{"id":"W4311827969","doi":"10.1371/journal.pdig.0000150","title":"Using Primary Care Clinical Text Data and Natural Language Processing to Identify Indicators of COVID-19 in Toronto, Canada","year":2022,"lang":"en","type":"article","venue":"PLOS Digital Health","topic":"Data-Driven Disease Surveillance","field":"Medicine","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"North York General Hospital; University of Toronto","funders":"","keywords":"Coronavirus disease 2019 (COVID-19); Medicine; Medical record; Retrospective cohort study; Severe acute respiratory syndrome; Primary care; Artificial intelligence; Family medicine; Pediatrics; Natural language processing; Computer science; Internal medicine; Disease","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001078797,0.0003137151,0.0002609031,0.002778555,0.000995303,0.001640187,0.0006857775,0.0003065928,0.001225406],"category_scores_gemma":[0.008579294,0.0002327822,0.0003027722,0.005129179,0.0006014142,0.0004364378,0.0006792494,0.0003194813,0.0002325986],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.02634275,"about_ca_system_score_gemma":0.03496978,"about_ca_topic_candidate":true,"about_ca_topic_consensus":true,"about_ca_topic_score_codex":0.991789,"about_ca_topic_score_gemma":0.9920748,"domain_scores_codex":[0.9985011,0.000185621,0.0001941474,0.0003225534,0.0005454404,0.0002512101],"domain_scores_gemma":[0.9919232,0.001670155,0.001295688,0.000251638,0.00420637,0.0006529582],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0001619004,0.00003381628,0.9755939,0.0002538447,0.00006500912,0.0004055646,0.00162775,0.001258721,0.001131164,0.0002164724,0.002725213,0.01652667],"study_design_scores_gemma":[0.00001261202,0.00004068291,0.9904581,0.00006835819,0.00004631208,0.0001138494,0.002003036,0.003566184,0.0006196555,0.0000621146,0.002988076,0.0000210265],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9611228,0.001129096,0.001347388,0.0008510354,0.00002315936,0.0002616225,0.03126223,0.000117306,0.003885304],"genre_scores_gemma":[0.9860036,0.0006018349,0.002419243,0.0001887683,0.00001210574,0.00005694872,0.00959275,0.0000118747,0.001112881],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02634275,"threshold_uncertainty_score":0.1911309,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05818390352170421,"score_gpt":0.4301188475247922,"score_spread":0.371934944003088,"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."}}