DVHnet：一种基于深度学习的放疗计划中个体化剂量体积直方图预测方法。

DVHnet: A deep learning-based prediction of patient-specific dose volume histograms for radiotherapy planning.

机构信息

National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, 100021, China.

出版信息

Med Phys. 2021 Jun;48(6):2705-2713. doi: 10.1002/mp.14758. Epub 2021 Apr 18.

DOI:10.1002/mp.14758

PMID:33550616

Abstract

PURPOSE

To develop a deep learning method to predict patient-specific dose volume histograms (DVHs) for radiotherapy planning.

METHODS

Patient data included 180 cases with nasopharyngeal cancer, of which 153 cases were used for training and 27 for testing. A network (named "DVHnet") based on a convolutional neural network (CNN) was designed for directly predicting DVHs of organs at risk (OARs). Two-channel images with contoured structures were generated as the inputs for training the model. A one-dimensional array consisting of 256 continuous volume percentages on a DVH curve for each slice was calculated as the corresponding output. The combined DVH was then calculated. Sixteen OARs were modeled in the study. Prediction accuracy was evaluated against the corresponding DVH curve of ground truth (GT) plans. A global DVH analysis and critical dosimetry metrics for each OAR were calculated for quantitative evaluation. The performance of DVHnet also was evaluated against two baselines: DosemapNet (developed by our research group) and commercial RapidPlan software.

RESULTS

The predicted mean difference in average dose of all OARs using DVHnet was 0.30 ± 0.95 Gy. And the predicted differences in D2% and D50 can be control within 2.32 and 0.69 Gy. For most OARs, there were no obvious differences between the dosimetric metrics of the predicted and GT values for both DVHnet and DosemapNet (P ≥ 0.05). Only the predicted D2% of the optic organs for DVHnet, and of brain stem PRV for DosemapNet displayed statistically significant differences. Except for the optic organs, DVHnet performs better than or comparably with RapidPlan. The mean difference in proportion of points of interest was 3.59% ± 7.78%.

CONCLUSIONS

A deep learning network model was developed to automatically extract useful features for accurate prediction of patient-specific DVH curves directly. The performance of DVHnet was comparable to DosemapNet and RapidPlan.

摘要

目的

开发一种深度学习方法，以预测放射治疗计划的患者特异性剂量体积直方图（DVH）。

方法

患者数据包括 180 例鼻咽癌患者，其中 153 例用于训练，27 例用于测试。设计了一种基于卷积神经网络（CNN）的网络（名为“DVHnet”），用于直接预测危及器官（OAR）的 DVH。将带有轮廓结构的双通道图像作为模型训练的输入。为每个切片的 DVH 曲线计算由 256 个连续体积百分比组成的一维数组作为相应的输出。然后计算联合 DVH。在研究中对 16 个 OAR 进行建模。将预测准确性与地面真实（GT）计划的相应 DVH 曲线进行对比评估。计算每个 OAR 的全局 DVH 分析和关键剂量学指标进行定量评估。还对 DVHnet 的性能与两个基线进行了评估：DosemapNet（由我们的研究小组开发）和商业 RapidPlan 软件。

结果

使用 DVHnet 预测的所有 OAR 的平均剂量的平均差异为 0.30±0.95 Gy。并且 D2%和 D50 的预测差异可以控制在 2.32 和 0.69 Gy 以内。对于大多数 OAR，DVHnet 和 DosemapNet 的预测值与 GT 值之间的剂量学指标没有明显差异（P≥0.05）。仅 DVHnet 的光学器官的预测 D2%和 DosemapNet 的脑干 PRV 的预测值存在统计学差异。除了光学器官外，DVHnet 的性能优于或与 RapidPlan 相当。感兴趣点的比例差异的平均值为 3.59%±7.78%。