{"id":"W2970660638","doi":"10.2196/12575","title":"Extracting Clinical Features From Dictated Ambulatory Consult Notes Using a Commercially Available Natural Language Processing Tool: Pilot, Retrospective, Cross-Sectional Validation Study","year":2019,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":16,"is_retracted":false,"has_abstract":true,"ca_institutions":"St. Michael's Hospital; Hamilton Health Sciences; Public Health Ontario; University of Toronto","funders":"","keywords":"Natural language processing; Artificial intelligence; Computer science; Abstraction; Information retrieval; Medicine; Health records; Machine learning; Health care","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":true,"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":[0.01384019,0.0005070089,0.0004424217,0.001227253,0.0006922606,0.0007636012,0.0006243741,0.0006101654,0.0007380705],"category_scores_gemma":[0.04359588,0.0004454761,0.000493672,0.0008026937,0.001015117,0.0008909167,0.0008915569,0.0006424726,0.0004531246],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000633791,"about_ca_system_score_gemma":0.001503203,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002318302,"about_ca_topic_score_gemma":0.003526856,"domain_scores_codex":[0.9918795,0.003381246,0.001305812,0.001311713,0.001855426,0.0002662413],"domain_scores_gemma":[0.9494976,0.02616675,0.007531772,0.005413356,0.0103658,0.0010247],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0006907614,0.001231423,0.9793729,0.0001726113,0.00009973413,0.0002906219,0.002669282,0.0001824934,0.002523939,0.00004813684,0.0003384768,0.01237971],"study_design_scores_gemma":[0.0001557403,0.005652845,0.9850768,0.0000914665,0.0001190912,0.001472716,0.001577605,0.001734526,0.002901321,0.00005318612,0.001135307,0.00002944526],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9958682,0.0001161133,0.002622292,0.00003277335,0.000009097212,0.0006760635,0.0003958689,0.00002355303,0.0002560923],"genre_scores_gemma":[0.9911336,0.0001175868,0.00650991,0.0001459334,0.00001918789,0.0007825302,0.001158683,0.00001612494,0.0001165271],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01384019,"threshold_uncertainty_score":0.07319474,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04302415869666514,"score_gpt":0.3867284598518602,"score_spread":0.343704301155195,"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."}}