{"id":"W4407187061","doi":"10.3389/frai.2025.1513674","title":"Targeted generative data augmentation for automatic metastases detection from free-text radiology reports","year":2025,"lang":"en","type":"article","venue":"Frontiers in Artificial Intelligence","topic":"Topic Modeling","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"Queen's University","funders":"National Cancer Institute; Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Context (archaeology); Task (project management); Natural language processing; Identification (biology); Artificial intelligence; F1 score; Language model; Data extraction; Machine learning; Information extraction; Information retrieval; MEDLINE","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006380577,0.0001560432,0.0002777498,0.0002785365,0.0001460365,0.0001273775,0.001105148,0.0000945991,0.00001697524],"category_scores_gemma":[0.0009219084,0.0001643193,0.0000469778,0.0004873661,0.00006898596,0.000594445,0.000358781,0.000128641,0.000004154335],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001456207,"about_ca_system_score_gemma":0.0001185149,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003695062,"about_ca_topic_score_gemma":0.0003976717,"domain_scores_codex":[0.9978676,0.0001790069,0.0007181444,0.0008198317,0.0001491746,0.0002662732],"domain_scores_gemma":[0.9980644,0.0002511232,0.0001997818,0.001367878,0.00007581141,0.00004097863],"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":[0.00003571069,0.00008231715,0.0004352387,0.000022192,0.00009349151,0.00002037725,0.001120992,0.01518756,0.006079133,0.008163883,0.005608723,0.9631504],"study_design_scores_gemma":[0.00003630104,0.00002391703,0.00006475532,0.00001951108,0.00001691533,0.000002142349,0.0002928061,0.730525,0.08231371,0.1862461,0.0003483662,0.0001104158],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0110023,0.0007967316,0.9818275,0.0004607737,0.005177603,0.0005376403,0.00004202742,0.0001000752,0.00005532927],"genre_scores_gemma":[0.3840344,0.00001911747,0.6154324,0.000190743,0.00009393598,0.00009038243,0.00009122605,0.000006519248,0.00004126888],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.96304,"threshold_uncertainty_score":0.6700748,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0575910859937608,"score_gpt":0.3186015932620943,"score_spread":0.2610105072683335,"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."}}