{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,16]],"date-time":"2026-06-16T17:53:02Z","timestamp":1781632382955,"version":"3.54.5"},"reference-count":18,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2026,1,24]],"date-time":"2026-01-24T00:00:00Z","timestamp":1769212800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,2,17]],"date-time":"2026-02-17T00:00:00Z","timestamp":1771286400000},"content-version":"vor","delay-in-days":24,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"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>Rationale and objectives<\/jats:title>\n                    <jats:p>To evaluate the impact of deploying AI as the first reader (R1) in a double-reading breast-screening workflow and to characterize second-reader (R2) behavior\u2014including the effect of disclosing whether R1 was AI or human.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Materials and methods<\/jats:title>\n                    <jats:p>\n                      This retrospective study used a cancer-enriched cohort of 220 women (95 cancers), with prevalence-weighted analyses performed to approximate population screening metrics. Five radiologists and one commercially available AI (Breast-SlimView\n                      <jats:sup>\u00ae<\/jats:sup>\n                      , Hera-MI) each served as R1; four radiologists served as R2. For each R2, cases were randomized 1:1 to AI-first versus human-first and, independently, to disclosure versus concealment of R1 identity. R2 could validate, dismiss, or add annotations. The primary endpoint was final decision correctness by breast. We used GEE logistic regression to estimate the overall effect of using AI as the first reader and to isolate second-reader behavior independently of first-reader accuracy.\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>\n                      At the prespecified R1 operating point, AI had sensitivity\/specificity\/accuracy of 85.2%\/79.5%\/80.8% versus 84.3%\/84.5%\/85.0% for human R1s; crude final accuracy was lower for AI-first. At 0.6% prevalence, AI-first yielded higher recalls (20.8% vs. 16.8%) with slightly lower PPV (2.7% vs. 3.0%). Conditioning on R1 correctness, R2s were approximately twice more likely to overturn an incorrect AI-initiated opinion than an incorrect human-initiated one (OR\u2009\u2248\u20092.05,\n                      <jats:italic>p<\/jats:italic>\n                      \u2009&lt;\u20090.001). Disclosure that R1 was AI increased R2 corrections (from 13.6% to 19.1%,\n                      <jats:italic>p<\/jats:italic>\n                      \u2009=\u20090.029). Thirteen AI-true-positive cues were dismissed by R2.\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>At this operating point, AI-first reduced crude accuracy due to lower specificity, yet reader-behavior analyses indicate greater scrutiny of AI-initiated opinions. Protocol, threshold, and user-interface choices may raise specificity while preserving beneficial human\u2013AI dynamics.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1186\/s40644-026-00995-0","type":"journal-article","created":{"date-parts":[[2026,1,24]],"date-time":"2026-01-24T08:36:07Z","timestamp":1769243767000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Human\u2013AI interaction in a cancer-enriched double-reading breast screening cohort: diagnostic accuracy and second-reader behavior"],"prefix":"10.1186","volume":"26","author":[{"given":"Elo\u00efse","family":"Sossavi","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Micka\u00ebl","family":"Tardy","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Florie","family":"Hurstel","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jean","family":"Schwartz","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Antoine","family":"Wackenthaler","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Claire","family":"Harter","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Julien","family":"Uttner","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"M\u00e9lanie","family":"Mollion","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Marie-Fran\u00e7oise","family":"Bretz","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"S\u00e9bastien","family":"Moli\u00e8re","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,1,24]]},"reference":[{"issue":"4","key":"995_CR1","doi-asserted-by":"publisher","first-page":"503","DOI":"10.1038\/s41416-018-0195-6","volume":"119","author":"AMP Coolen","year":"2018","unstructured":"Coolen AMP, Voogd AC, Strobbe LJ, Louwman MWJ, Tjan-Heijnen VCG, Duijm LEM. Impact of the second reader on screening outcome at blinded double reading of digital screening mammograms. Br J Cancer. 2018;119(4):503\u20137.","journal-title":"Br J Cancer"},{"key":"995_CR2","unstructured":"Directorate-General for Health and Consumers (European Commission) Now known as\u2026 Karsa L. von, Holland R, Broeders M, Wolf C de, Perry N, European guidelines for quality assurance in breast cancer screening and diagnosis.4th edition, supplements [Internet]. Publications Office of the European Union; 2013 [cited 2025 Mar 18]. Available from: https:\/\/data.europa.eu\/doi\/10.2772\/13196"},{"key":"995_CR3","doi-asserted-by":"crossref","unstructured":"Chen Y, Taib AG, Darker IT, James JJ. Performance of a breast cancer detection AI algorithm using the personal performance in mammographic screening scheme. Radiology. 2023 Sept;308(3):e223299.","DOI":"10.1148\/radiol.223299"},{"key":"995_CR4","doi-asserted-by":"crossref","unstructured":"Sossavi E, Roy C, Moli\u00e8re S. Artificial intelligence in breast cancer screening: a systematic review and meta-analysis of integration strategies. Eur J Radiol Open. 2026 June 1;16:100727.","DOI":"10.1016\/j.ejro.2026.100727"},{"issue":"4","key":"995_CR5","doi-asserted-by":"publisher","first-page":"e222176","DOI":"10.1148\/radiol.222176","volume":"307","author":"T Dratsch","year":"2023","unstructured":"Dratsch T, Chen X, Rezazade Mehrizi M, Kloeckner R, M\u00e4hringer-Kunz A, P\u00fcsken M, et al. Automation bias in mammography: the impact of artificial intelligence BI-RADS suggestions on reader performance. Radiology. 2023;307(4):e222176.","journal-title":"Radiology"},{"issue":"6","key":"995_CR6","doi-asserted-by":"publisher","first-page":"e230529","DOI":"10.1148\/ryai.230529","volume":"6","author":"MT Elhakim","year":"2024","unstructured":"Elhakim MT, Stougaard SW, Graumann O, Nielsen M, Gerke O, Larsen LB, et al. AI-integrated screening to replace double reading of mammograms: A Population-wide accuracy and feasibility study. Radiol Artif Intell. 2024;6(6):e230529.","journal-title":"Radiol Artif Intell"},{"key":"995_CR7","first-page":"706","volume-title":"Explanation in causal inference: methods for mediation and interaction","author":"TJ VanderWeele","year":"2015","unstructured":"VanderWeele TJ. Explanation in causal inference: methods for mediation and interaction. Oxfordf New York, NY: Oxford University Press; 2015. p. 706."},{"issue":"10","key":"995_CR8","doi-asserted-by":"publisher","first-page":"e703","DOI":"10.1016\/S2589-7500(23)00153-X","volume":"5","author":"K Dembrower","year":"2023","unstructured":"Dembrower K, Crippa A, Col\u00f3n E, Eklund M, Strand F. Artificial intelligence for breast cancer detection in screening mammography in sweden: a prospective, population-based, paired-reader, non-inferiority study. Lancet Digit Health. 2023;5(10):e703\u201311.","journal-title":"Lancet Digit Health"},{"key":"995_CR9","doi-asserted-by":"publisher","first-page":"104498","DOI":"10.1016\/j.ebiom.2023.104498","volume":"90","author":"ML Marinovich","year":"2023","unstructured":"Marinovich ML, Wylie E, Lotter W, Lund H, Waddell A, Madeley C, et al. Artificial intelligence (AI) for breast cancer screening: breastscreen population-based cohort study of cancer detection. EBioMedicine. 2023;90:104498.","journal-title":"EBioMedicine"},{"issue":"10","key":"995_CR10","doi-asserted-by":"publisher","first-page":"2697","DOI":"10.1177\/02841851231187382","volume":"64","author":"SH Heywang-K\u00f6brunner","year":"2023","unstructured":"Heywang-K\u00f6brunner SH, Hacker A, J\u00e4nsch A, Hertlein M, Mieskes C, Elsner S, et al. Use of novel artificial intelligence computer-assisted detection (AI-CAD) for screening mammography: an analysis of 17,884 consecutive two-view full-field digital mammography screening exams. Acta Radiol. 2023;64(10):2697\u2013703.","journal-title":"Acta Radiol"},{"issue":"3","key":"995_CR11","doi-asserted-by":"publisher","first-page":"e175","DOI":"10.1016\/S2589-7500(24)00267-X","volume":"7","author":"V Hernstr\u00f6m","year":"2025","unstructured":"Hernstr\u00f6m V, Josefsson V, Sartor H, Schmidt D, Larsson AM, Hofvind S, et al. Screening performance and characteristics of breast cancer detected in the mammography screening with artificial intelligence trial (MASAI): a randomised, controlled, parallel-group, non-inferiority, single-blinded, screening accuracy study. Lancet Digit Health. 2025;7(3):e175\u201383.","journal-title":"Lancet Digit Health"},{"issue":"8","key":"995_CR12","doi-asserted-by":"publisher","first-page":"936","DOI":"10.1016\/S1470-2045(23)00298-X","volume":"24","author":"K L\u00e5ng","year":"2023","unstructured":"L\u00e5ng K, Josefsson V, Larsson AM, Larsson S, H\u00f6gberg C, Sartor H, et al. Artificial intelligence-supported screen reading versus standard double reading in the mammography screening with artificial intelligence trial (MASAI): a clinical safety analysis of a randomised, controlled, non-inferiority, single-blinded, screening accuracy study. Lancet Oncol. 2023;24(8):936\u201344.","journal-title":"Lancet Oncol"},{"issue":"5","key":"995_CR13","doi-asserted-by":"publisher","first-page":"1097","DOI":"10.2214\/AJR.17.18907","volume":"210","author":"JN Itri","year":"2018","unstructured":"Itri JN, Patel SH. Heuristics and cognitive error in medical imaging. AJR Am J Roentgenol. 2018;210(5):1097\u2013105.","journal-title":"AJR Am J Roentgenol"},{"key":"995_CR14","unstructured":"Yilmaz E, Seker ME, Guldogan N, Turk EB, Erdemli S, Koyluoglu YO, et al. Clinical Application of AI in Mammography: Insights from a Prospective Study. Acad Radiol [Internet]. 2025 June 2 [cited 2025 July 2]; Available from: https:\/\/www.sciencedirect.com\/science\/article\/pii\/S1076633225004532."},{"issue":"3","key":"995_CR15","doi-asserted-by":"publisher","first-page":"e242566","DOI":"10.1148\/radiol.242566","volume":"314","author":"KE Dembrower","year":"2025","unstructured":"Dembrower KE, Crippa A, Eklund M, Strand F. Human-AI interaction in the screentrustcad trial: recall proportion and positive predictive value related to screening mammograms flagged by AI CAD versus a human reader. Radiology. 2025;314(3):e242566.","journal-title":"Radiology"},{"issue":"none","key":"995_CR16","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1214\/21-SS133","volume":"16","author":"C Rudin","year":"2022","unstructured":"Rudin C, Chen C, Chen Z, Huang H, Semenova L, Zhong C. Interpretable machine learning: fundamental principles and 10 grand challenges. Stat Surv. 2022;16(none):1\u201385.","journal-title":"Stat Surv"},{"key":"995_CR17","doi-asserted-by":"crossref","unstructured":"van Winkel SL, Sechopoulos I, Rodr\u00edguez-Ruiz A, Veldkamp WJH, Gennaro G, Chevalier M, et al. An Exploration of Discrepant Recalls Between AI and Human Readers of Malignant Lesions in Digital Mammography Screening. Diagnostics. 2025 June 19;15(12):1566.","DOI":"10.3390\/diagnostics15121566"},{"key":"995_CR18","doi-asserted-by":"crossref","unstructured":"Woo OH, Song SE, Choe SJ, Kim M, Cho KR, Seo BK. Invasive breast cancers missed by AI screening of mammograms. Radiology. 2025 June;315(3):e242408.","DOI":"10.1148\/radiol.242408"}],"container-title":["Cancer Imaging"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/article\/10.1186\/s40644-026-00995-0","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s40644-026-00995-0.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s40644-026-00995-0.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,16]],"date-time":"2026-06-16T17:04:55Z","timestamp":1781629495000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1186\/s40644-026-00995-0"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,1,24]]},"references-count":18,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,12]]}},"alternative-id":["995"],"URL":"https:\/\/doi.org\/10.1186\/s40644-026-00995-0","relation":{},"ISSN":["1470-7330"],"issn-type":[{"value":"1470-7330","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,1,24]]},"assertion":[{"value":"29 September 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"19 January 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"24 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 retrospective, non-interventional study was approved by the regional institutional review board (Comit\u00e9 d\u2019\u00c9thique du CHU de Strasbourg; reference CE-2020-133, approval 15 December 2020). The board waived the requirement for written informed consent because all images and clinical data were fully anonymised before analysis and no additional patient contact or intervention occurred. The study complies with the Declaration of Helsinki, the EU General Data Protection Regulation, and the STARD-AI (Standards for Reporting of Diagnostic Accuracy Studies involving Artificial Intelligence) reporting guideline.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethical approval and consent"}},{"value":"All authors have read and approved the final manuscript and consent to its publication.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent for publication"}},{"value":"Micka\u00ebl Tardy is Scientific Director at Hera-MI, the vendor of the AI software evaluated. Hera-MI provided the software and technical documentation only. Micka\u00ebl Tardy\u2019s involvement was limited to configuring the prespecified operating threshold and supplying documentation. Neither Hera-MI nor Micka\u00ebl Tardy had any role in study design, case selection, data collection, data curation, statistical analysis, interpretation of results, manuscript preparation, or the decision to submit. All analyses were conducted independently by the academic authors. All other authors declare no competing interests.","order":4,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}],"article-number":"29"}}