Deep Learning-Based Synthetic-CT Generation from MRI for Enhanced Precision in MRI-Only Radiotherapy Dose Planning

• Autorzy: Acquah Isaac Kwesi , Issahaku Shiraz , Tagoe Samuel Nii Adu , Sackey Theophilus Akumea

Identyfikatory

DOI

10.2478/pjmpe-2025-0025

• Języki publikacji: EN

Abstrakty

EN

Introduction: Radiotherapy aims to precisely target tumors while sparing healthy tissue, traditionally relying on CT imaging for accurate dose planning. However, CT has limitations in soft tissue contrast and exposes patients to ionizing radiation. MRI offers superior soft tissue contrast without radiation but lacks electron density information, restricting its use in dose planning. This study addresses this gap by developing deep learning models to generate pseudo-CT images from MRI, enabling MRI-only workflows in radiotherapy. Methodology: Paired MRI and CT scans from 12 subjects were processed using normalization, alignment, and masking. Four deep learning architectures (U-Net, Pix2Pix, CycleGAN, and conditional GAN (cGAN)) were trained to generate synthetic CT images from MRI data. Model performance was evaluated using metrics including mean absolute error (MAE), mean squared error (MSE), peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), and Pearson correlation coefficient (PCC). Results: Pix2Pix achieved the highest SSIM and PSNR, indicating strong structural preservation and reduced noise. It also had the lowest MAE and MSE, showing high accuracy in synthetic-CT generation. The cGAN model scored highest in PCC, highlighting its effective intensity alignment with real CT data. Statistical tests confirmed Pix2Pix’s superior performance, though CycleGAN and cGAN also showed notable results in alignment accuracy. Conclusion: Deep learning models, particularly Pix2Pix, can generate reliable pseudo-CT images from MRI, supporting MRI-only radiotherapy planning. This approach reduces radiation exposure and may streamline radiotherapy work-flows, offering a promising advance for patient-centered cancer care and MRI-only radiotherapy workflows.

Słowa kluczowe

EN

MRI-only radiotherapy; deep learning; synthetic CT images; cancer treatment planning

• Wydawca: Polskie Towarzystwo Fizyki Medycznej
• Czasopismo: Polish Journal of Medical Physics and Engineering
• Rocznik: 2025
• Tom: Vol. 31, Iss. 3
• Strony: 219--226
• Opis fizyczny: Bibliogr. 20 poz., rys., tab.

Twórcy

autor

Acquah Isaac Kwesi

• Department of Medical Physics, University of Ghana, Legon, Accra, Ghana
• Department of Physics Education, University of Education, Winneba, Ghan

• https://orcid.org/0000-0003-1117-1254

autor

Issahaku Shiraz

• Department of Medical Physics, University of Ghana, Legon, Accra, Ghana
• Ghana Atomic Energy Commission, Accra, Ghana

autor

Tagoe Samuel Nii Adu

• Department of Medical Physics, University of Ghana, Legon, Accra, Ghana
• Department of Radiation Oncology, National Centre for Radiotherapy and Nuclear Medicine, Korle-Bu Teaching Hospital, Guggisberg Avenue, Korle-Bu, Accra, Ghan

autor

Sackey Theophilus Akumea

• Ghana Atomic Energy Commission, Accra, Ghana

Bibliografia

• 1. Acquah IK, Inkoom S, Hasford F. Comparison of Methods of μ-Map Generation: MR-Based Method in PET/MR Imaging Versus Pseudo-CT Method in Radiotherapy Dose Planning. Iranian Journal of Medical Physics. 2024;21(6):355-364. doi:10.22038/ijmp.2024.77457.2369
• 2. Dinkla AM, Wolterink JM, Maspero M, et al. MR-Only Brain Radiation Therapy: Dosimetric Evaluation of Synthetic CTs Generated by a Dilated Convolutional Neural Network. International Journal of Radiation Oncology*Biology*Physics. 2018;102(4):801-812. doi:10.1016/j.ijrobp.2018.05.058
• 3. Xiang L, Wang Q, Nie D, et al. Deep embedding convolutional neural network for synthesizing CT image from T1-Weighted MR image. Medical Image Analysis. 2018;47:31-44. doi:10.1016/j.media.2018.03.011
• 4. Edmund JM, Nyholm T. A review of substitute CT generation for MRI-only radiation therapy. Radiat Oncol. 2017;12(1). doi:10.1186/s13014-016-0747-y
• 5. Singh A, Kwiecinski J, Cadet S, et al. Automated nonlinear registration of coronary PET to CT angiography using pseudo-CT generated from PET with generative adversarial networks. Journal of Nuclear Cardiology. 2023;30(2):604-615. doi:10.1007/s12350-022-03010-8
• 6. Ranjan A, Lalwani D, Misra R. GAN for synthesizing CT from T2-weighted MRI data towards MR-guided radiation treatment. Magn Reson Mater Phy. 2021;35(3):449-457. doi:10.1007/s10334-021-00974-5
• 7. Wolterink JM, Dinkla AM, Savenije MHF, Seevinck PR, van den Berg CAT, Išgum I. Deep MR to CT Synthesis Using Unpaired Data. Lecture Notes in Computer Science. Published online 2017:14-23. doi:10.1007/978-3-319-68127-6_2
• 8. Nie D, Trullo R, Lian J, et al. Medical Image Synthesis with Context-Aware Generative Adversarial Networks. Lecture Notes in Computer Science. Published online 2017:417-425. doi:10.1007/978-3-319-66179-7_48
• 9. Goodfellow IJ, Pouget-Abadie J, Mirza M, et al. Generative Adversarial Nets. In: Ghahramani Z, Welling M, Cortes C, Lawrence N, Weinberger KQ, eds. Advances in Neural Information Processing Systems. Vol 27. Curran Associates, Inc.; 2014. https://proceedings.neurips.cc/paper_files/paper/2014/file/f033ed80deb0234979a61f95710dbe25-Paper.pdf
• 10. Zhu JY, Park T, Isola P, Efros AA. Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks. 2017 IEEE International Conference on Computer Vision (ICCV). Published online October 2017:2242-2251. doi:10.1109/iccv.2017.244
• 11. Acquah IK, Issahaku S, Tagoe SNA. A systematic review of deep learning techniques for generating synthetic CT images from MRI data. Polish Journal of Medical Physics and Engineering. 2025;31(1):20-38. doi:10.2478/pjmpe-2025-0003
• 12. Tustison NJ, Avants BB, Cook PA, et al. N4ITK: Improved N3 Bias Correction. IEEE Trans Med Imaging. 2010;29(6):1310-1320. doi:10.1109/tmi.2010.2046908
• 13. Otsu N. A threshold selection method from gray-level histograms. Automatica. 1975;11(285-296):23-27
• 14. Ronneberger O, Fischer P, Brox T. U-Net: Convolutional Networks for Biomedical Image Segmentation. Lecture Notes in Computer Science. Published online 2015:234-241. doi:10.1007/978-3-319-24574-4_28
• 15. Liu L, Johansson A, Cao Y, Dow J, Lawrence TS, Balter JM. Abdominal synthetic CT generation from MR Dixon images using a U-net trained with ‘semi-synthetic’ CT data. Phys Med Biol. 2020;65(12):125001. doi:10.1088/1361-6560/ab8cd2
• 16. Rusanov B, Hassan GM, Reynolds M, et al. Transformer CycleGAN with uncertainty estimation for CBCT based synthetic CT in adaptive radiotherapy. Phys Med Biol. 2024;69(3):035014. doi:10.1088/1361-6560/ad1cfc
• 17. Maspero M, Bentvelzen LG, Savenije MHF, et al. Deep learning-based synthetic CT generation for paediatric brain MR-only photon and proton radiotherapy. Radiotherapy and Oncology. 2020;153:197-204. doi:10.1016/j.radonc.2020.09.029
• 18. Han X. MR-based synthetic CT generation using a deep convolutional neural network method. Med Phys. 2017;44(4):1408-1419. doi:10.1002/mp.12155
• 19. Kazemifar S, McGuire S, Timmerman R, et al. MRI-only brain radiotherapy: Assessing the dosimetric accuracy of synthetic CT images generated using a deep learning approach. Radiotherapy and Oncology. 2019;136:56-63. doi:10.1016/j.radonc.2019.03.026
• 20. Pan S, Abouei E, Wynne J, et al. Synthetic CT generation from MRI using 3D transformer‐based denoising diffusion model. Medical Physics. 2023;51(4):2538-2548. doi:10.1002/mp.16847