{"id":"W7117364293","doi":"10.2196/77409","title":"A Sentence Classification–Based Medical Status Extraction Pipeline for Electronic Health Records: Institutional Case Study","year":2025,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Information extraction; Sentence; Medical record; Medical information; Electronic health record; Health records; Electronic medical record; Health informatics","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00272712,0.0001997999,0.0003143343,0.0003054779,0.0004951238,0.00009477814,0.0007903969,0.000220829,0.00007447434],"category_scores_gemma":[0.002133853,0.0001757865,0.00007797244,0.000822962,0.0001242429,0.0004909909,0.0001719686,0.001126017,0.00002124776],"about_ca_system_candidate":true,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000614994,"about_ca_system_score_gemma":0.01175924,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005321044,"about_ca_topic_score_gemma":0.001125738,"domain_scores_codex":[0.9958567,0.0002549239,0.001340051,0.00028225,0.00154458,0.0007215041],"domain_scores_gemma":[0.9972261,0.0008131062,0.0003914397,0.0006051582,0.0003249187,0.0006393052],"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":[0.00005632486,0.001138405,0.005752624,0.001026747,0.00003908643,0.0001784713,0.004707556,0.0003286035,3.273509e-7,0.05361931,0.01176883,0.9213837],"study_design_scores_gemma":[0.001436986,0.0003524291,0.001431378,0.000189037,0.000005670063,0.000437843,0.002318191,0.9607539,0.0000010226,0.0003738535,0.03255809,0.0001415853],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02306438,0.0001127563,0.9506584,0.02320595,0.0007310589,0.001551608,0.000008681443,0.0003164654,0.0003507051],"genre_scores_gemma":[0.9542568,0.00008068534,0.03140517,0.01309522,0.0001832157,0.0007356399,0.00008475208,0.00001092837,0.0001475943],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9604253,"threshold_uncertainty_score":0.9938432,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03488286215087998,"score_gpt":0.4077768561638938,"score_spread":0.3728939940130138,"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."}}