{"id":"W4412163719","doi":"10.1158/1557-3265.aimachine-b015","title":"Abstract B015: Large language models to predict cancer risk from free-text clinical notes","year":2025,"lang":"en","type":"article","venue":"Clinical Cancer Research","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Sunnybrook Health Science Centre; University Health Network; University of Toronto; North Pacific Marine Science Organization; Princess Margaret Cancer Centre","funders":"","keywords":"Cancer; Medicine; Linguistics; Computer science; Internal medicine; Philosophy","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"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":[0.003124868,0.001510074,0.0006210192,0.001503436,0.000625818,0.001339462,0.002051718,0.001101745,0.005284008],"category_scores_gemma":[0.01452854,0.0005244246,0.001458035,0.0008790034,0.0003476513,0.00110704,0.001242872,0.001663474,0.002457957],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002464543,"about_ca_system_score_gemma":0.002643896,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.106866,"about_ca_topic_score_gemma":0.1309832,"domain_scores_codex":[0.9990264,0.0004126426,0.00007351633,0.0003060272,0.0001026562,0.00007875028],"domain_scores_gemma":[0.9932541,0.005200262,0.0002580898,0.0004236418,0.0006372391,0.0002265909],"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":[0.004030907,0.001386823,0.1962729,0.001105749,0.001226543,0.001012453,0.000730479,0.2848882,0.003001973,0.004105739,0.1466449,0.3555933],"study_design_scores_gemma":[0.0002678407,0.0001846249,0.0119413,0.00009918863,0.0001378893,0.00009652994,0.0001263498,0.9761326,0.0007827233,0.004250275,0.00592544,0.00005526505],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6067453,0.004608253,0.178247,0.01073442,0.0009183891,0.001398695,0.1655991,0.02373621,0.008012711],"genre_scores_gemma":[0.7743015,0.0007858332,0.09220918,0.001167137,0.0004114502,0.000802958,0.1235748,0.00032258,0.006424467],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.106866,"threshold_uncertainty_score":0.212488,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1761727981245693,"score_gpt":0.5355322737544392,"score_spread":0.3593594756298699,"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."}}