Wang Siqiu, Liao Chien-Yi, Choi Byongsu, All Sean, Bai Ti, Visak Justin, Moon Dominic, Pompos Arnold, Avkshtol Vladmir, Parsons David, Godley Andrew, Sher David, Lin Mu-Han
Department of Radiation Oncology, The University of Texas Southwestern Medical Center, Dallas, Texas.
Department of Radiation Oncology, The University of Texas Southwestern Medical Center, Dallas, Texas.
Pract Radiat Oncol. 2025 Mar-Apr;15(2):e211-e219. doi: 10.1016/j.prro.2024.09.005. Epub 2024 Oct 5.
Online adaptive radiation therapy (oART) has high resource costs especially for head and neck (H&N) cancer, which requires recontouring complex targets and numerous organs-at-risk (OARs). Adaptive radiation therapy systems provide autocontours to help. We aimed to explore the optimal level of editing automatic contours to maintain plan quality in a cone beam computed tomography-based oART system for H&N cancer. In this system, influencer OAR contours are generated and reviewed first, which then drives the autocontouring of the remaining OARs and targets.
Three-hundred-forty-nine adapted fractions of 44 patients with H&N cancer were retrospectively analyzed, with physician-edited OARs and targets. These contours and associated online-adapted plans served as the gold standard for comparison. We simulated 3 contour editing workflows: (1) no editing of contours; (2) only editing the influencers; and (3) editing the influencers and targets. The geometric difference was quantified using the Dice similarity coefficient (DSC) and Hausdorff distance (HD). The dosimetric differences in target coverage and OAR doses were calculated between the gold standard and these 3 simulated workflows.
Workflow 1 resulted in significantly inferior contour quality for all OARs (mean DSC, 0.85 ± 0.17 and HD95, 3.10 ± 5.80mm); hence, dosimetric data was not calculated for workflow 1. In workflow 2, the frequency of physician editing targets and remaining OARs were 80.8% to 95.7% and 2.3% (brachial plexus) to 67.7% (oral cavity), respectively, where the OAR differences were geometrically minor (mean DSC >0.95 with std ≤0.09). However, because of the unedited target contours of workflow 2 (mean DSC, 0.86-0.92 and mean HD95, 2.56-3.30 mm vs the ground-truth targets), plans were inadequate with insufficient coverage. In workflow 3, when both targets and influencers were edited (noninfluencer OARs were unedited), >95.5% of the adapted plans achieved the patient-specific dosimetry goals.
The cone beam computed tomography-based H&N oART workflow can be meaningfully accelerated by only editing the influencers and targets while omitting the remaining OARs without compromising the quality of the adaptive plans.
在线自适应放射治疗（oART）资源成本高昂，尤其是对于头颈部（H&N）癌，其需要对复杂靶区和众多危及器官（OARs）进行重新轮廓勾画。自适应放射治疗系统提供自动轮廓来提供帮助。我们旨在探索在基于锥形束计算机断层扫描的H&N癌oART系统中，编辑自动轮廓的最佳水平，以维持计划质量。在该系统中，首先生成并审查影响OAR轮廓，然后驱动其余OAR和靶区的自动轮廓勾画。
回顾性分析44例H&N癌患者的349个适配分次，包括医生编辑的OAR和靶区。这些轮廓和相关的在线适配计划用作比较的金标准。我们模拟了3种轮廓编辑工作流程：（1）不编辑轮廓；（2）仅编辑影响因素；（3）编辑影响因素和靶区。使用骰子相似系数（DSC）和豪斯多夫距离（HD）量化几何差异。计算金标准与这3种模拟工作流程之间靶区覆盖和OAR剂量的剂量差异。
工作流程1导致所有OAR的轮廓质量显著较差（平均DSC，0.85±0.17；HD95，3.10±5.80mm）；因此，未计算工作流程1的剂量数据。在工作流程2中，医生编辑靶区和其余OAR的频率分别为80.8%至95.7%和2.3%（臂丛神经）至67.7%（口腔），其中OAR差异在几何上较小（平均DSC>0.95，标准差≤0.09）。然而，由于工作流程2中靶区轮廓未编辑（平均DSC，0.86 - 0.92；平均HD95，2.56 - 3.30mm，与真实靶区相比），计划覆盖不足。在工作流程3中，当靶区和影响因素都被编辑时（非影响OAR未编辑），>95.5%的适配计划实现了患者特异性剂量目标。
基于锥形束计算机断层扫描的H&N oART工作流程可以通过仅编辑影响因素和靶区，同时省略其余OAR来显著加速，而不会损害自适应计划的质量。
