{"id":"W4403064636","doi":"10.1186/s12911-024-02677-y","title":"Validation of large language models for detecting pathologic complete response in breast cancer using population-based pathology reports","year":2024,"lang":"en","type":"article","venue":"BMC Medical Informatics and Decision Making","topic":"AI in cancer detection","field":"Computer Science","cited_by":18,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Calgary; Alberta Health Services","funders":"Canadian Institutes of Health Research","keywords":"Health informatics; Breast cancer; Medicine; Pathology; Population; Cancer; Computer science; Natural language processing; Internal medicine; Public health","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003172863,0.0001018337,0.0002074421,0.0003442785,0.0000900025,0.00009770878,0.0001498642,0.0001324141,0.00001024322],"category_scores_gemma":[0.0004607981,0.0000860506,0.00005289742,0.0003880203,0.00002241711,0.0004467626,0.0001378193,0.0001387802,3.212608e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000110608,"about_ca_system_score_gemma":0.000239768,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002715785,"about_ca_topic_score_gemma":0.00003379217,"domain_scores_codex":[0.9982677,0.00009400037,0.0008076568,0.0001899448,0.0004459087,0.0001947497],"domain_scores_gemma":[0.9974343,0.001927071,0.0002653692,0.0002427779,0.00007687639,0.00005358205],"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":[0.0005736813,0.00004545416,0.006716677,0.001007346,0.00001135846,0.000246156,0.004299121,0.4260486,0.0008954637,0.001782622,0.00002416618,0.5583494],"study_design_scores_gemma":[0.0002798872,0.00003515176,0.001545444,0.0008553213,0.000005736857,0.0003031544,0.0001669812,0.9895337,0.00009437554,0.007081125,0.000005462349,0.00009362146],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.472458,0.0001104798,0.5270388,0.00001440747,0.0002241291,0.0001079531,0.00001037789,0.00003244607,0.000003443464],"genre_scores_gemma":[0.7017179,0.000005738616,0.2981056,0.0001220105,0.00002428581,0.00001605793,0.000002594328,0.000005556544,2.717389e-7],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.5634851,"threshold_uncertainty_score":0.3509042,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05551034995085824,"score_gpt":0.3693355669972379,"score_spread":0.3138252170463796,"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."}}