{"id":"W4408405940","doi":"10.1007/s10916-025-02167-2","title":"Using Generative AI to Extract Structured Information from Free Text Pathology Reports","year":2025,"lang":"en","type":"article","venue":"Journal of Medical Systems","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":18,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"Taipei Medical University","keywords":"Computer science; Generalizability theory; Health informatics; Unstructured data; Artificial intelligence; Data science; Pathology; Medicine; Data mining; Big data; Public health; Psychology","routes":{"ca_aff":true,"ca_fund":false,"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":[0.003049989,0.001070668,0.0005001883,0.003333778,0.0006168231,0.00253102,0.001290188,0.0006798328,0.002910207],"category_scores_gemma":[0.01683914,0.000407053,0.001231096,0.001774238,0.0008350461,0.002044353,0.001816414,0.001007704,0.002223285],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005715787,"about_ca_system_score_gemma":0.001559379,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002232934,"about_ca_topic_score_gemma":0.003176511,"domain_scores_codex":[0.9976134,0.0009323179,0.0002497967,0.0004838451,0.0006335301,0.00008716121],"domain_scores_gemma":[0.9790396,0.01608677,0.0009777681,0.001734835,0.001935649,0.0002253592],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0004560059,0.0002395211,0.008100899,0.001141955,0.0001994915,0.001550192,0.003497749,0.02995708,0.06162779,0.01132915,0.01107238,0.8708279],"study_design_scores_gemma":[0.00009088831,0.0003053468,0.007036072,0.0003885239,0.0002553517,0.002276468,0.001680647,0.7789891,0.1018329,0.04598746,0.06092206,0.0002352489],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.02380922,0.0002536386,0.9559946,0.0004112157,0.0000904246,0.0003463779,0.0006384252,0.01549012,0.002965963],"genre_scores_gemma":[0.1571766,0.0003363772,0.8353974,0.0003842054,0.00009217386,0.0002658322,0.002638125,0.0008504,0.002858972],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.003333778,"threshold_uncertainty_score":0.01613003,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1086383209129768,"score_gpt":0.4531075349627653,"score_spread":0.3444692140497885,"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."}}