{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,23]],"date-time":"2026-09-23T07:01:29Z","timestamp":1790146889592,"version":"4.0.1"},"reference-count":37,"publisher":"IOP Publishing","issue":"5","license":[{"start":{"date-parts":[[2026,9,23]],"date-time":"2026-09-23T00:00:00Z","timestamp":1790121600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/publishingsupport.iopscience.iop.org\/iop-standard\/v1"},{"start":{"date-parts":[[2026,9,23]],"date-time":"2026-09-23T00:00:00Z","timestamp":1790121600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"},{"start":{"date-parts":[[2026,9,23]],"date-time":"2026-09-23T00:00:00Z","timestamp":1790121600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/publishingsupport.iopscience.iop.org\/questions\/alternative-author-rights-policies\/"},{"start":{"date-parts":[[2026,9,23]],"date-time":"2026-09-23T00:00:00Z","timestamp":1790121600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/publishingsupport.iopscience.iop.org\/questions\/alternative-author-rights-policies\/"}],"content-domain":{"domain":["iopscience.iop.org"],"crossmark-restriction":false},"short-container-title":["Biomed. Phys. Eng. Express"],"published-print":{"date-parts":[[2026,10,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Deep learning\u2013based MRI-to-CT synthesis supports MR-only radiotherapy and PET\/MR attenuation correction, but comparisons of U-Net variants are confounded by differences in datasets, preprocessing, training protocols, and evaluation metrics. This study performed a controlled, tissue-aware, and complexity-aware comparison of five U-Net-based architectures, including U-Net, ResU-Net, Attention U-Net, U-Net++, and attention deep residual U-Net (ADR-U-Net), for brain sCT generation from T1-weighted and FLAIR MRI in 37 subjects from the CERMEP-IDB-MRXFDG database. Subject-level five-fold cross-validation reduced split-dependent bias and prevented slice-level information leakage. Evaluation included mean absolute error (MAE), root mean square error (RMSE), peak signal-to-noise ratio, structural similarity index measure, tissue-specific errors in Hounsfield units (HU), paired Wilcoxon signed-rank tests with Bonferroni correction, and computational complexity analysis. ADR-U-Net achieved the lowest whole-image MAE of 37.38 \u00b1 6.01 HU and RMSE of 109.72 \u00b1 17.35 HU. Its MAE improvement was 1.60 HU relative to ResU-Net, indicating a modest gain over the closest comparator. A focused ADR-U-Net modality-ablation analysis showed lower whole-image MAE with dual-contrast input (37.38 \u00b1 6.01 HU) than with T1-only (40.79 \u00b1 5.00 HU) or FLAIR-only input (38.41 \u00b1 6.30 HU). Bone remained the dominant source of HU error, and U-Net++ had the highest number of floating-point operations without achieving the best accuracy. These findings suggest that model selection should consider tissue-specific HU errors, subject-level robustness, and computational cost. Because no external validation or downstream clinical endpoint was included, the results should be interpreted as controlled within-dataset technical evidence rather than evidence of clinical readiness.<\/jats:p>","DOI":"10.1088\/2057-1976\/aea580","type":"journal-article","created":{"date-parts":[[2026,9,10]],"date-time":"2026-09-10T22:47:55Z","timestamp":1789080475000},"page":"055023","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Evaluation of U-Net based architectures for synthetic CT generation from dual-contrast MRI"],"prefix":"10.1088","volume":"12","author":[{"ORCID":"https:\/\/orcid.org\/0009-0004-5418-5351","authenticated-orcid":true,"given":"Seyede Sajede Mousavi","family":"Eshkiki","sequence":"first","affiliation":[{"name":"University of Guilan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3239-1520","authenticated-orcid":true,"given":"Alireza","family":"Sadremomtaz","sequence":"additional","affiliation":[{"name":"University of Guilan"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"266","published-online":{"date-parts":[[2026,9,23]]},"reference":[{"key":"bpexaea580bib1","doi-asserted-by":"publisher","first-page":"3772","DOI":"10.1002\/mp.17867","type":"journal-article","article-title":"Generation of synthetic CT from MRI for MRI\u2010based attenuation correction of brain PET images using radiomics and machine learning","volume":"52","author":"Hoseinipourasl","year":"2025","journal-title":"Med. Phys."},{"key":"bpexaea580bib2","doi-asserted-by":"publisher","first-page":"52","DOI":"10.1186\/s40658-023-00569-0","type":"journal-article","article-title":"A review of PET attenuation correction methods for PET-MR","volume":"10","author":"Krokos","year":"2023","journal-title":"EJNMMI Phys."},{"key":"bpexaea580bib3","doi-asserted-by":"publisher","first-page":"20","DOI":"10.2478\/pjmpe-2025-0003","type":"journal-article","article-title":"A systematic review of deep learning techniques for generating synthetic CT images from MRI data","volume":"31","author":"Acquah","year":"2025","journal-title":"Pol. J. Med. Phys. Eng."},{"key":"bpexaea580bib4","doi-asserted-by":"publisher","first-page":"10076","DOI":"10.1109\/TPAMI.2024.3435571","type":"journal-article","article-title":"Medical image segmentation review: the success of U-net","volume":"46","author":"Azad","year":"2024","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"bpexaea580bib5","doi-asserted-by":"publisher","first-page":"1408","DOI":"10.1002\/mp.12155","type":"journal-article","article-title":"MR\u2010based synthetic CT generation using a deep convolutional neural network method","volume":"44","author":"Han","year":"2017","journal-title":"Med. Phys."},{"key":"bpexaea580bib6","doi-asserted-by":"publisher","first-page":"125629","DOI":"10.1109\/ACCESS.2024.3383880","type":"journal-article","article-title":"Image generation and lesion segmentation of brain tumors and stroke based on GAN and 3D ResU-Net","volume":"13","author":"Sun","year":"2024","journal-title":"IEEE Access"},{"key":"bpexaea580bib7","doi-asserted-by":"publisher","first-page":"117","DOI":"10.1002\/mp.15986","type":"journal-article","article-title":"SARU: a self\u2010attention ResUNet to generate synthetic CT images for MR\u2010only BNCT treatment planning","volume":"50","author":"Zhao","year":"2023","journal-title":"Med. Phys."},{"key":"bpexaea580bib8","doi-asserted-by":"publisher","first-page":"875","DOI":"10.3390\/diagnostics16060875","type":"journal-article","article-title":"An adaptive attention 3D U-Net for high-fidelity MRI-to-CT synthesis: bridging the anatomical gap with CBAM","volume":"16","author":"Bensebihi","year":"2026","journal-title":"Diagnostics"},{"key":"bpexaea580bib9","doi-asserted-by":"publisher","DOI":"10.3389\/fonc.2021.665807","type":"journal-article","article-title":"CT-based pelvic T1-weighted MR image synthesis using UNet, UNet++ and cycle-consistent generative adversarial network (Cycle-GAN)","volume":"11","author":"Kalantar","year":"2021","journal-title":"Front. Oncol."},{"key":"bpexaea580bib10","doi-asserted-by":"publisher","first-page":"477","DOI":"10.3390\/info16060477","type":"journal-article","article-title":"Towards MR-Only radiotherapy in head and neck: generation of synthetic CT from zero-TE MRI using deep learning","volume":"16","author":"Aouadi","year":"2025","journal-title":"Information"},{"key":"bpexaea580bib11","doi-asserted-by":"publisher","DOI":"10.1016\/j.compbiomed.2023.106738","type":"journal-article","article-title":"CT synthesis from multi-sequence MRI using adaptive fusion network","volume":"157","author":"Li","year":"2023","journal-title":"Comput. Biol. Med."},{"key":"bpexaea580bib12","doi-asserted-by":"publisher","first-page":"108","DOI":"10.1186\/s41983-021-00360-x","type":"journal-article","article-title":"Role of contrast-enhanced FLAIR MRI in diagnosis of intracranial lesions","volume":"57","author":"Mustafa","year":"2021","journal-title":"Egypt J. Neurol. Psychiatr. Neurosurg."},{"key":"bpexaea580bib13","doi-asserted-by":"publisher","first-page":"92","DOI":"10.1093\/jrr\/rrz063","type":"journal-article","article-title":"Feasibility of synthetic computed tomography generated with an adversarial network for multi-sequence magnetic resonance-based brain radiotherapy","volume":"61","author":"Koike","year":"2020","journal-title":"J. Radiat. Res."},{"key":"bpexaea580bib14","doi-asserted-by":"publisher","DOI":"10.1088\/2057-1976\/abe3a7","type":"journal-article","article-title":"Synthesizing CT images from MR images with deep learning: model generalization for different datasets through transfer learning","volume":"7","author":"Li","year":"2021","journal-title":"Biomed. Phys. Eng. Express"},{"key":"bpexaea580bib15","doi-asserted-by":"publisher","DOI":"10.1016\/j.radonc.2023.110052","type":"journal-article","article-title":"Deep learning MRI-only synthetic-CT generation for pelvis, brain and head and neck cancers","volume":"191","author":"Bird","year":"2024","journal-title":"Radiother. Oncol."},{"key":"bpexaea580bib16","doi-asserted-by":"publisher","DOI":"10.1002\/acm2.14499","type":"journal-article","article-title":"Advancements in synthetic CT generation from MRI: a review of techniques, and trends in radiation therapy planning","volume":"25","author":"Bahloul","year":"2024","journal-title":"J. Appl. Clin. Med. Phys."},{"key":"bpexaea580bib17","doi-asserted-by":"publisher","DOI":"10.3389\/fradi.2024.1385742","type":"journal-article","article-title":"A systematic literature review: deep learning techniques for synthetic medical image generation and their applications in radiotherapy","volume":"4","author":"Sherwani","year":"2024","journal-title":"Front. Radiol."},{"key":"bpexaea580bib18","doi-asserted-by":"publisher","DOI":"10.1016\/j.neuroimage.2020.117221","type":"journal-article","article-title":"AI-driven attenuation correction for brain PET\/MRI: clinical evaluation of a dementia cohort and importance of the training group size","volume":"222","author":"Ladefoged","year":"2020","journal-title":"Neuroimage"},{"key":"bpexaea580bib19","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2023.103046","type":"journal-article","article-title":"Deep learning based synthesis of MRI, CT and PET: review and analysis","volume":"92","author":"Dayarathna","year":"2024","journal-title":"Med. Image Anal."},{"key":"bpexaea580bib20","doi-asserted-by":"publisher","DOI":"10.1016\/j.compmedimag.2023.102241","type":"journal-article","article-title":"Investigation and benchmarking of U-Nets on prostate segmentation tasks","volume":"107","author":"Bhandary","year":"2023","journal-title":"Comput. Med. Imaging Graph."},{"key":"bpexaea580bib21","doi-asserted-by":"publisher","first-page":"203","DOI":"10.1038\/s41592-020-01008-z","type":"journal-article","article-title":"nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation","volume":"18","author":"Isensee","year":"2021","journal-title":"Nat. Methods"},{"key":"bpexaea580bib22","doi-asserted-by":"publisher","first-page":"91","DOI":"10.1186\/s13550-021-00830-6","type":"journal-article","article-title":"CERMEP-IDB-MRXFDG: a database of 37 normal adult human brain [18F] FDG PET, T1 and FLAIR MRI, and CT images available for research","volume":"11","author":"M\u00e9rida","year":"2021","journal-title":"EJNMMI Res."},{"key":"bpexaea580bib23","author":"Ronneberger","year":"2015","type":"conference-proceedings"},{"key":"bpexaea580bib24","type":"conference-proceedings","article-title":"Deep residual learning for image recognition","author":"He","year":"2016","edition":"(eds)"},{"key":"bpexaea580bib25","article-title":"Attention u-net: learning where to look for the pancreas","author":"Oktay","year":"2018","type":"preprint"},{"key":"bpexaea580bib26","doi-asserted-by":"publisher","DOI":"10.1016\/j.ejmp.2024.103315","type":"journal-article","article-title":"Attention-based deep neural network for partial volume correction in brain 18F-FDG PET imaging","volume":"119","author":"Azimi","year":"2024","journal-title":"Phys. Med."},{"key":"bpexaea580bib27","type":"book","article-title":"UNet++: a nested U-Net architecture for medical image segmentation","author":"Zhou","year":"2018","edition":"(eds)"},{"key":"bpexaea580bib28","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6560\/aac763","type":"journal-article","article-title":"Attenuation correction for brain PET imaging using deep neural network based on Dixon and ZTE MR images","volume":"63","author":"Gong","year":"2018","journal-title":"Phys. Med. Biol."},{"key":"bpexaea580bib29","doi-asserted-by":"publisher","first-page":"2896","DOI":"10.1109\/TUFFC.2022.3198522","type":"journal-article","article-title":"Classical and learned MR to pseudo-CT mappings for accurate transcranial ultrasound simulation","volume":"69","author":"Miscouridou","year":"2022","journal-title":"IEEE Trans. Ultrasonics, Ferroelectrics, and Frequency Control"},{"key":"bpexaea580bib30","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2024.103276","type":"journal-article","article-title":"Generating synthetic computed tomography for radiotherapy: synthRAD2023 challenge report","volume":"97","author":"Huijben","year":"2024","journal-title":"Med. Image Anal."},{"key":"bpexaea580bib31","article-title":"Deep learning-based synthetic CT generation from MR images: comparison of generative adversarial and residual neural networks","author":"Gholamiankhah","year":"2021","type":"preprint"},{"key":"bpexaea580bib32","doi-asserted-by":"publisher","first-page":"56","DOI":"10.1016\/j.radonc.2019.03.026","type":"journal-article","article-title":"MRI-only brain radiotherapy: assessing the dosimetric accuracy of synthetic CT images generated using a deep learning approach","volume":"136","author":"Kazemifar","year":"2019","journal-title":"Radiother. Oncol."},{"key":"bpexaea580bib33","doi-asserted-by":"publisher","DOI":"10.3390\/app122211600","type":"journal-article","article-title":"How to pseudo-CT: a comparative review of deep convolutional neural network architectures for CT synthesis","volume":"12","author":"Vera-olmos","year":"2022","journal-title":"Appl. Sci."},{"key":"bpexaea580bib34","doi-asserted-by":"publisher","first-page":"182","DOI":"10.1186\/s13014-023-02349-7","type":"journal-article","article-title":"Comprehensive evaluation of similarity between synthetic and real CT images for nasopharyngeal carcinoma","volume":"18","author":"Yuan","year":"2023","journal-title":"Radiat. Oncol."},{"key":"bpexaea580bib35","doi-asserted-by":"publisher","first-page":"23NT03","DOI":"10.1088\/1361-6560\/abc5cb","type":"journal-article","article-title":"Comparison of deep learning synthesis of synthetic CTs using clinical MRI inputs","volume":"65","author":"Massa","year":"2020","journal-title":"Phys. Med. Biol."},{"key":"bpexaea580bib36","article-title":"Enhancing CT image synthesis from multi-modal MRI data based on a multi-task neural network framework","author":"Xin","year":"2023","type":"preprint"},{"key":"bpexaea580bib37","doi-asserted-by":"publisher","DOI":"10.1002\/mp.17981","type":"journal-article","article-title":"SynthRAD2025 grand challenge dataset: generating synthetic CTs for radiotherapy from head to abdomen","volume":"52","author":"Thummerer","year":"2025","journal-title":"Med. Phys."}],"container-title":["Biomedical Physics &amp; Engineering Express"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/aea580","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/aea580\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/aea580","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/aea580\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/aea580\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/aea580\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/aea580\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/aea580\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,23]],"date-time":"2026-09-23T06:17:33Z","timestamp":1790144253000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/aea580"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,23]]},"references-count":37,"journal-issue":{"issue":"5","published-online":{"date-parts":[[2026,9,23]]},"published-print":{"date-parts":[[2026,10,1]]}},"URL":"https:\/\/doi.org\/10.1088\/2057-1976\/aea580","relation":{},"ISSN":["2057-1976"],"issn-type":[{"value":"2057-1976","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,23]]},"assertion":[{"value":"Evaluation of U-Net based architectures for synthetic CT generation from dual-contrast MRI","name":"article_title","label":"Article Title"},{"value":"Biomedical Physics & Engineering Express","name":"journal_title","label":"Journal Title"},{"value":"paper","name":"article_type","label":"Article Type"},{"value":"\u00a9 2026 IOP Publishing Ltd. All rights, including for text and data mining, AI training, and similar technologies, are reserved.","name":"copyright_information","label":"Copyright Information"},{"value":"2025-12-29","name":"date_received","label":"Date Received","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2026-09-10","name":"date_accepted","label":"Date Accepted","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2026-09-23","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}