{"id":"W4408364279","doi":"10.1148/ryai.240287","title":"External Testing of a Commercial AI Algorithm for Breast Cancer Detection at Screening Mammography","year":2025,"lang":"en","type":"article","venue":"Radiology Artificial Intelligence","topic":"AI in cancer detection","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"BC Cancer Agency; University of Calgary; Kelowna General Hospital","funders":"Mitacs; Canadian Cancer Society","keywords":"Receiver operating characteristic; Medicine; Breast cancer; Mammography; Algorithm; Breast cancer screening; Area under the curve; Area under curve; Retrospective cohort study; Machine learning; Cancer; Artificial intelligence; Internal medicine; Oncology; Mathematics; Computer science","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0004901309,0.0001790913,0.0002854586,0.000334462,0.0003903393,0.00004122629,0.000626276,0.0001705025,0.00002447155],"category_scores_gemma":[0.00009055476,0.00019116,0.0001408081,0.001057593,0.0002630834,0.0002479029,0.0001948743,0.000233114,0.00000503224],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001675352,"about_ca_system_score_gemma":0.00009677906,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0004042326,"about_ca_topic_score_gemma":0.0004157655,"domain_scores_codex":[0.9983028,0.0001208324,0.0005319083,0.0005204945,0.0001439365,0.0003800397],"domain_scores_gemma":[0.9985247,0.0005078032,0.000234093,0.0003466965,0.0003302888,0.00005647238],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0000835507,0.0000264477,0.00198211,0.00001836187,0.00003754006,0.000001637127,0.00009848984,0.001781075,0.01657481,0.002806983,0.00005655411,0.9765325],"study_design_scores_gemma":[0.00007607161,0.000191461,0.01152018,0.0001102132,0.00003186115,0.00009837576,0.00002760626,0.6804549,0.2877973,0.01918969,0.0002884442,0.0002139219],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0136762,0.0003146245,0.9829167,0.0006373576,0.001891723,0.0003313226,0.00003517885,0.0001323606,0.00006455992],"genre_scores_gemma":[0.8409838,0.00002688288,0.1579766,0.0004135828,0.0003711295,0.000171772,9.054535e-7,0.00001206827,0.00004326411],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9763185,"threshold_uncertainty_score":0.7795278,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04239276374823154,"score_gpt":0.322733987064202,"score_spread":0.2803412233159705,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}