{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00139815,0.001170977,0.0005815019,0.000771265,0.0002381029,0.0006136464,0.001523353,0.0009512192,0.002247514],"category_scores_gemma":[0.005421485,0.0004607451,0.001267004,0.0004906499,0.0006020944,0.0009724327,0.001322953,0.001773652,0.00209685],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004846864,"about_ca_system_score_gemma":0.0009526378,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003454108,"about_ca_topic_score_gemma":0.005137037,"domain_scores_codex":[0.9991375,0.0003828029,0.00005240997,0.0002342161,0.0001285292,0.00006454933],"domain_scores_gemma":[0.9964082,0.002636437,0.0001741953,0.0004373863,0.0002589931,0.00008475184],"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":[0.001439514,0.0006475955,0.009811871,0.0006509676,0.0001993923,0.001047887,0.0007750111,0.2889082,0.06192926,0.003167314,0.01875565,0.6126674],"study_design_scores_gemma":[0.00003946895,0.0001663614,0.001573935,0.00003109239,0.00003876052,0.0002401556,0.00008857642,0.9656436,0.02544609,0.00253504,0.004161112,0.00003581563],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2368405,0.001578418,0.7238599,0.001312722,0.0003820418,0.0002878324,0.003570792,0.02933328,0.002834597],"genre_scores_gemma":[0.7926858,0.000379232,0.190567,0.000555755,0.0001263605,0.0003951521,0.01111238,0.0005710598,0.003607299],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003454108,"threshold_uncertainty_score":0.007518649,"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."}}