{"id":"W4414980730","doi":"10.14740/gr2062","title":"Enabling Drug-Induced Liver Injury Surveillance Through Automated Medication Extraction From Clinical Notes: A Medical Information Mart for Intensive Care IV Real-World Large Language Models Validation Study","year":2025,"lang":"en","type":"article","venue":"Gastroenterology Research","topic":"Pharmacovigilance and Adverse Drug Reactions","field":"Pharmacology, Toxicology and Pharmaceutics","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Intensive care; Data extraction; Information extraction; Identification (biology); Recall; Liver injury; Electronic medical record; Extraction (chemistry)","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"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":[0.006340703,0.001006058,0.0004638704,0.001054761,0.0003683081,0.001104259,0.001077234,0.0007430119,0.0008982076],"category_scores_gemma":[0.01784988,0.0002540728,0.000999505,0.0005634401,0.0003924458,0.0007546847,0.001522112,0.0007010901,0.0007835455],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001118271,"about_ca_system_score_gemma":0.001654703,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006968205,"about_ca_topic_score_gemma":0.006843904,"domain_scores_codex":[0.9971855,0.001424713,0.0003680904,0.0005068498,0.00040521,0.0001096243],"domain_scores_gemma":[0.9903744,0.005793157,0.0006456961,0.001515891,0.001405905,0.0002649911],"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":[0.0044621,0.003347145,0.2454092,0.002253838,0.001549697,0.002586059,0.00248989,0.1567759,0.05756955,0.002005226,0.02673297,0.4948185],"study_design_scores_gemma":[0.0006711612,0.002742131,0.1118228,0.0003061999,0.0004189842,0.001525023,0.001302408,0.8033412,0.059242,0.002164466,0.01627145,0.0001922738],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9274908,0.0006127934,0.05415555,0.0006842827,0.00009805675,0.0009323794,0.009574847,0.005169192,0.001282138],"genre_scores_gemma":[0.8611224,0.0002252291,0.1116557,0.0003630698,0.00004903935,0.0006364416,0.02485906,0.0001773823,0.0009116387],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.006968205,"threshold_uncertainty_score":0.03353322,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2097857129953876,"score_gpt":0.57128462922677,"score_spread":0.3614989162313824,"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."}}