In previous works, we showed that incorporating individual airways as organs-at-risk (OARs) in the treatment of lung stereotactic ablative radiotherapy (SAbR) patients potentially mitigates post-SAbR radiation injury. However, the performance of common clinical dose calculation algorithms in airways has not been thoroughly studied. Airways are of particular concern because their small size and the density differences they create have the potential to hinder dose calculation accuracy. To address this gap in knowledge, here we investigate dosimetric accuracy in airways of two commonly used dose calculation algorithms, the anisotropic analytical algorithm (AAA) and Acuros-XB (AXB), recreating clinical treatment plans on a cohort of four SAbR patients. A virtual bronchoscopy software was used to delineate 856 airways on a high-resolution breath-hold CT (BHCT) image acquired for each patient. The planning target volumes (PTVs) and standard thoracic OARs were contoured on an average CT (AVG) image over the breathing cycle. Conformal and intensity-modulated radiation therapy plans were recreated on the BHCT image and on the AVG image, for a total of four plan types per patient. Dose calculations were performed using AAA and AXB, and the differences in maximum and mean dose in each structure were calculated. The median differences in maximum dose among all airways were ≤0.3Gy in magnitude for all four plan types. With airways grouped by dose-to-structure or diameter, median dose differences were still ≤0.5Gy in magnitude, with no clear dependence on airway size. These results, along with our previous airway radiosensitivity works, suggest that dose differences between AAA and AXB correspond to an airway collapse variation ≤0.7% in magnitude. This variation in airway injury risk can be considered as not clinically relevant, and the use of either AAA or AXB is therefore appropriate when including patient airways as individual OARs so as to reduce risk of radiation-induced lung toxicity.

在之前的工作中，我们发现将个体气道作为风险器官 (OAR) 用于治疗肺立体定向消融放射治疗 (SAbR) 患者可能会减轻 SAbR 后的辐射损伤。然而，尚未彻底研究气道中常见临床剂量计算算法的性能。气道特别值得关注，因为它们的小尺寸和它们产生的密度差异有可能阻碍剂量计算的准确性。为了解决这一知识差距，我们在此研究了两种常用剂量计算算法（各向异性分析算法 (AAA) 和 Acuros-XB (AXB)）在气道中的剂量学准确性，对一组四名 SAbR 患者重新制定临床治疗计划。使用虚拟支气管镜软件在为每位患者采集的高分辨率屏气 CT (BHCT) 图像上描绘 856 条气道。计划目标容积 (PTV) 和标准胸 OAR 在呼吸周期内的平均 CT (AVG) 图像上绘制轮廓。在 BHCT 图像和 AVG 图像上重新创建了适形和调强放射治疗计划，每位患者总共有四种计划类型。使用 AAA 和 AXB 进行剂量计算，并计算每个结构中最大和平均剂量的差异。对于所有四种计划类型，所有气道之间最大剂量的中位数差异均≤0.3Gy。根据剂量-结构或直径对气道进行分组，中位剂量差异的幅度仍≤0.5Gy，与气道大小没有明显的相关性。这些结果，连同我们之前的气道放射敏感性研究，表明 AAA 和 AXB 之间的剂量差异对应于气道塌陷变化幅度≤0.7%。可以认为气道损伤风险的这种变化与临床无关，因此在将患者气道作为单独的 OAR 时使用 AAA 或 AXB 是合适的，以降低辐射引起的肺毒性风险。