# DeepDoseNet: A Deep Learning model for 3D Dose Prediction in Radiation Therapy

arXiv: https://arxiv.org/abs/2111.00077 · HTML: https://mlsanity.com/paper/2111.00077/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 18 of its 63 references that are in the map.
- First posted on arXiv 29 Oct 2021.

## Builds on (most cited first, 18 of 18)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks](https://mlsanity.com/paper/1506.01497/) (2015)
- [Densely Connected Convolutional Networks](https://mlsanity.com/paper/1608.06993/) (2016)
- [DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs](https://mlsanity.com/paper/1606.00915/) (2016)
- [3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation](https://mlsanity.com/paper/1606.06650/) (2016)
- [Dual Attention Network for Scene Segmentation](https://mlsanity.com/paper/1809.02983/) (2018)
- [DetectoRS: Detecting Objects with Recursive Feature Pyramid and Switchable Atrous Convolution](https://mlsanity.com/paper/2006.02334/) (2020)
- [Understanding Convolution for Semantic Segmentation](https://mlsanity.com/paper/1702.08502/) (2017)
- [A Deep Learning Framework for Unsupervised Affine and Deformable Image Registration](https://mlsanity.com/paper/1809.06130/) (2018)
- [Automatic 3D Cardiovascular MR Segmentation with Densely-Connected Volumetric ConvNets](https://mlsanity.com/paper/1708.00573/) (2017)
- [OpenKBP: The open-access knowledge-based planning grand challenge](https://mlsanity.com/paper/2011.14076/) (2020)
- [Three-Dimensional Radiotherapy Dose Prediction on Head and Neck Cancer Patients with a Hierarchically Densely Connected U-net Deep Learning Architecture](https://mlsanity.com/paper/1805.10397/) (2018)
- [Three-Dimensional Dose Prediction for Lung IMRT Patients with Deep Neural Networks: Robust Learning from Heterogeneous Beam Configurations](https://mlsanity.com/paper/1812.06934/) (2018)
- [Incorporating human and learned domain knowledge into training deep neural networks: A differentiable dose volume histogram and adversarial inspired framework for generating Pareto optimal dose distributions in radiation therapy](https://mlsanity.com/paper/1908.05874/) (2019)
- [3D Densely Convolutional Networks for Volumetric Segmentation](https://mlsanity.com/paper/1709.03199/) (2017)

## Built on by (most cited first, 3 of 3)

- [Swin UNETR++: Advancing Transformer-Based Dense Dose Prediction Towards Fully Automated Radiation Oncology Treatments](https://mlsanity.com/paper/2311.06572/) (2023)
- [Deep Evidential Learning for Radiotherapy Dose Prediction](https://mlsanity.com/paper/2404.17126/) (2024)
- [Any2Any 3D Diffusion Models with Knowledge Transfer: A Radiotherapy Planning Study](https://mlsanity.com/paper/2605.09622/) (2026)

## Abstract (arXiv)

The DeepDoseNet 3D dose prediction model based on ResNet and Dilated DenseNet is proposed. The 340 head-and-neck datasets from the 2020 AAPM OpenKBP challenge were utilized, with 200 for training, 40 for validation, and 100 for testing. Structures include 56Gy, 63Gy, 70Gy PTVs, and brainstem, spinal cord, right parotid, left parotid, larynx, esophagus, and mandible OARs. Mean squared error (MSE) loss, mean absolute error (MAE) loss, and MAE plus dose-volume histogram (DVH) based loss functions were investigated. Each model's performance was compared using a 3D dose score, barSD, (mean absolute difference between ground truth and predicted 3D dose distributions) and a DVH score, barSDVH (mean absolute difference between ground truth and predicted dose-volume metrics).Furthermore, DVH metrics Mean[Gy] and D0.1cc [Gy] for OARs and D99%, D95%, D1% for PTVs were computed. DeepDoseNet with the MAE plus DVH-based loss function had the best dose score performance of the OpenKBP entries. MAE+DVH model had the lowest prediction error (P<0.0001, Wilcoxon test) on validation and test datasets (validation: barSD=2.3Gy, barSDVH=1.9Gy; test: barSD=2.0Gy, barSDVH=1.6Gy) followed by the MAE model (validation: barSD=3.6Gy, barSDVH=2.4Gy; test: barSD=3.5Gy, barSDVH=2.3Gy). The MSE model had the highest prediction error (validation: barSD=3.7Gy, barSDVH=3.2Gy; test: barSD=3.6Gy, barSDVH=3.0Gy). No significant difference was found among models in terms of Mean [Gy], but the MAE+DVH model significantly outperformed the MAE and MSE models in terms of D0.1cc[Gy], particularly for mandible and parotids on both validation (P<0.01) and test (P<0.0001) datasets. MAE+DVH outperformed (P<0.0001) in terms of D99%, D95%, D1% for targets. MAE+DVH reduced barSD by ~60% and barSDVH by ~70%.