{"id":"W4409229222","doi":"10.1093/jncics/pkaf037","title":"Deep learning analysis of hematoxylin and eosin-stained benign breast biopsies to predict future invasive breast cancer","year":2025,"lang":"en","type":"article","venue":"JNCI Cancer Spectrum","topic":"AI in cancer detection","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Calgary","funders":"Division of Cancer Epidemiology and Genetics, National Cancer Institute; National Cancer Institute; National Institutes of Health; Breast Cancer Research Foundation","keywords":"Medicine; H&E stain; Pathological; Receiver operating characteristic; Breast cancer; Logistic regression; Diagnostic accuracy; Disease; Pathology; Artificial intelligence; Cancer; Internal medicine; Computer science; Immunohistochemistry","routes":{"ca_aff":true,"ca_fund":false,"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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0002278684,0.0003050701,0.0006226819,0.0009336534,0.0002437128,0.0001530512,0.0006584818,0.0001394095,0.0001377686],"category_scores_gemma":[0.00002040732,0.0002947577,0.0001717703,0.004509466,0.0000986532,0.0004048197,0.0004031585,0.0003100379,0.000002750305],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006763639,"about_ca_system_score_gemma":0.0004461679,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003886018,"about_ca_topic_score_gemma":0.01287543,"domain_scores_codex":[0.9977335,0.00009299927,0.0004393795,0.0008162054,0.0004344941,0.0004834577],"domain_scores_gemma":[0.9986343,0.0001016669,0.0002704912,0.0006040121,0.00022235,0.0001671547],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0006270514,0.0001767064,0.6152224,0.000644776,0.005338766,0.00004084354,0.008549941,0.1078298,0.006169863,0.01633636,0.00301867,0.2360448],"study_design_scores_gemma":[0.0009672694,0.0001909159,0.8590099,0.0003932164,0.0009705201,0.00005509789,0.001072086,0.1184462,0.01504058,0.001071443,0.002074599,0.000708168],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6473903,0.007607888,0.2911875,0.04687331,0.003561279,0.001106583,0.0003479446,0.0006603329,0.001264875],"genre_scores_gemma":[0.9961465,0.001046685,0.001351897,0.0004542669,0.0004125871,0.0001656309,0.000005044412,0.00002056444,0.0003968135],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3487562,"threshold_uncertainty_score":0.9999505,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.004232689961310168,"score_gpt":0.2419677979020703,"score_spread":0.2377351079407601,"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."}}