Department of Radiation Physics, The University of Texas MD Anderson Cancer Center, Houston, Texas; The University of Texas MD Anderson Cancer Center UTHealth Houston Graduate School of Biomedical Sciences, Houston, Texas.
Department of Radiation Oncology, Stanford University, Stanford, California.
Pract Radiat Oncol. 2023 May-Jun;13(3):e282-e291. doi: 10.1016/j.prro.2022.12.003. Epub 2023 Jan 24.
This study aimed to use deep learning-based dose prediction to assess head and neck (HN) plan quality and identify suboptimal plans.
A total of 245 volumetric modulated arc therapy HN plans were created using RapidPlan knowledge-based planning (KBP). A subset of 112 high-quality plans was selected under the supervision of an HN radiation oncologist. We trained a 3D Dense Dilated U-Net architecture to predict 3-dimensional dose distributions using 3-fold cross-validation on 90 plans. Model inputs included computed tomography images, target prescriptions, and contours for targets and organs at risk (OARs). The model's performance was assessed on the remaining 22 test plans. We then tested the application of the dose prediction model for automated review of plan quality. Dose distributions were predicted on 14 clinical plans. The predicted versus clinical OAR dose metrics were compared to flag OARs with suboptimal normal tissue sparing using a 2 Gy dose difference or 3% dose-volume threshold. OAR flags were compared with manual flags by 3 HN radiation oncologists.
The predicted dose distributions were of comparable quality to the KBP plans. The differences between the predicted and KBP-planned D,D, and D across the targets were within -2.53% ± 1.34%, -0.42% ± 1.27%, and -0.12% ± 1.97%, respectively, and the OAR mean and maximum doses were within -0.33 ± 1.40 Gy and -0.96 ± 2.08 Gy, respectively. For the plan quality assessment study, radiation oncologists flagged 47 OARs for possible plan improvement. There was high interphysician variability; 83% of physician-flagged OARs were flagged by only one of 3 physicians. The comparative dose prediction model flagged 63 OARs, including 30 of 47 physician-flagged OARs.
Deep learning can predict high-quality dose distributions, which can be used as comparative dose distributions for automated, individualized assessment of HN plan quality.
本研究旨在利用基于深度学习的剂量预测来评估头颈部（HN）计划质量并识别次优计划。
共创建了 245 例容积调强弧形治疗 HN 计划，使用 RapidPlan 基于知识的计划（KBP）。在一名 HN 放射肿瘤学家的监督下，选择了一组 112 例高质量计划。我们使用 3 倍交叉验证在 90 个计划上训练了一个 3D 密集扩张 U-Net 架构，以预测 3 维剂量分布。模型输入包括 CT 图像、靶区处方和靶区及危及器官（OARs）的轮廓。然后在剩余的 22 个测试计划上评估模型的性能。然后，我们测试了剂量预测模型在自动审查计划质量中的应用。对 14 例临床计划进行了剂量分布预测。比较预测与临床 OAR 剂量指标，使用 2Gy 剂量差异或 3%剂量-体积阈值标记 OAR 正常组织保护不足的情况。三位 HN 放射肿瘤学家将 OAR 标记与手动标记进行了比较。
预测的剂量分布与 KBP 计划质量相当。预测与 KBP 计划靶区的 D,D 和 D 之间的差异分别在-2.53%±1.34%、-0.42%±1.27%和-0.12%±1.97%之间，OAR 的平均和最大剂量分别在-0.33±1.40Gy 和-0.96±2.08Gy 之间。在计划质量评估研究中，放射肿瘤学家标记了 47 个 OAR 可能需要改进计划。存在高度的医师间变异性；只有 3 位医师中的 1 位标记了 83%的医师标记的 OAR。比较剂量预测模型标记了 63 个 OAR，包括 30 个医师标记的 OAR。
深度学习可以预测高质量的剂量分布，可用于自动、个体化评估 HN 计划质量的比较剂量分布。
