Department of Radiation Oncology, Huadong Hospital, Fudan University, Shanghai, China.
Front Public Health. 2022 Mar 22;10:860135. doi: 10.3389/fpubh.2022.860135. eCollection 2022.
The purpose of this study was to evaluate the accuracy of a lung stereotactic body radiotherapy (SBRT) treatment plan with the target of a newly predicted internal target volume (ITV) and the feasibility of its clinical application. ITV was automatically generated by our in-house deep learning model according to the cone-beam CT (CBCT) image database.
A retrospective study of 45 patients who underwent SBRT was involved, and Mask R-CNN based algorithm model helped to predict the internal target volume (ITV) using the CBCT image database. The geometric accuracy of ITV was verified by the Dice Similarity Coefficient (DSC), 3D Motion Range (R), Relative Volume Index (RVI), and Hausdorff Distance (HD). The PTV was generated by ITV, which was registered and then projected on free-breath CT (FBCT) images. The PTV was margined from the GTV on FBCT images gross tumor volume on free-breath CT (GTV). Treatment plans with the target of Predict planning target volume on CBCT images (PTV) and planning target volume on free-breath CT (PTV) were respectively re-established, and the dosimetric parameters included the ratio of the volume of patients receiving at least the prescribed dose to the volume of PTV (R), the ratio of the volume of patients receiving at least 50% of the prescribed dose to the volume of PTV in the Radiation Therapy Oncology Group (RTOG) 0813 Trial (R), Gradient Index (GI), and the maximum dose 2 cm from the PTV (D), which were evaluated Plan, plan which based on PTV (Plan), and plan which based on PTV (Plan).
The geometric results showed that there existed a good correlation between ITV and ITV on the 4-dimensional CT [ITV; DSC= 0.83 ±0.18]. However, the average volume of ITV was 10% less than that of ITV ( = 0.333). No significant difference in dose coverage was found in V for the ITV with 99.98 ± 0.04% in the ITV vs. 97.56 ± 4.71% in the ITV ( = 0.162). Dosimetry parameters of PTV, including R, R, GI and D showed no statistically significant difference between each plan ( > 0.05).
Dosimetric parameters of Plan are clinically comparable to those of the original Plan This study confirmed that the treatment plan based on ITV produced by our model could automatically meet clinical requirements. Thus, for patients undergoing lung SBRT, the model has great potential for using CBCT images for ITV contouring which can be used in treatment planning.
本研究旨在评估使用新预测的内部靶区(ITV)和临床应用可行性的肺立体定向体放射治疗(SBRT)治疗计划的准确性。ITV 是根据锥形束 CT(CBCT)图像数据库由我们内部的深度学习模型自动生成的。
回顾性分析了 45 例接受 SBRT 的患者,基于 Mask R-CNN 的算法模型利用 CBCT 图像数据库预测 ITV。使用 Dice 相似系数(DSC)、3D 运动范围(R)、相对体积指数(RVI)和 Hausdorff 距离(HD)验证 ITV 的几何精度。PTV 由 ITV 生成,然后对 ITV 进行注册并将其投影到自由呼吸 CT(FBCT)图像上。PTV 是从 FBCT 图像上的 GTV 上 Margined 的 Gross Tumor Volume on Free-Breath CT(GTV)。分别为基于 CBCT 图像预测计划靶区(PTV)和基于自由呼吸 CT 的计划靶区(PTV)的治疗计划重新建立,剂量学参数包括至少接受规定剂量的患者的体积与 PTV 体积的比值(R)、至少接受 50%规定剂量的患者的体积与 RTOG 0813 试验(R)中 PTV 体积的比值(R)、梯度指数(GI)和距离 PTV 2cm 处的最大剂量(D),评估了计划、基于 PTV 的计划(Plan)和基于 PTV 的计划(Plan)。
几何结果表明,ITV 与 4D CT 上的 ITV 之间存在良好的相关性[ITV;DSC=0.83±0.18]。然而,ITV 的平均体积比 ITV 小 10%(=0.333)。99.98±0.04%的 ITV 和 97.56±4.71%的 ITV 的 V 剂量覆盖率没有显著差异(=0.162)。PTV 的剂量学参数,包括 R、R、GI 和 D,在每个计划之间没有统计学上的显著差异(>0.05)。
Plan 的剂量学参数与原始 Plan 具有临床可比性。本研究证实,基于我们模型生成的 ITV 的治疗计划可以自动满足临床要求。因此,对于接受肺 SBRT 的患者,该模型具有利用 CBCT 图像进行 ITV 轮廓勾画并应用于治疗计划的巨大潜力。