{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,24]],"date-time":"2026-09-24T12:01:50Z","timestamp":1790251310854,"version":"4.1.0"},"reference-count":31,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,9,29]],"date-time":"2025-09-29T00:00:00Z","timestamp":1759104000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Oncol."],"abstract":"<jats:sec><jats:title>Background<\/jats:title><jats:p>Magnetic resonance (MR) guided radiation therapy combines high-resolution image capabilities of MRI with the precise targeting of radiation therapy. However, MRI does not provide the essential electron density information for accurate dose calculation, which limit the application of MRI. In this presented work, we evaluated the potential for Deep Learning (DL) based synthetic CT (sCT) generation using 3D MRI setup scans acquired during real-time adaptive MRI-guided radiation therapy.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>We trained and evaluated a Cycle-consistent Generative Adversarial Network (Cycle-GAN) using paired MRI and deformably registered CT scan slices (dCT) in the context of real-time adaptive MRI-guided radiation therapy. Synthetic CT (sCT) volumes are output from the MR to CT generator of the Cycle-GAN network. A retrospective study was conducted to train and evaluate the DL model using data from patients undergoing treatment for kidney, pancreas, liver, lung, bone, and prostate tumors. Data was partitioned by patients using a stratified k-fold approach to ensure balanced representation of treatment sites in the training and testing sets. Synthetic CT images were evaluated using mean absolute error in Hounsfield Units (HU) relative to dCT, and four image quality metrics (mean absolute error, structural similarity index measure, peak signal-to-noise ratio, and normalized cross correlation) using the deformed CT scans as a reference standard. Synthetic CT volumes were also imported into a clinical treatment planning system and dosimetric calculations re-evaluated for each treatment plan (absolute difference in delivered dose to 3cm radius of PTV).<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>We trained the model using 8405 frames from 57 patients and evaluated it using a test set of 357 sCT frames from 17 patients. Quantitatively, sCTs were comparable to electron density of dCTs, while improving structural similarity with on-table MRI scans. The MAE between sCT and dCT was 49.2\u00b113.2 HU, sCT NCC outperformed dCT by 0.06, and SSIM and PSNR were 0.97\u00b10.01 and 19.9\u00b11.6 respectively. Furthermore, dosimetric evaluations revealed minimal differences between sCTs and dCTs. Qualitatively, superior reconstruction of air-bubbles in sCT compared to dCT reveal higher alignment between sCT than dCT with the associated MR.<\/jats:p><\/jats:sec><jats:sec><jats:title>Conclusions<\/jats:title><jats:p>Accuracy of deep learning based synthetic CT generation using setup scans on MR-Linacs was adequate for dose calculation\/optimization. This can enable MR-only treatment planning workflows on MR-Linacs, thereby increasing the efficiency of simulation and adaptive planning for MRgRT.<\/jats:p><\/jats:sec>","DOI":"10.3389\/fonc.2025.1672778","type":"journal-article","created":{"date-parts":[[2025,9,29]],"date-time":"2025-09-29T04:14:12Z","timestamp":1759119252000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":2,"title":["Dosimetric evaluations using cycle-consistent generative adversarial network synthetic CT for MR-guided adaptive radiation therapy"],"prefix":"10.3389","volume":"15","author":[{"given":"Gabriel L.","family":"Asher","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shiru","family":"Wang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bassem I.","family":"Zaki","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Gregory A.","family":"Russo","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Gobind S.","family":"Gill","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Charles R.","family":"Thomas","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Temiloluwa O.","family":"Prioleau","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yuting","family":"Li","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Rongxiao","family":"Zhang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yue","family":"Yan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Brady","family":"Hunt","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2025,9,29]]},"reference":[{"key":"B1","doi-asserted-by":"crossref","DOI":"10.1093\/neuonc\/nox006","article-title":"MR guided radiation therapy: transformative technology and its role in the central nervous system","volume":"19","author":"Cao","year":"2017","journal-title":"Neuro Oncol"},{"key":"B2","doi-asserted-by":"crossref","DOI":"10.1111\/1754-9485.12968","article-title":"Magnetic resonance-guided radiation therapy: a review","volume":"64","author":"Chin","year":"2020","journal-title":"J Med Imaging Radiat Oncol"},{"key":"B3","doi-asserted-by":"crossref","DOI":"10.1002\/mrm.25300","article-title":"Real-time active MR-tracking of metallic stylets in MR-guided radiation therapy","volume":"73","author":"Wang","year":"2015","journal-title":"Magn Reson Med"},{"key":"B4","doi-asserted-by":"crossref","DOI":"10.1016\/j.radonc.2015.01.015","article-title":"Benchmark IMRT evaluation of a Co-60 MRI-guided radiation therapy system","volume":"114","author":"Wooten","year":"2015","journal-title":"Radiother Oncol"},{"key":"B5","doi-asserted-by":"crossref","DOI":"10.1080\/0284186X.2018.1445284","article-title":"MRI-only based radiation therapy of prostate cancer: workflow and early clinical experience","volume":"57","author":"Tenhunen","year":"2018","journal-title":"Acta Oncol"},{"key":"B6","doi-asserted-by":"crossref","DOI":"10.1016\/j.radonc.2017.07.028","article-title":"Fast and robust online adaptive planning in stereotactic MR-guided adaptive radiation therapy (SMART) for pancreatic cancer","volume":"125","author":"Bohoudi","year":"2017","journal-title":"Radiother Oncol"},{"key":"B7","doi-asserted-by":"crossref","first-page":"20160667","DOI":"10.1259\/bjr.20160667","article-title":"The future of image-guided radiotherapy will be MR guided","volume":"90","author":"Pollard","year":"2017","journal-title":"Br J Radiol"},{"key":"B8","doi-asserted-by":"crossref","DOI":"10.2214\/AJR.11.7364","article-title":"Metal-induced artifacts in MRI","volume":"197","author":"Hargreaves","year":"2011","journal-title":"AJR Am J Roentgenol"},{"key":"B9","doi-asserted-by":"crossref","first-page":"28","DOI":"10.1186\/s13014-016-0747-y","article-title":"A review of substitute CT generation for MRI-only radiation therapy","volume":"12","author":"Edmund","year":"2017","journal-title":"Radiat Oncol"},{"key":"B10","doi-asserted-by":"crossref","first-page":"05TR01","DOI":"10.1088\/1361-6560\/aaaca4","article-title":"MRI-only treatment planning: benefits and challenges","volume":"63","author":"Owrangi","year":"2018","journal-title":"Phys Med Biol"},{"key":"B11","doi-asserted-by":"crossref","first-page":"20190067","DOI":"10.1259\/bjr.20190067","article-title":"MRI-based treatment planning for liver stereotactic body radiotherapy: validation of a deep learning-based synthetic CT generation method","volume":"92","author":"Liu","year":"2019","journal-title":"Br J Radiol"},{"key":"B12","doi-asserted-by":"crossref","first-page":"5218","DOI":"10.1002\/mp.13187","article-title":"Comparative study of algorithms for synthetic CT generation from MRI: consequences for MRI-guided radiation planning in the pelvic region","volume":"45","author":"Arabi","year":"2018","journal-title":"Med Phys"},{"key":"B13","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1186\/1748-717X-8-51","article-title":"MRI-based treatment plan simulation and adaptation for ion radiotherapy using a classification-based approach","volume":"8","author":"Rank","year":"2013","journal-title":"Radiat Oncol"},{"key":"B14","doi-asserted-by":"crossref","DOI":"10.1016\/j.media.2018.11.010","article-title":"A deep learning framework for unsupervised affine and deformable image registration","volume":"52","author":"De Vos","year":"2019","journal-title":"Med Image Anal"},{"key":"B15","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-030-00928-1_86","article-title":"Gdl-fire: Deep learning-based fast 4d ct image registration","volume-title":"International conference on medical image computing and computer-assisted intervention","author":"Sentker","year":"2018"},{"key":"B16","doi-asserted-by":"crossref","DOI":"10.21203\/rs.3.rs-464091\/v1","article-title":"Deep learning in brain MRI: effect of data leakage due to slice-level split using 2D convolutional neural networks","author":"Yagis","year":"2021"},{"key":"B17","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-319-66179-7_48","article-title":"Medical image synthesis with context-aware generative adversarial networks","author":"Nie","year":"2017"},{"key":"B18","doi-asserted-by":"crossref","DOI":"10.1109\/TMI.2020.3015379","article-title":"Unsupervised MR-to-CT synthesis using structure-constrained CycleGAN","volume":"39","author":"Yang","year":"2020","journal-title":"IEEE Trans Med Imaging"},{"key":"B19","doi-asserted-by":"crossref","DOI":"10.1002\/mp.16087","article-title":"Ensemble learning and personalized training for the improvement of unsupervised deep learning-based synthetic CT reconstruction","volume":"50","author":"Olberg","year":"2023","journal-title":"Med Phys"},{"key":"B20","first-page":"14","article-title":"Deep MR to CT synthesis using unpaired data","author":"Wolterink","year":"2017"},{"key":"B21","doi-asserted-by":"crossref","DOI":"10.1002\/mp.13617","article-title":"MRI-only based synthetic CT generation using dense cycle consistent generative adversarial networks","volume":"46","author":"Lei","year":"2019","journal-title":"Med Phys"},{"key":"B22","doi-asserted-by":"crossref","DOI":"10.1007\/s13534-021-00195-8","article-title":"Synthetic CT generation from weakly paired MR images using cycle-consistent GAN for MR-guided radiotherapy","volume":"11","author":"Kang","year":"2021","journal-title":"BioMed Eng Lett"},{"key":"B23","doi-asserted-by":"crossref","first-page":"93","DOI":"10.1002\/acm2.13327","article-title":"Deep learning-based synthetic CT generation for MR-only radiotherapy of prostate cancer patients with 0.35 T MRI linear accelerator","volume":"22","author":"Farjam","year":"2021","journal-title":"J Appl Clin Med Phys"},{"key":"B24","doi-asserted-by":"crossref","DOI":"10.1109\/ICCV.2017.244","article-title":"Unpaired image-to-image translation using cycle-consistent adversarial networks","author":"Zhu","year":"2017"},{"key":"B25","doi-asserted-by":"crossref","DOI":"10.1145\/3422622","article-title":"Generative adversarial networks","volume":"63","author":"Goodfellow","year":"2020","journal-title":"Commun ACM"},{"key":"B26","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR.2017.632","article-title":"Image-to-image translation with conditional adversarial networks","author":"Isola","year":"2017"},{"key":"B27","doi-asserted-by":"crossref","DOI":"10.1109\/ACSSC.2003.1292216","article-title":"Multiscale structural similarity for image quality assessment","volume-title":"The thrity-seventh asilomar conference on signals, systems & Computers, 2003","author":"Wang","year":"2003"},{"key":"B28","doi-asserted-by":"crossref","DOI":"10.1002\/mp.14866","article-title":"Improving generalization in MR-to-CT synthesis in radiotherapy by using an augmented cycle generative adversarial network with unpaired data","volume":"48","author":"Brou Boni","year":"2021","journal-title":"Med Phys"},{"key":"B29","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-030-00928-1_60","article-title":"Distribution matching losses can hallucinate features in medical image translation","author":"Cohen","year":"2018"},{"key":"B30","article-title":"Conversion between CT and MRI images using diffusion and score-matching models","author":"Lyu","year":"2022","journal-title":"arXiv preprint arXiv:2209.12104"},{"key":"B31","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3554729","article-title":"Diffusion models: A comprehensive survey of methods and applications. ACM","volume":"56","author":"Yang","year":"2023","journal-title":"Comput Surv"}],"container-title":["Frontiers in Oncology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fonc.2025.1672778\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,9,29]],"date-time":"2025-09-29T04:14:16Z","timestamp":1759119256000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fonc.2025.1672778\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,9,29]]},"references-count":31,"alternative-id":["10.3389\/fonc.2025.1672778"],"URL":"https:\/\/doi.org\/10.3389\/fonc.2025.1672778","relation":{},"ISSN":["2234-943X"],"issn-type":[{"value":"2234-943X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,9,29]]},"article-number":"1672778"}}