{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003987579,0.0005311336,0.0003902313,0.001144553,0.0002662592,0.0006986035,0.0007206042,0.0005977654,0.0009488563],"category_scores_gemma":[0.01100479,0.0002053553,0.0008754628,0.0005190606,0.0004556253,0.0005358633,0.0004983214,0.0007398989,0.0003870126],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006702764,"about_ca_system_score_gemma":0.0007055152,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00607582,"about_ca_topic_score_gemma":0.005295674,"domain_scores_codex":[0.999012,0.0004413361,0.00008366678,0.000244532,0.0001408916,0.00007756204],"domain_scores_gemma":[0.9878698,0.008029348,0.001031154,0.001293912,0.001466347,0.0003094717],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.001185503,0.002152378,0.883391,0.0002045263,0.0005057263,0.0003782217,0.0004906072,0.04745151,0.003217647,0.000583041,0.001883903,0.058556],"study_design_scores_gemma":[0.00007494217,0.001519498,0.3194562,0.00008544906,0.0001691732,0.0004231544,0.0005894076,0.6716537,0.003926057,0.0009584266,0.001093956,0.00005011003],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9899539,0.0001086039,0.008515293,0.00007079382,0.00001874228,0.00009109823,0.0009189748,0.0001087205,0.0002139716],"genre_scores_gemma":[0.9926588,0.00006052627,0.00470076,0.00002596357,0.00001241238,0.00006973523,0.002306639,0.000005576709,0.0001596825],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.00607582,"threshold_uncertainty_score":0.0210886,"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."}}