Division of Medical Radiation Physics, University of Stockholm at Karolinska Hospital, SE-171 76 Stockholm, Sweden.
Phys Med Biol. 2012 Apr 21;57(8):2117-36. doi: 10.1088/0031-9155/57/8/2117. Epub 2012 Mar 27.
A systematic analysis of the available data has been carried out for mass energy-absorption coefficients and their ratios for air, graphite and water for photon energies between 1 keV and 2 MeV, using representative kilovoltage x-ray spectra for mammography and diagnostic radiology below 100 kV, and for ¹⁹²Ir and ⁶⁰Co gamma-ray spectra. The aim of this work was to establish 'an envelope of uncertainty' based on the spread of the available data. Type A uncertainties were determined from the results of Monte Carlo (MC) calculations with the PENELOPE and EGSnrc systems, yielding mean values for µ(en)/ρ with a given statistical standard uncertainty. Type B estimates were based on two groupings. The first grouping consisted of MC calculations based on a similar implementation but using different data and/or approximations. The second grouping was formed by various datasets, obtained by different authors or methods using the same or different basic data, and with different implementations (analytical, MC-based, or a combination of the two); these datasets were the compilations of NIST, Hubbell, Johns-Cunningham, Attix and Higgins, plus MC calculations with PENELOPE and EGSnrc. The combined standard uncertainty, u(c), for the µ(en)/ρ values for the mammography x-ray spectra is 2.5%, decreasing gradually to 1.6% for kilovoltage x-ray spectra up to 100 kV. For ⁶⁰Co and ¹⁹²Ir, u(c) is approximately 0.1%. The Type B uncertainty analysis for the ratios of µ(en)/ρ values includes four methods of analysis and concludes that for the present data the assumption that the data interval represents 95% confidence limits is a good compromise. For the mammography x-ray spectra, the combined standard uncertainties of (µ(en)/ρ)(graphite,air) and (µ(en)/ρ)(graphite,water) are 1.5%, and 0.5% for (µ(en)/ρ)(water,air), decreasing gradually down to u(c) = 0.1% for the three µ(en)/ρ ratios for the gamma-ray spectra. The present estimates are shown to coincide well with those of Hubbell (1977 Rad. Res. 70 58-81), except for the lowest energy range (radiodiagnostic) where it is concluded that current databases and their systematic analysis represent an improvement over the older Hubbell estimations. The results for (µ(en)/ρ)(graphite,air) for the gamma-ray dosimetry range are moderately higher than those of Seltzer and Bergstrom (2005 private communication).
已经对空气、石墨和水的质量能量吸收系数及其比值进行了系统分析,这些系数和比值适用于光子能量在 1keV 至 2MeV 之间,使用的是代表乳腺摄影和诊断放射学低于 100kV 的千伏 X 射线谱以及 ¹⁹²Ir 和 ⁶⁰Co 伽马射线谱的代表性谱。这项工作的目的是根据现有数据的分散程度建立一个“不确定度范围”。A 类不确定度是根据 PENELOPE 和 EGSnrc 系统的蒙特卡罗(MC)计算结果确定的,给出了给定统计标准不确定度的µ(en)/ρ 平均值。B 类估计值基于两个分组。第一个分组由基于类似实现但使用不同数据和/或近似值的 MC 计算组成。第二个分组由不同作者或方法使用相同或不同基本数据获得的各种数据集组成,并且实现方式不同(分析、基于 MC 或两者的组合);这些数据集是 NIST、Hubbell、Johns-Cunningham、Attix 和 Higgins 的汇编,以及使用 PENELOPE 和 EGSnrc 的 MC 计算。乳腺 X 射线谱的µ(en)/ρ 值的组合标准不确定度 u(c)为 2.5%,逐渐降低至 100kV 以下千伏 X 射线谱的 1.6%。对于 ⁶⁰Co 和 ¹⁹²Ir,u(c)约为 0.1%。µ(en)/ρ 值比的 B 类不确定度分析包括四种分析方法,并得出结论,对于当前数据,假设数据间隔代表 95%置信限是一个很好的折衷方案。对于乳腺 X 射线谱,(µ(en)/ρ)(石墨,空气)和(µ(en)/ρ)(石墨,水)的组合标准不确定度分别为 1.5%和 0.5%,对于(µ(en)/ρ)(水,空气)为 0.5%,对于伽马射线谱的三个µ(en)/ρ 比,逐渐降低至 u(c)=0.1%。本研究的结果与 Hubbell(1977 Rad. Res. 70 58-81)的结果吻合良好,除了最低能量范围(放射诊断)外,本研究认为当前的数据库及其系统分析比旧的 Hubbell 估计有所改进。伽马射线剂量测定范围内(µ(en)/ρ)(石墨,空气)的结果略高于 Seltzer 和 Bergstrom(2005 年私人交流)的结果。