Abstract Accurate superficial dose calculation is of major importance because of the skin toxicity in radiotherapy, especially within the initial 2 mm depth being considered more clinically relevant. The aim of this study is to evaluate superficial dose calculation accuracy of four commonly used algorithms in commercially available treatment planning systems (TPS) by Monte Carlo (MC) simulation and film measurements. The superficial dose in a simple geometrical phantom with size of 30 cm×30 cm×30 cm was calculated by PBC (Pencil Beam Convolution), AAA (Analytical Anisotropic Algorithm), AXB (Acuros XB) in Eclipse system and CCC (Collapsed Cone Convolution) in Raystation system under the conditions of source to surface distance (SSD) of 100 cm and field size (FS) of 10×10 cm 2 . EGSnrc (BEAMnrc/DOSXYZnrc) program was performed to simulate the central axis dose distribution of Varian Trilogy accelerator, combined with measurements of superficial dose distribution by an extrapolation method of multilayer radiochromic films, to estimate the dose calculation accuracy of four algorithms in the superficial region which was recommended in detail by the ICRU (International Commission on Radiation Units and Measurement) and the ICRP (International Commission on Radiological Protection). In superficial region, good agreement was achieved between MC simulation and film extrapolation method, with the mean differences less than 1%, 2% and 5% for 0°, 30° and 60°, respectively. The relative skin dose errors were 0.84%, 1.88% and 3.90%; the mean dose discrepancies (0°, 30° and 60°) between each of four algorithms and MC simulation were (2.41±1.55%, 3.11±2.40%, and 1.53±1.05%), (3.09±3.00%, 3.10±3.01%, and 3.77±3.59%), (3.16±1.50%, 8.70±2.84%, and 18.20±4.10%) and (14.45±4.66%, 10.74±4.54%, and 3.34±3.26%) for AXB, CCC, AAA and PBC respectively. Monte Carlo simulation verified the feasibility of the superficial dose measurements by multilayer Gafchromic films. And the rank of superficial dose calculation accuracy of four algorithms was AXB>CCC>AAA>PBC. Care should be taken when using the AAA and PBC algorithms in the superficial dose calculation.

中文翻译：

---

四种剂量计算算法的表面剂量评估

摘要 由于放射治疗中的皮肤毒性，尤其是在最初的 2 毫米深度内被认为更具有临床相关性，因此准确的表面剂量计算非常重要。本研究的目的是通过蒙特卡罗 (MC) 模拟和胶片测量来评估市售治疗计划系统 (TPS) 中四种常用算法的表面剂量计算精度。通过PBC（铅笔束卷积）、AAA（分析各向异性算法）、Eclipse系统中的AXB（Acuros XB）和CCC（Collapsed Cone Convolution）计算大小为30 cm×30 cm×30 cm的简单几何体模的表面剂量) 在光源到地表距离 (SSD) 为 100 cm 和场尺寸 (FS) 为 10×10 cm 2 的条件下，在 Raystation 系统中。使用EGSnrc（BEAMnrc/DOSXYZnrc）程序模拟Varian Trilogy加速器的中轴剂量分布，结合多层放射致变色薄膜外推法测量表面剂量分布，估计四种算法在表面区域的剂量计算精度这是 ICRU（国际辐射单位和测量委员会）和 ICRP（国际辐射防护委员会）详细推荐的。在表面区域，MC模拟和薄膜外推法之间取得了很好的一致性，0°、30°和60°的平均差异分别小于1%、2%和5%。相对皮肤剂量误差分别为0.84%、1.88%和3.90%；平均剂量差异（0°，4 种算法和 MC 仿真之间的 30° 和 60°）分别为（2.41±1.55%、3.11±2.40% 和 1.53±1.05%）、（3.09±3.00%、3.10±3.01% 和 3.77±3.59%） , AXB、CCC、AAA 和 PBC 分别为 (3.16±1.50%、8.70±2.84% 和 18.20±4.10%) 和 (14.45±4.66%、10.74±4.54% 和 3.34±3.26%)。蒙特卡罗模拟验证了多层 Gafchromic 薄膜表面剂量测量的可行性。4种算法的表观剂量计算精度排序为AXB>CCC>AAA>PBC。在表面剂量计算中使用 AAA 和 PBC 算法时应小心。蒙特卡罗模拟验证了多层 Gafchromic 薄膜表面剂量测量的可行性。4种算法的表观剂量计算精度等级为AXB>CCC>AAA>PBC。在表面剂量计算中使用 AAA 和 PBC 算法时应小心。Monte Carlo 模拟验证了多层 Gafchromic 薄膜表面剂量测量的可行性。4种算法的表观剂量计算精度排序为AXB>CCC>AAA>PBC。在表面剂量计算中使用 AAA 和 PBC 算法时应小心。