{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,11,7]],"date-time":"2026-11-07T04:02:06Z","timestamp":1794024126067,"version":"4.8.0"},"reference-count":22,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2026,1,19]],"date-time":"2026-01-19T00:00:00Z","timestamp":1768780800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,2,11]],"date-time":"2026-02-11T00:00:00Z","timestamp":1770768000000},"content-version":"vor","delay-in-days":23,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"name":"NIH\/NCI"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Cancer Imaging"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Current MRI prostate cancer risk assessment methods focus mainly on detecting tumor lesions, ignoring the prostate gland macro-environment which may also impact disease progression. A generalized deep-learning model for prostate may help capture these gland-level characteristics through deep embeddings which can be used for a variety of downstream tasks. This study aims to assess whether U-Found, a generalized multiparametric (mp)MRI-based model, offers added value in predicting histopathological progression in active surveillance (AS) patients. The prostate macro-environment, captured in U-Found embeddings, is hypnotized to play a significant role in differentiating patients who progress to definitive treatment from those whose tumor is kept at bay.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>\n                      U-Found was trained on a dataset comprising over 3000 mpMRIs from in-house and public sources using self-supervised learning. Axial slices were represented in a 128-dimentional space. The physical interpretation of the embeddings was explored by investigating images that are closest to the centroid of embeddings clusters. U-Found was tested on a downstream task: identifying cancer in an independent dataset (publicly available UCLA dataset,\n                      <jats:italic>n<\/jats:italic>\n                      \u2009=\u20091,151). To determine the added value of U-Found embeddings to clinical and intratumoral radiomics features, we compared models for predicting histopathological progression in 144 participants of a prospective AS trial. In addition, associations between U-Found embeddings and lesion- and prostate radiomics were investigated.\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Our findings suggest that U-Found captures key characteristics of the prostate gland\u2019s macro-environment. U-Found successfully detected cancer in an independent UCLA dataset without being explicitly trained for lesion detection (AUC\u2009=\u20090.79). The prediction model incorporating a combination of clinical variables, mpMRI-derived intratumoral radiomics features and deep embeddings generated by U-Found achieved AUC\u2009=\u20090.86, outperforming models solely based on clinical and\/or radiomics features. There were clear associations between U-Found embeddings and radiomics features.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>U-Found was designed as a generalized self-supervised foundation model for prostate imaging, enabling the model to learn intrinsic imaging structures. We demonstrate that U-Found embeddings capture key features of the prostate macro-environment, which appear to contribute to disease progression, albeit to a lesser extent than tumor-specific imaging features.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1186\/s40644-026-00988-z","type":"journal-article","created":{"date-parts":[[2026,1,19]],"date-time":"2026-01-19T09:14:00Z","timestamp":1768814040000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":3,"title":["MRI-based patient selection for active surveillance in prostate cancer using U-Found: a generalized deep learning model"],"prefix":"10.1186","volume":"26","author":[{"given":"Noah C.","family":"Lowry","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Adrian L.","family":"Breto","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Veronica","family":"Wallaengen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ahmad","family":"Algohary","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nicolas","family":"Tapia-Stoll","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sandra M.","family":"Gaston","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nachiketh S.","family":"Prakash","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Pedro F. S.","family":"Freitas","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Oleksandr N.","family":"Kryvenko","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Patricia","family":"Castillo","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Joel","family":"Saltz","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tahsin","family":"Kurc","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chad R.","family":"Ritch","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bruno","family":"Nahar","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mark L.","family":"Gonzalgo","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Dipen J.","family":"Parekh","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Brandon","family":"Mahal","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Benjamin O.","family":"Spieler","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alan Dal","family":"Pra","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Matthew C.","family":"Abramowitz","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alan","family":"Pollack","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sanoj","family":"Punnen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Radka","family":"Stoyanova","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,1,19]]},"reference":[{"key":"988_CR1","doi-asserted-by":"publisher","first-page":"390","DOI":"10.1001\/jama.2014.17942","volume":"313","author":"MM Siddiqui","year":"2015","unstructured":"Siddiqui MM, Rais-Bahrami S, Turkbey B, et al. Comparison of MR\/ultrasound fusion-guided biopsy with ultrasound-guided biopsy for the diagnosis of prostate cancer. JAMA. 2015;313:390\u20137.","journal-title":"JAMA"},{"key":"988_CR2","doi-asserted-by":"crossref","unstructured":"Weinreb JC, Barentsz JO, Choyke PL, et al. PI-RADS prostate imaging - reporting and data system: 2015, version 2. Eur Urol. 2016;69:16\u201340.","DOI":"10.1016\/j.eururo.2015.08.052"},{"key":"988_CR3","doi-asserted-by":"publisher","first-page":"675","DOI":"10.1016\/j.eururo.2019.12.017","volume":"77","author":"JJ Tosoian","year":"2020","unstructured":"Tosoian JJ, Mamawala M, Epstein JI, et al. Active surveillance of grade group 1 prostate cancer: Long-term outcomes from a large prospective cohort. Eur Urol. 2020;77:675\u201382.","journal-title":"Eur Urol"},{"key":"988_CR4","doi-asserted-by":"publisher","first-page":"537","DOI":"10.1016\/j.euo.2020.11.007","volume":"5","author":"CE Chu","year":"2022","unstructured":"Chu CE, Cowan JE, Lonergan PE, et al. Diagnostic accuracy and prognostic value of serial prostate multiparametric magnetic resonance imaging in men on active surveillance for prostate cancer. Eur Urol Oncol. 2022;5:537\u201343.","journal-title":"Eur Urol Oncol"},{"key":"988_CR5","doi-asserted-by":"publisher","first-page":"e231439","DOI":"10.1001\/jamanetworkopen.2023.1439","volume":"6","author":"MR Cooperberg","year":"2023","unstructured":"Cooperberg MR, Meeks W, Fang R, Gaylis FD, Catalona WJ, Makarov DV. Time trends and variation in the use of active surveillance for management of Low-risk prostate cancer in the US. JAMA Netw Open. 2023;6:e231439.","journal-title":"JAMA Netw Open"},{"key":"988_CR6","doi-asserted-by":"publisher","first-page":"879","DOI":"10.1016\/S1470-2045(24)00220-1","volume":"25","author":"A Saha","year":"2024","unstructured":"Saha A, Bosma JS, Twilt JJ, et al. Artificial intelligence and radiologists in prostate cancer detection on MRI (PI-CAI): an international, paired, non-inferiority, confirmatory study. Lancet Oncol. 2024;25:879\u201387.","journal-title":"Lancet Oncol"},{"key":"988_CR7","unstructured":"Chen T, Kornblith S, Norouzi M, Hinton G. A simple framework for contrastive learning of visual representations. International Conference on Machine Learning. PMLR; 2020. pp. 1597\u2009\u2013\u2009607."},{"key":"988_CR8","doi-asserted-by":"crossref","unstructured":"Bosma JS, Saha A, Hosseinzadeh M, Slootweg I, de Rooij M, Huisman H. Semisupervised learning with report-guided pseudo labels for deep learning-based prostate cancer detection using biparametric MRI. Radiol Artif Intell. 2023;5.","DOI":"10.1148\/ryai.230031"},{"key":"988_CR9","unstructured":"Cardoso MJ, Li W, Brown R, et al. MONAI: an open-source framework for deep learning in healthcare. arXiv Preprint. 2022. arXiv:221102701."},{"key":"988_CR10","first-page":"399","volume":"944","author":"JW Johnson","year":"2020","unstructured":"Johnson JW. Automatic nucleus segmentation with Mask-RCNN. Adv Intell Syst Comp. 2020;944:399\u2013407.","journal-title":"Adv Intell Syst Comp"},{"key":"988_CR11","doi-asserted-by":"crossref","unstructured":"Akiba T, Sano S, Yanase T, Ohta T, Koyama M, Optuna: a next-generation hyperparameter optimization framework. Kdd\u201919: Proceedings of the 25th Acm Sigkdd International Conferencce on Knowledge Discovery and Data Mining. 2019. p. 2623-31.","DOI":"10.1145\/3292500.3330701"},{"key":"988_CR12","unstructured":"Paszke A, Gross S, Massa F, et al. PyTorch: an imperative style, high-performance deep learning library. Advances in Neural Information Processing Systems. 2019;32."},{"key":"988_CR13","doi-asserted-by":"crossref","unstructured":"McInnes L, Healy J, Melville J. UMAP: uniform manifold approximation and projection for dimension reduction. arXiv Preprint. 2018. arXiv:180203426.","DOI":"10.21105\/joss.00861"},{"key":"988_CR14","first-page":"7937","volume":"10","author":"S Natarajan","year":"2020","unstructured":"Natarajan S, Priester A, Margolis D, Huang J, Marks L. Prostate MRI and ultrasound with pathology and coordinates of tracked biopsy (prostate-MRI-US-biopsy). Cancer Imaging Arch. 2020;10:7937.","journal-title":"Cancer Imaging Arch"},{"key":"988_CR15","doi-asserted-by":"publisher","first-page":"41","DOI":"10.1016\/j.eururo.2015.08.038","volume":"69","author":"JO Barentsz","year":"2016","unstructured":"Barentsz JO, Weinreb JC, Verma S, et al. Synopsis of the PI-RADS v2 guidelines for multiparametric prostate magnetic resonance imaging and recommendations for use. Eur Urol. 2016;69:41\u20139.","journal-title":"Eur Urol"},{"key":"988_CR16","first-page":"034502","volume":"5","author":"D Kwon","year":"2018","unstructured":"Kwon D, Reis IM, Breto AL, et al. Classification of suspicious lesions on prostate multiparametric MRI using machine learning. J Med Imaging. 2018;5:034502.","journal-title":"J Med Imaging"},{"key":"988_CR17","doi-asserted-by":"publisher","first-page":"1226","DOI":"10.1109\/TPAMI.2005.159","volume":"27","author":"H Peng","year":"2005","unstructured":"Peng H, Long F, Ding C. Feature selection based on mutual information: criteria of max-dependency, max-relevance, and min-redundancy. IEEE Trans Pattern Anal Mach Intell. 2005;27:1226\u201338.","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"988_CR18","doi-asserted-by":"publisher","first-page":"e240597","DOI":"10.1148\/radiol.240597","volume":"314","author":"M Paschali","year":"2025","unstructured":"Paschali M, Chen Z, Blankemeier L, et al. Foundation models in radiology: What, How, why, and why not. Radiology. 2025;314:e240597.","journal-title":"Radiology"},{"key":"988_CR19","doi-asserted-by":"publisher","first-page":"863","DOI":"10.1038\/s41591-024-02856-4","volume":"30","author":"MY Lu","year":"2024","unstructured":"Lu MY, Chen B, Williamson DFK, et al. A visual-language foundation model for computational pathology. Nat Med. 2024;30:863\u201374.","journal-title":"Nat Med"},{"key":"988_CR20","doi-asserted-by":"publisher","first-page":"769","DOI":"10.1038\/s41586-024-08378-w","volume":"638","author":"J Xiang","year":"2025","unstructured":"Xiang J, Wang X, Zhang X, et al. A vision-language foundation model for precision oncology. Nature. 2025;638:769\u201378.","journal-title":"Nature"},{"key":"988_CR21","doi-asserted-by":"publisher","first-page":"2924","DOI":"10.1038\/s41591-024-03141-0","volume":"30","author":"E Vorontsov","year":"2024","unstructured":"Vorontsov E, Bozkurt A, Casson A, et al. A foundation model for clinical-grade computational pathology and rare cancers detection. Nat Med. 2024;30:2924\u201335.","journal-title":"Nat Med"},{"key":"988_CR22","unstructured":"Lee JH, Xinran Li C, Jahanandish H, et al. Prostate-specific foundation models for enhanced detection of clinically significant cancer. arXiv e-prints. 2025. arXiv:2502.00366."}],"container-title":["Cancer Imaging"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/article\/10.1186\/s40644-026-00988-z","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s40644-026-00988-z.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s40644-026-00988-z.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,16]],"date-time":"2026-06-16T17:49:25Z","timestamp":1781632165000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1186\/s40644-026-00988-z"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,1,19]]},"references-count":22,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,12]]}},"alternative-id":["988"],"URL":"https:\/\/doi.org\/10.1186\/s40644-026-00988-z","relation":{},"ISSN":["1470-7330"],"issn-type":[{"value":"1470-7330","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,1,19]]},"assertion":[{"value":"21 October 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"1 January 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"19 January 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"This study was conducted in accordance with the ethical standards of the institutional and\/or national research committee and with the 1964 Helsinki Declaration and its later amendments. U-Found was trained in part with University of Miami mpMRI exams from the Departments of Radiation Oncology and Urology. The mpMRI was reviewed under an IRB-approved protocol for retrospective review of prostate imaging (Protocol #20090554). The MAST study protocol was approved by the Institutional Review Board of University of Miami (IRB# 20140372, ClinicalTrials.gov: NCT02242773). Written informed consent was obtained from all individual participants included in the study.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethical approval and consent to participate"}},{"value":"All participants provided consent for publication of anonymized data. No identifying information is included in this manuscript.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent for publication"}},{"value":"The authors declare no competing interests.","order":4,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}],"article-number":"26"}}