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 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.

Keywords

Dose prediction / Deep learning / Monte Carlo simulation / Low-energy electron beam / Superficial treatment / Radiotherapy

References

[1] Hogstrom KR, Almond PR. Review of electron beam therapy physics. Phys Med Biol. 2006; 51(13): R455-R489.

[2] Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2021; 71(3): 209-249.

[3] Shen J, Lian X, Sun Y, et al. Hypofractionated electron-beam radiation therapy for keloids: retrospective study of 568 cases with 834 lesions. J Radiat Res. 2015; 56(5): 811-817.

[4] Debela DT, Muzazu SG, Heraro KD, et al. New approaches and procedures for cancer treatment: Current perspectives. SAGE Open Med. 2021; 9: 20503121211034366.

[5] Skinner L, Knopp R, Wang YC, et al. CT-less electron radiotherapy simulation and planning with a consumer 3D camera. J Appl Clin Med Phys. 2021; 22(7): 128-136.

[6] Shiu AS, Tung S, Hogstrom KR, et al. Verification data for electron beam dose algorithms. Med Phys. 1992; 19(3): 623-636.

[7] Storchi P, Woudstra E. Calculation of the absorbed dose distribution due to irregularly shaped photon beams using pencil beam kernels derived form basic beam data. Phys Med Biol. 1996; 41(4): 637-656.

[8] Fogliata A, Nicolini G, Vanetti E, Clivio A, Cozzi L. Dosimetric validation of the anisotropic analytical algorithm for photon dose calculation: fundamental characterization in water. Phys Med Biol. 2006; 51(6): 1421-1438.

[9] Fogliata A, Nicolini G, Clivio A, Vanetti E, Cozzi L. Dosimetric evaluation of Acuros XB Advanced Dose Calculation algorithm in heterogeneous media. Radiat Oncol. 2011; 6: 82.

[10] Hasenbalg F, Neuenschwander H, Mini R, Born EJ. Collapsed cone convolution and analytical anisotropic algorithm dose calculations compared to VMC++ Monte Carlo simulations in clinical cases. Phys Med Biol. 2007; 52(13): 3679-3691.

[11] Kui X, Liu F, Yang M, et al. A review of dose prediction methods for tumor radiation therapy. Meta-Radiology. 2024; 2(1): 100057.

[12] Ronneberger O, Fischer P, Brox T. U-Net: Convolutional Networks for Biomedical Image Segmentation. arXiv.

[13] Gautam S, Osman AFI, Richeson D, et al. Attention 3D UNET for dose distribution prediction of high-dose-rate brachytherapy of cervical cancer: Intracavitary applicators. J Appl Clin Med Phys. 2025; 26(2): e14568.

[14] Nguyen D, Jia X, Sher D, et al. 3D radiotherapy dose prediction on head and neck cancer patients with a hierarchically densely connected U-net deep learning architecture. Phys Med Biol. 2019; 64(6): 065020.

[15] Xiong T, Cai J, Zhou F, Liu B, Zhang J, Wu Q. An end-to-end deep convolutional neural network-based dose engine for parotid gland cancer seed implant brachytherapy. Med Phys. 2024; 51(9): 6365-6377.

[16] National Research Council of Canada. Metrology Research Centre. Ionizing Radiation Standards. EGSnrc: software for Monte Carlo simulation of ionizing radiation. National Research Council of Canada. 2021.

[17] Walters BRB, Kawrakow I. Dosxyznrc users manual. 2016. https://api.semanticscholar.org/CorpusID:62929952

[18] Kawrakow I, Fippel M, Friedrich K. 3D electron dose calculation using a Voxel based Monte Carlo algorithm (VMC). Med Phys. 1996; 23(4): 445-457.

[19] Xue J, He K, Nie D, et al. Cascaded MultiTask 3-D Fully Convolutional Networks for Pancreas Segmentation. IEEE Trans Cybern. 2021; 51(4): 2153-2165.

[20] Zhan B, Xiao J, Cao C, et al. Multi-constraint generative adversarial network for dose prediction in radiotherapy. Med Image Anal. 2022; 77: 102339.

[21] Kingma DP, Ba J. Adam: A Method for Stochastic Optimization. arXiv.

[22] Galapon AV, Thummerer A, Langendijk JA, Wagenaar D, Both S. Feasibility of Monte Carlo dropout-based uncertainty maps to evaluate deep learning-based synthetic CTs for adaptive proton therapy. Med Phys. 2024; 51(4): 2499-2509.

[23] Rusanov B, Hassan GM, Reynolds M, et al. Deep learning methods for enhancing cone-beam CT image quality toward adaptive radiation therapy: A systematic review. Med Phys. 2022; 49(9): 6019-6054.

[24] Low DA, Harms WB, Mutic S, Purdy JA. A technique for the quantitative evaluation of dose distributions. Med Phys. 1998; 25(5): 656-661.

[25] Wendling M, Zijp LJ, McDermott LN, et al. A fast algorithm for gamma evaluation in 3D. Med Phys. 2007; 34(5): 1647-1654.

[26] Hussein M, Clark CH, Nisbet A. Challenges in calculation of the gamma index in radiotherapy - Towards good practice. Phys Med. 2017; 36: 1-11.

[27] Hu N, Tanaka H, Kakino R, et al. Evaluation of a treatment planning system developed for clinical boron neutron capture therapy and validation against an independent Monte Carlo dose calculation system. Radiat Oncol. 2021; 16(1): 243.

[28] Li WG, Chang C, Qin Y, et al. Gpu-based cross-platform monte carlo proton dose calculation engine in the framework of taichi. Nucl Sci Tech. 2023; 34(5): 77

[29] Lee CM, Ye SJ. A GPU-accelerated Monte Carlo code, RT2for coupled transport of photon, electron/positron, and neutron. Phys Med Biol. 2024; 69(17).

[30] Babier A, Zhang B, Mahmood R, et al. OpenKBP: The open-access knowledge-based planning grand challenge and dataset. Med Phys. 2021; 48(9): 5549-5561.

[31] Technical Note: A cascade 3D U-Net for dose prediction in radiotherapy. Med Phys. 2021; 48(9): 5574-5582.

[32] Bourhis J, Montay-Gruel P, Jorge PG. Clinical translation of flash radio- therapy: Why and how? Radiot Oncol. 2019; 139: 11–17.

[33] Vozenin MC, Bourhis J, Durante M. Towards clinical translation of FLASH radiotherapy. Nat Rev Clin Oncol. 2022; 19(12): 791-803

[34] Fang W, Huang X, Tan J, et al. Proton linac-based therapy facility for ultra-high dose rate (flash) treatment. Nucl. Sci. Tech. 2021; 32(4): 12

[35] Zhang G, Wang J, Wang Y, Peng H. Proton FLASH: passive scattering or pencil beam scanning? Phys Med Biol. 2021; 66(3): 03NT01.

[36] Fan Y, Jiang D, Wang M, Song Z. A new markerless patient-to-image registration method using a portable 3D scanner. Med Phys. 2014; 41(10): 101910.

[37] Zhao YJ, Xiong YX, Wang Y. Three-Dimensional Accuracy of Facial Scan for Facial Deformities in Clinics: A New Evaluation Method for Facial Scanner Accuracy. PLoS One. 2017; 12(1): e0169402.

[38] Cho H, Lee JS, Kim JS, et al. Generating 3D images of VMAT plans for predictive models and activation maps associated with plan deliverability. Med Phys. 2024; 51(10): 7415-7424.

RIGHTS & PERMISSIONS

2025 The Author(s). Precision Radiation Oncology published by John Wiley & Sons Australia, Ltd on behalf of Shandong Cancer Hospital & Institute.