially in the case of an irradiation scenario with dominating lateral exposure. For such a reason, the stylized phantom arm structure definition should be revised in order to obtain more realistic evaluations.
通过为两类不同的计算人体模型实施的蒙特卡罗辐射传输技术,已估算出人体解剖结构内的辐射吸收剂量分布:(1)基于数学方程的简化模型和(2)基于断层图像的体素模型。自20世纪80年代末以来,人们积极开发由真实人体解剖结构的断层图像构建的体素模型,以克服简化模型所必需的解剖学近似,简化模型自20世纪60年代中期就已开始使用。然而,自体素模型发展以来,简化模型也在并行修订,因为体素模型(1)最初仅限于最初成像者的个体特定解剖结构,(2)必须逐个器官地进行重构以符合参考个体解剖结构,(3)不易表示比整体模型体素尺寸更薄的非常精细的解剖结构和组织层。尽管已努力提高简化模型的解剖学逼真度,但这些努力大多仅限于尝试改变内部器官结构。除了内部器官,简化模型的外部形状,尤其是手臂结构,也远非对人体解剖结构的现实描述,并且可能在外部照射场景下器官吸收剂量的计算中导致剂量测定误差。本研究旨在通过比较简化成人模型与三个成人体素模型在侧向光子照射几何条件下的器官剂量,强调修订简化模型中现有手臂结构的必要性。选用具有代表性的简化模型、橡树岭国家实验室(ORNL)系列的成人模型以及两个成人男性体素模型KTMAN - 2和VOXTISS8进行蒙特卡罗剂量计算,并从文献来源获取另一个体素模型VIP - Man的数据。在侧向单能光子照射几何条件下,针对两类模型获取了在侧向照射几何条件下可能受手臂结构影响的肺、食管、肝脏和肾脏的吸收剂量。正如预期的那样,ORNL模型中的那些器官所接受的吸收剂量明显高于体素模型中的器官。这种高估主要归因于ORNL模型中手臂结构的表示相对较差,其中手臂骨骼嵌入描述模型躯干的区域内。本研究结果表明,当使用简化模型进行等效或有效剂量估计时,尤其是在侧向照射占主导的照射场景中,应考虑由于手臂表示不现实而导致的器官剂量高估。出于这个原因,应修订简化模型手臂结构的定义,以获得更现实的评估。