{"id":"W4417343483","doi":"10.64898/2025.12.14.25342223","title":"OMOP CDM for breast cancer research: transforming the Breast Cancer Now Biobank data","year":2025,"lang":"","type":"article","venue":"medRxiv","topic":"AI in cancer detection","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Barts Health NHS Trust; McMaster University","keywords":"Biobank; Interoperability; Breast cancer; Translational research; General partnership; Process (computing); Data quality; Quality (philosophy)","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.04966275,0.001019297,0.0008864183,0.005892711,0.001609027,0.0070622,0.002894562,0.001317273,0.009471248],"category_scores_gemma":[0.1489928,0.001033854,0.002452792,0.00651169,0.001307682,0.004771421,0.01401917,0.00488446,0.004832216],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002959805,"about_ca_system_score_gemma":0.008794663,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008122127,"about_ca_topic_score_gemma":0.005945027,"domain_scores_codex":[0.971283,0.0126678,0.003506342,0.004554428,0.007272509,0.0007158902],"domain_scores_gemma":[0.9103456,0.03653568,0.004996304,0.0340023,0.0120573,0.002062783],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.001330866,0.0004625102,0.05735115,0.002671349,0.0007529964,0.001221448,0.007765002,0.01391911,0.01840706,0.07710965,0.3664847,0.4525241],"study_design_scores_gemma":[0.0003088614,0.000164413,0.02130658,0.001364031,0.0001970797,0.000844652,0.002852414,0.05158738,0.0328893,0.1085376,0.7797684,0.0001793924],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.03255354,0.0009738904,0.7611337,0.01598481,0.00102949,0.003434572,0.1060942,0.06337164,0.01542424],"genre_scores_gemma":[0.1106324,0.0005040385,0.7749105,0.003076856,0.0001874097,0.003734422,0.09470807,0.008526774,0.00371938],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.04966275,"threshold_uncertainty_score":0.2626449,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1260531515446123,"score_gpt":0.3992900830746032,"score_spread":0.273236931529991,"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."}}