{"id":"W4409250979","doi":"10.1016/j.compbiomed.2025.110161","title":"Integrating large language models with human expertise for disease detection in electronic health records","year":2025,"lang":"en","type":"article","venue":"Computers in Biology and Medicine","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"Calgary Laboratory Services; Alberta Health Services; Libin Cardiovascular Institute of Alberta; University of Calgary","funders":"Canadian Institutes of Health Research","keywords":"Health records; Computer science; Electronic health record; Data science; Natural language processing; Artificial intelligence; Health care","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006487225,0.0001064637,0.0002348429,0.0002756658,0.000118352,0.000006724074,0.0002034412,0.00005414168,7.368806e-7],"category_scores_gemma":[0.00006151041,0.00007935891,0.00001208866,0.0002828655,0.00005744016,0.00006691822,0.00007676556,0.0002895247,6.946999e-8],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001445556,"about_ca_system_score_gemma":0.0001155155,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0007789393,"about_ca_topic_score_gemma":0.003577561,"domain_scores_codex":[0.9988289,0.0001981663,0.0002370903,0.0003536023,0.00003705198,0.0003451192],"domain_scores_gemma":[0.9994686,0.0001725285,0.0000675705,0.0002000388,0.00002193726,0.0000693475],"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":[0.0001654752,0.00008043379,0.06588762,0.0003407784,0.00001500281,0.00001539141,0.009341541,0.000269738,0.0001804359,0.3346503,0.0001764542,0.5888768],"study_design_scores_gemma":[0.002646286,0.001372192,0.01886511,0.0008943168,0.000003150577,0.000006926518,0.000336076,0.9391549,0.00001856055,0.03604145,0.0005156244,0.0001454457],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1261493,0.003182028,0.8641148,0.005869653,0.0002364104,0.0003355521,8.071879e-7,0.00005473684,0.00005674516],"genre_scores_gemma":[0.9889362,0.00009395958,0.008293724,0.00251677,0.00005558707,0.00006413233,0.0000136973,0.000004217927,0.00002173716],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9388852,"threshold_uncertainty_score":0.3236162,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01367169068403938,"score_gpt":0.3675827736706654,"score_spread":0.353911082986626,"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."}}