{"id":"W4310368227","doi":"10.1101/2022.11.28.22282767","title":"Natural Language Processing for Clinical Laboratory Data Repository Systems: Implementation and Evaluation for Respiratory Viruses","year":2022,"lang":"en","type":"preprint","venue":"medRxiv","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Sunnybrook Hospital; Sinai Health System; Institute for Clinical Evaluative Sciences; Vector Institute; Public Health Ontario; University Health Network; University of Toronto","funders":"Vector Institute; Canadian Institutes of Health Research; Hospital for Sick Children","keywords":"Computer science; Artificial intelligence; Natural language processing; Generalizability theory; Parsing; Machine learning; Classifier (UML); F1 score; Information extraction","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"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":[0.00609035,0.001147629,0.0006161483,0.001077418,0.0007084818,0.001367642,0.003709599,0.001550638,0.003242744],"category_scores_gemma":[0.01510696,0.0005679806,0.0007618477,0.001002067,0.0007005222,0.002682708,0.001711377,0.002072568,0.001927554],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002167004,"about_ca_system_score_gemma":0.002959854,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02507743,"about_ca_topic_score_gemma":0.01941498,"domain_scores_codex":[0.9969766,0.001030423,0.0003622846,0.0008654546,0.0005838281,0.0001814633],"domain_scores_gemma":[0.99204,0.004794008,0.0003258147,0.0007095373,0.00172318,0.0004074623],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.003545332,0.00573429,0.03067631,0.003171277,0.0009161049,0.002096057,0.002449985,0.1775878,0.04060695,0.002144326,0.05226688,0.6788048],"study_design_scores_gemma":[0.000384839,0.0006763323,0.005732359,0.0000967606,0.0001242697,0.0002563085,0.0004685679,0.9571471,0.02707722,0.001020841,0.006943511,0.00007179715],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.7548903,0.001396138,0.143699,0.001870368,0.0004137084,0.003332363,0.006193922,0.08307398,0.005130196],"genre_scores_gemma":[0.7083008,0.0006805421,0.2702346,0.0005531689,0.00005704022,0.001250574,0.01451858,0.0009340377,0.003470655],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.02507743,"threshold_uncertainty_score":0.04986292,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2407022410614401,"score_gpt":0.4900222128435923,"score_spread":0.2493199717821522,"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."}}