{"id":"W3111416811","doi":"10.2196/24490","title":"Automated Categorization of Systemic Disease and Duration From Electronic Medical Record System Data Using Finite-State Machine Modeling: Prospective Validation Study","year":2020,"lang":"en","type":"article","venue":"JMIR Formative Research","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Computer science; Medical record; Unstructured data; Categorization; Medical history; Big data; Electronic medical record; Artificial intelligence; Analytics; Data mining; Machine learning; Medicine","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":[0.002265656,0.0001622009,0.0002916997,0.0002371574,0.0003132126,0.0001805958,0.001022555,0.00006904211,0.000005144428],"category_scores_gemma":[0.001044225,0.0001438343,0.00002014591,0.001111522,0.00004880591,0.00137137,0.001000484,0.0006757861,0.00001032405],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000371116,"about_ca_system_score_gemma":0.0007798732,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00146188,"about_ca_topic_score_gemma":0.00004464798,"domain_scores_codex":[0.994895,0.001857824,0.0007150137,0.0006277546,0.001512681,0.0003916959],"domain_scores_gemma":[0.9976609,0.0004000263,0.0002878443,0.0006803465,0.0006632548,0.0003076141],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.002392933,0.001687756,0.3543819,0.0108961,0.0008282432,0.0003156259,0.3067465,0.27005,0.001773297,0.0199792,0.0002031769,0.03074523],"study_design_scores_gemma":[0.0006208015,0.000423872,0.002781833,0.0002462277,0.000009813575,0.00000525375,0.001529535,0.9939385,0.00003945623,0.0002838635,0.000001359202,0.0001195005],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5969605,0.0001189695,0.4008335,0.0004263466,0.00005234449,0.001282679,0.00006116013,0.0002564261,0.000007966846],"genre_scores_gemma":[0.9989172,0.00003074575,0.0005912472,0.00001195511,0.00005628384,0.0000860308,0.0002868082,0.000018008,0.000001768265],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7238885,"threshold_uncertainty_score":0.5865393,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1028373134323186,"score_gpt":0.4174706598475131,"score_spread":0.3146333464151945,"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."}}