Deep learning‐based dose prediction for low‐energy electron beam superficial radiotherapy

Abstract Background Accurate surface dose calculation is crucial in superficial low‐energy electron beam radiotherapy owing to shallow treatment depths and the risk of skin toxicity. Traditional Monte Carlo (MC) simulations are precise but computationally expensive and time‐consuming. Methods This s...

Full description

Saved in:
Bibliographic Details
Main Authors: Jialin Huang, Zhitao Dai, Shuai Hu, Yuanchun Ye, Yuling Chen, Ming Li, Tianye Niu, Jinfen Zheng, Yongsheng Huang, Yuanjie Bi
Format: Article
Language:English
Published: Wiley 2025-06-01
Series:Precision Radiation Oncology
Subjects:
Online Access:https://doi.org/10.1002/pro6.70015
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850107006699962368
author Jialin Huang
Zhitao Dai
Shuai Hu
Yuanchun Ye
Yuling Chen
Ming Li
Tianye Niu
Jinfen Zheng
Yongsheng Huang
Yuanjie Bi
author_facet Jialin Huang
Zhitao Dai
Shuai Hu
Yuanchun Ye
Yuling Chen
Ming Li
Tianye Niu
Jinfen Zheng
Yongsheng Huang
Yuanjie Bi
author_sort Jialin Huang
collection DOAJ
description Abstract Background Accurate surface dose calculation is crucial in superficial low‐energy electron beam radiotherapy owing to shallow treatment depths and the risk of skin toxicity. Traditional Monte Carlo (MC) simulations are precise but computationally expensive and time‐consuming. Methods This study combined MC simulations with deep learning to improve both accuracy and speed. DOSXYZnrc was used to simulate low‐energy electron beams for six body sites, generating computed tomography phantoms and corresponding dose distributions. A cascaded 3D U‐Net (C3D) model was trained on these datasets to predict dose distributions rapidly. Results The C3D model demonstrated significant improvements over traditional 3D U‐Net models, achieving a minimum Gamma pass rate of 92.09% and a minimum dose difference pass rate of 93.58%. The model completed dose predictions in just 0.42 seconds, making predictions approximately 140,000 times faster than MC simulations. In the evaluation of dose distributions across six anatomical regions, C3D consistently outperformed other deep learning models (3D U‐Net, Deep Convolutional Neural Network, and HD U‐Net) in both accuracy and robustness. Conclusion The integration of deep learning with MC simulations significantly enhances the efficiency of surface dose calculations in superficial electron beam radiotherapy. The C3D model provides rapid and accurate dose predictions, facilitating efficient treatment planning while maintaining high accuracy.
format Article
id doaj-art-340c3624102d4ebdbb28a8de7ff68c29
institution OA Journals
issn 2398-7324
language English
publishDate 2025-06-01
publisher Wiley
record_format Article
series Precision Radiation Oncology
spelling doaj-art-340c3624102d4ebdbb28a8de7ff68c292025-08-20T02:38:41ZengWileyPrecision Radiation Oncology2398-73242025-06-019210811910.1002/pro6.70015Deep learning‐based dose prediction for low‐energy electron beam superficial radiotherapyJialin Huang0Zhitao Dai1Shuai Hu2Yuanchun Ye3Yuling Chen4Ming Li5Tianye Niu6Jinfen Zheng7Yongsheng Huang8Yuanjie Bi9School of ScienceShenzhen Campus of Sun Yat‐sen UniversityShenzhenChinaDepartment of Radiation OncologyCancer Hospital Chinese Academy of Medical SciencesShenzhenGuangdongChinaSchool of ScienceShenzhen Campus of Sun Yat‐sen UniversityShenzhenChinaSchool of ScienceShenzhen Campus of Sun Yat‐sen UniversityShenzhenChinaDepartment of Rheumatology and Immunology, The Seventh AffiliatedHospital Sun Yat‐sen UniversityShenzhen GuangdongChinaIncubation Center of Guangdong University of TechnologyShenzhenGuangdongChinaShenzhen Bay LaboratoryShenzhenChinaDermatology, Center for Chronic Disease Prevention of ShenzhenShenzhenChinaSchool of ScienceShenzhen Campus of Sun Yat‐sen UniversityShenzhenChinaSchool of ScienceShenzhen Campus of Sun Yat‐sen UniversityShenzhenChinaAbstract Background Accurate surface dose calculation is crucial in superficial low‐energy electron beam radiotherapy owing to shallow treatment depths and the risk of skin toxicity. Traditional Monte Carlo (MC) simulations are precise but computationally expensive and time‐consuming. Methods This study combined MC simulations with deep learning to improve both accuracy and speed. DOSXYZnrc was used to simulate low‐energy electron beams for six body sites, generating computed tomography phantoms and corresponding dose distributions. A cascaded 3D U‐Net (C3D) model was trained on these datasets to predict dose distributions rapidly. Results The C3D model demonstrated significant improvements over traditional 3D U‐Net models, achieving a minimum Gamma pass rate of 92.09% and a minimum dose difference pass rate of 93.58%. The model completed dose predictions in just 0.42 seconds, making predictions approximately 140,000 times faster than MC simulations. In the evaluation of dose distributions across six anatomical regions, C3D consistently outperformed other deep learning models (3D U‐Net, Deep Convolutional Neural Network, and HD U‐Net) in both accuracy and robustness. Conclusion The integration of deep learning with MC simulations significantly enhances the efficiency of surface dose calculations in superficial electron beam radiotherapy. The C3D model provides rapid and accurate dose predictions, facilitating efficient treatment planning while maintaining high accuracy.https://doi.org/10.1002/pro6.70015Dose predictionDeep learningMonte Carlo simulationLow‐energy electron beamSuperficial treatmentRadiotherapy
spellingShingle Jialin Huang
Zhitao Dai
Shuai Hu
Yuanchun Ye
Yuling Chen
Ming Li
Tianye Niu
Jinfen Zheng
Yongsheng Huang
Yuanjie Bi
Deep learning‐based dose prediction for low‐energy electron beam superficial radiotherapy
Precision Radiation Oncology
Dose prediction
Deep learning
Monte Carlo simulation
Low‐energy electron beam
Superficial treatment
Radiotherapy
title Deep learning‐based dose prediction for low‐energy electron beam superficial radiotherapy
title_full Deep learning‐based dose prediction for low‐energy electron beam superficial radiotherapy
title_fullStr Deep learning‐based dose prediction for low‐energy electron beam superficial radiotherapy
title_full_unstemmed Deep learning‐based dose prediction for low‐energy electron beam superficial radiotherapy
title_short Deep learning‐based dose prediction for low‐energy electron beam superficial radiotherapy
title_sort deep learning based dose prediction for low energy electron beam superficial radiotherapy
topic Dose prediction
Deep learning
Monte Carlo simulation
Low‐energy electron beam
Superficial treatment
Radiotherapy
url https://doi.org/10.1002/pro6.70015
work_keys_str_mv AT jialinhuang deeplearningbaseddosepredictionforlowenergyelectronbeamsuperficialradiotherapy
AT zhitaodai deeplearningbaseddosepredictionforlowenergyelectronbeamsuperficialradiotherapy
AT shuaihu deeplearningbaseddosepredictionforlowenergyelectronbeamsuperficialradiotherapy
AT yuanchunye deeplearningbaseddosepredictionforlowenergyelectronbeamsuperficialradiotherapy
AT yulingchen deeplearningbaseddosepredictionforlowenergyelectronbeamsuperficialradiotherapy
AT mingli deeplearningbaseddosepredictionforlowenergyelectronbeamsuperficialradiotherapy
AT tianyeniu deeplearningbaseddosepredictionforlowenergyelectronbeamsuperficialradiotherapy
AT jinfenzheng deeplearningbaseddosepredictionforlowenergyelectronbeamsuperficialradiotherapy
AT yongshenghuang deeplearningbaseddosepredictionforlowenergyelectronbeamsuperficialradiotherapy
AT yuanjiebi deeplearningbaseddosepredictionforlowenergyelectronbeamsuperficialradiotherapy