{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001260339,0.00008645125,0.0003584069,0.0002350671,0.0000701609,0.00004326618,0.0001240688,0.0002624424,0.000158911],"category_scores_gemma":[0.003686906,0.0000623563,0.000070545,0.000218114,0.00003398547,0.0002119158,0.00002968069,0.000450165,0.000008741131],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002339701,"about_ca_system_score_gemma":0.001169733,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001377548,"about_ca_topic_score_gemma":0.00004922712,"domain_scores_codex":[0.9974934,0.0001266743,0.001351868,0.00009146144,0.0007834992,0.0001530302],"domain_scores_gemma":[0.9982026,0.0001627989,0.0004936971,0.0002159929,0.0006032005,0.0003217577],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.001032013,0.0004056007,0.1462762,0.0009395735,0.0005822617,0.003190579,0.01996803,0.001941345,0.01409778,0.001736194,0.2378416,0.5719888],"study_design_scores_gemma":[0.002879021,0.004384634,0.1147338,0.02644487,0.001629256,0.03982971,0.06948441,0.07554522,0.07120753,0.04279837,0.5495756,0.001487625],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8624753,0.0009228163,0.1088823,0.01927042,0.00791134,0.0003039046,0.00000445688,0.00001036739,0.0002191359],"genre_scores_gemma":[0.9930934,0.00004021591,0.001730797,0.003477746,0.001600137,0.000004165964,0.000008476844,0.000004596782,0.00004046133],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5705012,"threshold_uncertainty_score":0.4413837,"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."}}