Department of Medical Physics, Memorial Sloan-Kettering Cancer Center, New York, NY, United States of America.
Department of Electrical & Computer Engineering, Georgia Institute of Technology, Atlanta, GA, United States of America.
Phys Med Biol. 2022 Sep 14;67(18). doi: 10.1088/1361-6560/ac8d45.
To propose a novel moment-based loss function for predicting 3D dose distribution for the challenging conventional lung intensity modulated radiation therapy plans. The moment-based loss function is convex and differentiable and can easily incorporate clinical dose volume histogram (DVH) domain knowledge in any deep learning (DL) framework without computational overhead.We used a large dataset of 360 (240 for training, 50 for validation and 70 for testing) conventional lung patients with 2 Gy × 30 fractions to train the DL model using clinically treated plans at our institution. We trained a UNet like convolutional neural network architecture using computed tomography, planning target volume and organ-at-risk contours as input to infer corresponding voxel-wise 3D dose distribution. We evaluated three different loss functions: (1) the popular mean absolute error (MAE) loss, (2) the recently developed MAE + DVH loss, and (3) the proposed MAE + moments loss. The quality of the predictions was compared using different DVH metrics as well as dose-score and DVH-score, recently introduced by theModel with (MAE + moment) loss function outperformed the model with MAE loss by significantly improving the DVH-score (11%,< 0.01) while having similar computational cost. It also outperformed the model trained with (MAE + DVH) by significantly improving the computational cost (48%) and the DVH-score (8%,< 0.01).DVH metrics are widely accepted evaluation criteria in the clinic. However, incorporating them into the 3D dose prediction model is challenging due to their non-convexity and non-differentiability. Moments provide a mathematically rigorous and computationally efficient way to incorporate DVH information in any DL architecture. The code, pretrained models, docker container, and Google Colab project along with a sample dataset are available on our DoseRTX GitHub (https://github.com/nadeemlab/DoseRTX).
提出了一种新的基于矩的损失函数，用于预测具有挑战性的传统肺部强度调制放射治疗计划的三维剂量分布。基于矩的损失函数是凸的和可微的，可以在任何深度学习（DL）框架中轻松纳入临床剂量体积直方图（DVH）域知识，而不会增加计算开销。
我们使用了一个包含 360 名（240 名用于训练，50 名用于验证，70 名用于测试）常规肺部患者的大型数据集，这些患者接受了 2 Gy×30 个分次的治疗，使用我们机构治疗的计划来训练 DL 模型。我们使用 CT、计划靶区和危及器官轮廓作为输入，训练了一个类似于 UNet 的卷积神经网络架构，以推断相应的体素三维剂量分布。我们评估了三种不同的损失函数：（1）流行的平均绝对误差（MAE）损失，（2）最近开发的 MAE+DVH 损失，和（3）建议的 MAE+矩损失。使用不同的 DVH 指标以及剂量评分和 DVH 评分来比较预测的质量，最近由模型引入。
与 MAE 损失函数的模型相比，具有（MAE+矩）损失函数的模型通过显著提高 DVH 评分（11%，<0.01），同时具有相似的计算成本，从而显著提高了模型的性能。与使用（MAE+DVH）训练的模型相比，它还通过显著提高计算成本（48%）和 DVH 评分（8%，<0.01）来提高性能。
DVH 指标是临床中广泛接受的评估标准。然而，由于其非凸性和不可微性，将其纳入 3D 剂量预测模型是具有挑战性的。矩为在任何 DL 架构中纳入 DVH 信息提供了一种数学上严格且计算效率高的方法。代码、预训练模型、Docker 容器和 Google Colab 项目以及一个示例数据集都可在我们的 DoseRTX GitHub 上获得（https://github.com/nadeemlab/DoseRTX）。
