{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.007466185,0.0006489742,0.0003771438,0.001140186,0.0002433312,0.0008279628,0.001068939,0.001182136,0.003302146],"category_scores_gemma":[0.0300208,0.0001879375,0.0003309308,0.0007409938,0.0004064747,0.0008115341,0.0008338077,0.0003890434,0.001360427],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006153132,"about_ca_system_score_gemma":0.0004548259,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001817243,"about_ca_topic_score_gemma":0.001115134,"domain_scores_codex":[0.9966887,0.001517946,0.0003233462,0.0006117079,0.0007322078,0.0001261784],"domain_scores_gemma":[0.9761428,0.01426532,0.0006795694,0.001785573,0.00668498,0.0004417543],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.01507317,0.005735848,0.2474249,0.0006569879,0.0008577715,0.0008986313,0.000840082,0.04464296,0.05262785,0.002484274,0.01679858,0.611959],"study_design_scores_gemma":[0.001231216,0.006240345,0.1191471,0.00007711402,0.0002882265,0.001377978,0.000309648,0.7909045,0.06862933,0.001666152,0.0100467,0.00008177031],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9466217,0.0003882825,0.03983292,0.0004986221,0.0002304817,0.0006931469,0.0009147768,0.003835207,0.006984927],"genre_scores_gemma":[0.9666562,0.00006873221,0.02889273,0.0002472941,0.0000714614,0.00026802,0.00143234,0.0001769012,0.002186291],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.007466185,"threshold_uncertainty_score":0.03948539,"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."}}