{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,15]],"date-time":"2026-10-15T06:44:43Z","timestamp":1792046683331,"version":"4.4.0"},"reference-count":28,"publisher":"IOP Publishing","issue":"19","license":[{"start":{"date-parts":[[2026,9,30]],"date-time":"2026-09-30T00:00:00Z","timestamp":1790726400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"},{"start":{"date-parts":[[2026,9,30]],"date-time":"2026-09-30T00:00:00Z","timestamp":1790726400000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"}],"content-domain":{"domain":["iopscience.iop.org"],"crossmark-restriction":false},"short-container-title":["Phys. Med. Biol."],"published-print":{"date-parts":[[2026,10,14]]},"abstract":"Abstract<\/jats:title>\n \n Objective.<\/jats:italic>\n Computed tomography (CT) field-of-view (FOV) truncation leaves peripheral anatomy entirely absent, introducing systematic errors in radiotherapy dose calculation and body-composition analysis. Existing methods either rely on unavailable raw projection data or suffer from over-smoothing and hallucination under severe truncation.\n Approach.<\/jats:italic>\n We propose the self-cascade latent Schr\u00f6dinger bridge (SCL-SB), combining a latent image-to-image Schr\u00f6dinger bridge (I\n \n \n \n <\/jats:tex-math>\n \n \n \n \n 2<\/mml:mn>\n <\/mml:msup>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n SB) with a self-cascade mechanism in a single shared-weight U-Net. I\n \n \n \n <\/jats:tex-math>\n \n \n \n \n 2<\/mml:mn>\n <\/mml:msup>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n SB constructs a diffusion bridge from the truncated-image distribution to the full-FOV distribution, enabling high-quality reconstruction in ten denoising steps (\n \n \n \n <\/jats:tex-math>\n \n \n \n \u2248<\/mml:mo>\n <\/mml:mrow>\n 29<\/mml:mn>\n \u00d7<\/mml:mo>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n faster than a 100-step denoising diffusion implicit model baseline). The self-cascade mechanism runs two sequential rounds sharing identical weights: the first produces an initial extrapolation; the second refines it using the first-round output as a hint. A frequency-aware residual mechanism amplifies gradient contributions at high-frequency inter-round residuals to stabilise cascade training and recover anatomically faithful detail.The truncation radius is sampled from a stratified distribution spanning mild to severe truncation during training, enabling a single model to cover the full severity spectrum without radius-specific fine-tuning.\n Main results.<\/jats:italic>\n On a head-to-chest CT dataset spanning three truncation severities, SCL-SB, trained as a single model without radius-specific tuning, achieves the highest structural similarity index measure (SSIM) and a high-frequency ratio (HF-Ratio) closest to ideal among all compared methods at every severity, running nearly\n \n \n \n <\/jats:tex-math>\n \n \n 9<\/mml:mn>\n \u00d7<\/mml:mo>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n faster than the latest vision transformer-based baseline. Ablation shows that cascade training alone already improves the first round\u2019s output without running the second round at inference, an effect we term cascade implicit enhancement. Relative to the same architecture trained without cascade (I\n \n \n \n <\/jats:tex-math>\n \n \n \n \n 2<\/mml:mn>\n <\/mml:msup>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n SB), the model recovers more high-frequency anatomical detail (HF-Ratio\n \n \n \n <\/jats:tex-math>\n \n \n +<\/mml:mo>\n 45.3<\/mml:mn>\n %<\/mml:mi>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n ) and achieves a higher SSIM by\n \n \n \n <\/jats:tex-math>\n \n \n 0.076<\/mml:mn>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n (\n \n \n \n <\/jats:tex-math>\n \n \n 9.0<\/mml:mn>\n %<\/mml:mi>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n ) at severe truncation, at no additional inference cost.\n Significance.<\/jats:italic>\n SCL-SB addresses the posterior-mean regression that blurs high-frequency anatomical detail in ill-posed FOV extrapolation, produces spatially continuous Hounsfield unit reconstructions free of the patch-boundary discontinuities introduced by Vision-transformer-based architectures, and avoids the spectral hallucination characteristic of generative adversarial network-based methods. Although downstream clinical validation on larger, multi-centre cohorts remains necessary, these properties can potentially improve image guidance in clinical applications such as adaptive radiotherapy, body-composition assessment, and spine surgery.\n <\/jats:p>","DOI":"10.1088\/1361-6560\/aea5d0","type":"journal-article","created":{"date-parts":[[2026,9,10]],"date-time":"2026-09-10T22:53:59Z","timestamp":1789080839000},"page":"195012","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Self-cascade latent Schr\u00f6dinger bridge for CT field-of-view extension"],"prefix":"10.1088","volume":"71","author":[{"ORCID":"https:\/\/orcid.org\/0009-0006-8874-5649","authenticated-orcid":true,"given":"Xihao","family":"Wu","sequence":"first","affiliation":[{"name":"Peking University Health Science Center"},{"name":"Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"conceptualization","vocabulary":"credit"},{"role":"data-curation","vocabulary":"credit"},{"role":"formal-analysis","vocabulary":"credit"},{"role":"investigation","vocabulary":"credit"},{"role":"methodology","vocabulary":"credit"},{"role":"software","vocabulary":"credit"},{"role":"validation","vocabulary":"credit"},{"role":"visualization","vocabulary":"credit"},{"role":"writing-original-draft","vocabulary":"credit"},{"role":"writing-review-editing","vocabulary":"credit"}]},{"ORCID":"https:\/\/orcid.org\/0009-0005-8850-1102","authenticated-orcid":false,"given":"Jiacheng","family":"Liu","sequence":"additional","affiliation":[{"name":"Beihang University"},{"name":"Peking University Cancer Hospital & Institute"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"data-curation","vocabulary":"credit"},{"role":"writing-review-editing","vocabulary":"credit"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1596-7962","authenticated-orcid":false,"given":"Haijun","family":"Yu","sequence":"additional","affiliation":[{"name":"Peking University Health Science Center"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"validation","vocabulary":"credit"},{"role":"writing-review-editing","vocabulary":"credit"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9550-5284","authenticated-orcid":true,"given":"Andreas","family":"Maier","sequence":"additional","affiliation":[{"name":"Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"project-administration","vocabulary":"credit"},{"role":"supervision","vocabulary":"credit"},{"role":"writing-review-editing","vocabulary":"credit"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2627-3077","authenticated-orcid":true,"given":"Yixing","family":"Huang","sequence":"additional","affiliation":[{"name":"Peking University Health Science Center"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"conceptualization","vocabulary":"credit"},{"role":"project-administration","vocabulary":"credit"},{"role":"resources","vocabulary":"credit"},{"role":"supervision","vocabulary":"credit"},{"role":"writing-review-editing","vocabulary":"credit"}]}],"member":"266","published-online":{"date-parts":[[2026,9,30]]},"reference":[{"key":"pmbaea5d0bib1","doi-asserted-by":"publisher","first-page":"1798","DOI":"10.1109\/TPAMI.2013.50","type":"journal-article","article-title":"Representation learning: a review and new perspectives","volume":"35","author":"Bengio","year":"2013","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"pmbaea5d0bib2","first-page":"pp 6228","type":"conference-proceedings","article-title":"The perception-distortion tradeoff","author":"Blau","year":"2018"},{"key":"pmbaea5d0bib3","article-title":"An image is worth 16x16 words: transformers for image recognition at scale","author":"Dosovitskiy","year":"2021","type":"conference-proceedings"},{"key":"pmbaea5d0bib4","doi-asserted-by":"publisher","DOI":"10.1016\/j.cmpb.2023.107767","type":"journal-article","article-title":"A transformer-based dual-domain network for reconstructing FOV extended cone-beam CT images from truncated sinograms in radiation therapy","volume":"241","author":"Gao","year":"2023","journal-title":"Comput. Methods Programs Biomed."},{"key":"pmbaea5d0bib5","first-page":"pp 15979","type":"conference-proceedings","article-title":"Masked autoencoders are scalable vision learners","author":"He","year":"2022"},{"key":"pmbaea5d0bib6","first-page":"pp 6840","type":"conference-proceedings","article-title":"Denoising diffusion probabilistic models","volume":"vol 33","author":"Ho","year":"2020"},{"key":"pmbaea5d0bib7","doi-asserted-by":"publisher","first-page":"2385","DOI":"10.1118\/1.1776673","type":"journal-article","article-title":"A novel reconstruction algorithm to extend the CT scan field-of-view","volume":"31","author":"Hsieh","year":"2004","journal-title":"Med. Phys."},{"key":"pmbaea5d0bib8","doi-asserted-by":"publisher","first-page":"186","DOI":"10.1007\/978-3-658-29267-6_40","type":"journal-article","article-title":"Field of view extension in computed tomography using deep learning prior","volume":"2020","author":"Huang","year":"2020","journal-title":"Bildverarb. f\u00fcr Med."},{"key":"pmbaea5d0bib9","doi-asserted-by":"publisher","first-page":"3042","DOI":"10.1109\/TMI.2021.3072568","type":"journal-article","article-title":"Data extrapolation from learned prior images for truncation correction in computed tomography","volume":"40","author":"Huang","year":"2021","journal-title":"IEEE Trans. Med. Imaging"},{"key":"pmbaea5d0bib10","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":"pmbaea5d0bib11","first-page":"pp 1125","type":"conference-proceedings","article-title":"Image-to-image translation with conditional adversarial networks","author":"Isola","year":"2017"},{"key":"pmbaea5d0bib12","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2023.102846","type":"journal-article","article-title":"Diffusion models in medical imaging: a comprehensive survey","volume":"88","author":"Kazerouni","year":"2023","journal-title":"Med. Image Anal."},{"key":"pmbaea5d0bib13","doi-asserted-by":"publisher","first-page":"e245","DOI":"10.2196\/jmir.2930","type":"journal-article","article-title":"The virtual skeleton database: an open access repository for biomedical research and collaboration","volume":"15","author":"Kistler","year":"2013","journal-title":"J. Med. Internet Res."},{"key":"pmbaea5d0bib14","doi-asserted-by":"publisher","DOI":"10.1002\/mp.17898","type":"journal-article","article-title":"Diffusion Schr\u00f6dinger bridge models for high-quality MR-to-CT synthesis for proton treatment planning","volume":"52","author":"Li","year":"2025","journal-title":"Med. Phys."},{"key":"pmbaea5d0bib15","first-page":"pp 36115","type":"conference-proceedings","article-title":"Back to basics: let denoising generative models denoise","author":"Li","year":"2026"},{"key":"pmbaea5d0bib16","doi-asserted-by":"publisher","DOI":"10.1109\/TRPMS.2026.3705435","type":"conference-proceedings","article-title":"Efficient image-to-image Schr\u00f6dinger bridge for CT field of view extension","author":"Li","year":"2025"},{"key":"pmbaea5d0bib17","first-page":"pp 21274","type":"conference-proceedings","article-title":"I2SB: image-to-image Schr\u00f6dinger bridge","author":"Liu","year":"2023"},{"key":"pmbaea5d0bib18","doi-asserted-by":"publisher","first-page":"39","DOI":"10.1118\/1.598855","type":"journal-article","article-title":"Efficient correction for CT image artifacts caused by objects extending outside the scan field of view","volume":"27","author":"Ohnesorge","year":"2000","journal-title":"Med. Phys."},{"key":"pmbaea5d0bib19","doi-asserted-by":"publisher","first-page":"7460","DOI":"10.3390\/s24237460","type":"journal-article","article-title":"Unsupervised deep learning for synthetic CT generation from CBCT images for proton and carbon ion therapy for paediatric patients","volume":"24","author":"Pepa","year":"2024","journal-title":"Sensors"},{"key":"pmbaea5d0bib20","first-page":"pp 10684","type":"conference-proceedings","article-title":"High-resolution image synthesis with latent diffusion models","author":"Rombach","year":"2022"},{"key":"pmbaea5d0bib21","doi-asserted-by":"publisher","first-page":"459","DOI":"10.1088\/0031-9155\/45\/2\/314","type":"journal-article","article-title":"Correlation between ct numbers and tissue parameters needed for Monte Carlo simulations of clinical dose distributions","volume":"45","author":"Schneider","year":"2000","journal-title":"Phys. Med. Biol."},{"key":"pmbaea5d0bib22","doi-asserted-by":"publisher","first-page":"4709","DOI":"10.1002\/mp.16994","type":"journal-article","article-title":"Impact of cone-beam computed tomography artifacts on dose calculation accuracy for lung cancer","volume":"51","author":"Schr\u00f6der","year":"2024","journal-title":"Med. Phys."},{"key":"pmbaea5d0bib23","article-title":"Score-based generative modeling through stochastic differential equations (SDE)","author":"Song","year":"2021","type":"conference-proceedings"},{"key":"pmbaea5d0bib24","doi-asserted-by":"publisher","DOI":"10.1002\/acm2.13961","type":"journal-article","article-title":"A roadmap for implementation of kv-CBCT online adaptive radiation therapy and initial first year experiences","volume":"24","author":"Stanley","year":"2023","journal-title":"J. Appl. Clin. Med. Phys."},{"key":"pmbaea5d0bib25","first-page":"pp 3172","type":"conference-proceedings","article-title":"Resolution-robust large mask inpainting with Fourier convolutions","author":"Suvorov","year":"2022"},{"key":"pmbaea5d0bib26","doi-asserted-by":"publisher","first-page":"1661","DOI":"10.1162\/NECO_a_00142","type":"journal-article","article-title":"A connection between score matching and denoising autoencoders","volume":"23","author":"Vincent","year":"2011","journal-title":"Neural Comput."},{"key":"pmbaea5d0bib27","doi-asserted-by":"publisher","first-page":"600","DOI":"10.1109\/TIP.2003.819861","type":"journal-article","article-title":"Image quality assessment: from error visibility to structural similarity","volume":"13","author":"Wang","year":"2004","journal-title":"IEEE Trans. Image Process."},{"key":"pmbaea5d0bib28","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2023.102852","type":"journal-article","article-title":"Body composition assessment with limited field-of-view computed tomography: a semantic image extension perspective","volume":"88","author":"Xu","year":"2023","journal-title":"Med. Image Anal."}],"container-title":["Physics in Medicine & Biology"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aea5d0","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aea5d0\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aea5d0","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aea5d0\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aea5d0\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aea5d0\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aea5d0\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aea5d0\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,15]],"date-time":"2026-10-15T05:57:23Z","timestamp":1792043843000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aea5d0"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,30]]},"references-count":28,"journal-issue":{"issue":"19","published-online":{"date-parts":[[2026,9,30]]},"published-print":{"date-parts":[[2026,10,14]]}},"URL":"https:\/\/doi.org\/10.1088\/1361-6560\/aea5d0","relation":{},"ISSN":["0031-9155","1361-6560"],"issn-type":[{"value":"0031-9155","type":"print"},{"value":"1361-6560","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,30]]},"assertion":[{"value":"Self-cascade latent Schr\u00f6dinger bridge for CT field-of-view extension","name":"article_title","label":"Article Title"},{"value":"Physics in Medicine & Biology","name":"journal_title","label":"Journal Title"},{"value":"paper","name":"article_type","label":"Article Type"},{"value":"\u00a9 2026 The Author(s). Published on behalf of Institute of Physics and Engineering in Medicine by IOP Publishing Ltd","name":"copyright_information","label":"Copyright Information"},{"value":"2026-05-19","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":"2026-09-30","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}