{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"medRxiv"}],"indexed":{"date-parts":[[2026,6,26]],"date-time":"2026-06-26T15:52:27Z","timestamp":1782489147131,"version":"3.54.5"},"posted":{"date-parts":[[2026,6,24]]},"group-title":"Oncology","reference-count":45,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,6,24]],"date-time":"2026-06-24T00:00:00Z","timestamp":1782259200000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,6,24]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:sec>\n                  <jats:title>Objective<\/jats:title>\n                  <jats:p>To develop and evaluate a deep learning model for five-year breast cancer risk prediction from screening breast ultrasound (BUS) examinations.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Methods<\/jats:title>\n                  <jats:p>This retrospective study included 295,298 breast ultrasound examinations from 122,072 women imaged between 2012 and 2020. Patients were split into training, validation, and test sets; the test set included screening examinations only. BUS-Risk-Net aggregated image features using attention-based multiple instance learning and combined them with age and ultrasound-estimated breast density to predict 2- to 5-year risk. Performance was compared with the full Tyrer\u2013Cuzick model in a matched case-control cohort and with a reduced Tyrer\u2013Cuzick model in the held-out test set. Risk stratification was evaluated within BI-RADS density categories.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Results<\/jats:title>\n                  <jats:p>\n                    In the matched case-control cohort (n = 240 women), BUS-Risk-Net achieved a 5-year AUC of 0.632 (95% CI, 0.562\u20130.702), versus 0.514 for the full Tyrer\u2013Cuzick model (95% CI, 0.440\u20130.588;\n                    <jats:italic>p<\/jats:italic>\n                    = 0.04). Among 19,548 examinations from 9,015 women eligible for 5-year evaluation in the test set, BUS-Risk-Net achieved an AUC of 0.679 (95% CI, 0.653\u20130.706), versus 0.594 for the reduced Tyrer\u2013Cuzick model (95% CI, 0.564\u20130.623;\n                    <jats:italic>P<\/jats:italic>\n                    &lt; .001). Observed 5-year cancer incidence increased across AI-defined risk tiers within each BI-RADS density category, ranging from 0.0% to 5.8% after AI stratification, compared with 2.1% to 3.6% across density categories alone.\n                  <\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Discussion<\/jats:title>\n                  <jats:p>Deep learning models applied to screening breast ultrasound could enable long-term breast cancer risk prediction and stratify risk beyond breast density alone. External and prospective validation is needed before clinical use.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Summary Sentence<\/jats:title>\n                  <jats:p>BUS-Risk-Net enabled long-term breast cancer risk prediction from screening ultrasound, outperformed the Tyrer\u2013Cuzick\u2013based comparators, and identified distinct 5-year risk groups with differing observed cancer incidence within each breast density category.<\/jats:p>\n                <\/jats:sec>","DOI":"10.64898\/2026.06.21.26356188","type":"posted-content","created":{"date-parts":[[2026,6,24]],"date-time":"2026-06-24T18:45:24Z","timestamp":1782326724000},"source":"Crossref","is-referenced-by-count":0,"title":["Five-Year Breast Cancer Risk Prediction From Screening Breast Ultrasound Using Deep Learning"],"prefix":"10.64898","author":[{"given":"Yuxuan","family":"Chen","sequence":"first","affiliation":[{"name":"Perlmutter Cancer Center, NYU Langone Health, New York, NY, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Haoyan","family":"Yang","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Stony Brook University, Stony Brook, NY, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yanqi","family":"Xu","sequence":"additional","affiliation":[{"name":"Center for Data Science, New York University, New York, NY, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ria","family":"Soni","sequence":"additional","affiliation":[{"name":"Department of Internal Medicine, NYU Langone Health, New York, NY, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Laura","family":"Heacock","sequence":"additional","affiliation":[{"name":"Department of Radiology, NYU Langone Health, New York, NY, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Maciej","family":"Lis","sequence":"additional","affiliation":[{"name":"Clinical Diagnostic Imaging Center, 5th Military Hospital with Polyclinic in Krakow, 30-901 Krakow, Poland"},{"name":"Department of Anatomy, Jagiellonian University Medical College, 31-008 Krakow, Poland"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Aleksandra","family":"Stanek","sequence":"additional","affiliation":[{"name":"Clinical Diagnostic Imaging Center, 5th Military Hospital with Polyclinic in Krakow, 30-901 Krakow, Poland"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tomasz","family":"Puto","sequence":"additional","affiliation":[{"name":"Clinical Diagnostic Imaging Center, 5th Military Hospital with Polyclinic in Krakow, 30-901 Krakow, Poland"},{"name":"Department of Diagnostic Imaging, Faculty of Medicine, Andrzej Frycz Modrzewski Krakow University, 30-705 Krakow, Poland"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alana Amarosa","family":"Lewin","sequence":"additional","affiliation":[{"name":"Department of Radiology, NYU Langone Health, New York, NY, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Linda","family":"Moy","sequence":"additional","affiliation":[{"name":"Department of Radiology, NYU Langone Health, New York, NY, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Freya","family":"Schnabel","sequence":"additional","affiliation":[{"name":"Department of Surgery, NYU Langone Health, New York, NY, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yiqiu","family":"Shen","sequence":"additional","affiliation":[{"name":"Department of Radiology, NYU Langone Health, New York, NY, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2026062608323655000_2026.06.21.26356188v1.1","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-025-03502-3"},{"key":"2026062608323655000_2026.06.21.26356188v1.2","doi-asserted-by":"publisher","DOI":"10.3389\/fonc.2025.1656842"},{"key":"2026062608323655000_2026.06.21.26356188v1.3","doi-asserted-by":"publisher","DOI":"10.1016\/j.jacr.2023.04.002"},{"key":"2026062608323655000_2026.06.21.26356188v1.4","doi-asserted-by":"publisher","DOI":"10.1038\/s41416-021-01550-3"},{"key":"2026062608323655000_2026.06.21.26356188v1.5","doi-asserted-by":"publisher","DOI":"10.1093\/jnci\/81.24.1879"},{"key":"2026062608323655000_2026.06.21.26356188v1.6","doi-asserted-by":"publisher","DOI":"10.7326\/0003-4819-148-5-200803040-00004"},{"key":"2026062608323655000_2026.06.21.26356188v1.7","doi-asserted-by":"publisher","DOI":"10.1200\/JCO.2015.60.8869"},{"key":"2026062608323655000_2026.06.21.26356188v1.8","doi-asserted-by":"publisher","DOI":"10.1002\/sim.1668"},{"key":"2026062608323655000_2026.06.21.26356188v1.9","doi-asserted-by":"publisher","DOI":"10.1148\/radiol.222733"},{"key":"2026062608323655000_2026.06.21.26356188v1.10","doi-asserted-by":"publisher","DOI":"10.1126\/scitranslmed.aba4373"},{"key":"2026062608323655000_2026.06.21.26356188v1.11","doi-asserted-by":"publisher","DOI":"10.1126\/scitranslmed.abn3971"},{"key":"2026062608323655000_2026.06.21.26356188v1.12","doi-asserted-by":"publisher","DOI":"10.1158\/1940-6207.CAPR-24-0427"},{"key":"2026062608323655000_2026.06.21.26356188v1.13","doi-asserted-by":"publisher","DOI":"10.1148\/ryai.240417"},{"key":"2026062608323655000_2026.06.21.26356188v1.14","doi-asserted-by":"publisher","DOI":"10.1148\/radiol.232780"},{"key":"2026062608323655000_2026.06.21.26356188v1.15","doi-asserted-by":"publisher","DOI":"10.3390\/cancers15072112"},{"key":"2026062608323655000_2026.06.21.26356188v1.16","doi-asserted-by":"publisher","DOI":"10.1200\/JGO.19.00041"},{"key":"2026062608323655000_2026.06.21.26356188v1.17","doi-asserted-by":"publisher","DOI":"10.1158\/1055-9965.EPI-23-0049"},{"key":"2026062608323655000_2026.06.21.26356188v1.18","doi-asserted-by":"publisher","DOI":"10.1038\/bjc.2015.33"},{"key":"2026062608323655000_2026.06.21.26356188v1.19","doi-asserted-by":"publisher","DOI":"10.7326\/0003-4819-138-3-200302040-00008"},{"key":"2026062608323655000_2026.06.21.26356188v1.20","doi-asserted-by":"publisher","DOI":"10.3348\/kjr.2025.0338"},{"key":"2026062608323655000_2026.06.21.26356188v1.21","doi-asserted-by":"publisher","DOI":"10.7314\/APJCP.2013.14.4.2277"},{"key":"2026062608323655000_2026.06.21.26356188v1.22","doi-asserted-by":"publisher","DOI":"10.3348\/kjr.2022.0099"},{"key":"2026062608323655000_2026.06.21.26356188v1.23","doi-asserted-by":"publisher","DOI":"10.1177\/0284185117695665"},{"key":"2026062608323655000_2026.06.21.26356188v1.24","doi-asserted-by":"publisher","DOI":"10.1016\/j.lana.2025.101096"},{"key":"2026062608323655000_2026.06.21.26356188v1.25","unstructured":"Checklist for Artificial Intelligence in Medical Imaging (CLAIM). Accessed May 20, 2026. https:\/\/pubs.rsna.org\/page\/ai\/claim"},{"key":"2026062608323655000_2026.06.21.26356188v1.26","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2504.05636"},{"key":"2026062608323655000_2026.06.21.26356188v1.27","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2303.16417"},{"key":"2026062608323655000_2026.06.21.26356188v1.28","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-020-01174-9"},{"key":"2026062608323655000_2026.06.21.26356188v1.29","unstructured":"Shamout FE , Shen Y , Witowski J , et al. The NYU Breast Ultrasound Dataset v1.0. Published online 2021. https:\/\/cs.nyu.edu\/~kgeras\/reports\/ultrasound_datav1.0.pdf"},{"key":"2026062608323655000_2026.06.21.26356188v1.30","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-021-26023-2"},{"key":"2026062608323655000_2026.06.21.26356188v1.31","unstructured":"Newell MS , Destounis SV , Leung JWT , DeMartini WB , Lee CH , Eby PR. ACR BI-RADS\u00ae v2025 Manual. American College of Radiology, Reston , VA. Published online 2025."},{"key":"2026062608323655000_2026.06.21.26356188v1.32","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2024.103202"},{"key":"2026062608323655000_2026.06.21.26356188v1.33","unstructured":"Ilse M, Tomczak J, Welling M. Attention-based Deep Multiple Instance Learning. In: Dy J, Krause A, eds. Proceedings of the 35th International Conference on Machine Learning. Vol 80. Proceedings of Machine Learning Research. PMLR; 10--15 Jul 2018:2127-2136. https:\/\/proceedings.mlr.press\/v80\/ilse18a.html"},{"key":"2026062608323655000_2026.06.21.26356188v1.34","unstructured":"Platt J. Probabilistic outputs for support vector machines and comparisons to regularized likelihood methods. Advances in large margin classifiers. Published online 1999. https:\/\/www.researchgate.net\/profile\/John-Platt-2\/publication\/2594015_Probabilistic_Outputs_for_Support_Vector_Machines_and_Comparisons_to_Regularized_Likelihood_Methods\/links\/004635154cff5262d6000000\/Probabilistic-Outputs-for-Support-Vector-Machines-and-Comparisons-to-Regularized-Likelihood-Methods.pdf"},{"key":"2026062608323655000_2026.06.21.26356188v1.35","unstructured":"Cuzick J. IBIS Breast Cancer Risk Evaluation Tool. Accessed April 6, 2026. https:\/\/ems-trials.org\/riskevaluator\/"},{"key":"2026062608323655000_2026.06.21.26356188v1.36","doi-asserted-by":"publisher","DOI":"10.1001\/jamanetworkopen.2025.12681"},{"key":"2026062608323655000_2026.06.21.26356188v1.37","doi-asserted-by":"publisher","DOI":"10.1148\/radiol.223077"},{"key":"2026062608323655000_2026.06.21.26356188v1.38","doi-asserted-by":"publisher","DOI":"10.6004\/jnccn.2015.0105"},{"key":"2026062608323655000_2026.06.21.26356188v1.39","doi-asserted-by":"publisher","DOI":"10.1200\/JCO.19.01472"},{"key":"2026062608323655000_2026.06.21.26356188v1.40","doi-asserted-by":"publisher","DOI":"10.1016\/j.jacr.2025.02.023"},{"key":"2026062608323655000_2026.06.21.26356188v1.41","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0221535"},{"key":"2026062608323655000_2026.06.21.26356188v1.42","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0303670"},{"key":"2026062608323655000_2026.06.21.26356188v1.43","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-022-18747-y"},{"key":"2026062608323655000_2026.06.21.26356188v1.44","doi-asserted-by":"publisher","DOI":"10.1007\/s00330-022-09263-8"},{"key":"2026062608323655000_2026.06.21.26356188v1.45","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-16437-8_45"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.06.21.26356188","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,26]],"date-time":"2026-06-26T15:32:53Z","timestamp":1782487973000},"score":1,"resource":{"primary":{"URL":"http:\/\/medrxiv.org\/lookup\/doi\/10.64898\/2026.06.21.26356188"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,24]]},"references-count":45,"URL":"https:\/\/doi.org\/10.64898\/2026.06.21.26356188","relation":{},"subject":[],"published":{"date-parts":[[2026,6,24]]},"subtype":"preprint"}}