{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001027791,0.0002259292,0.0003479161,0.00005740301,0.0001977908,0.000166906,0.0003194121,0.000459711,0.0002211154],"category_scores_gemma":[0.001832579,0.0001854951,0.00009732655,0.0001428169,0.0003204168,0.00004075097,0.0002276,0.0007788965,0.00005753986],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004888654,"about_ca_system_score_gemma":0.0003777634,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00009840084,"about_ca_topic_score_gemma":0.00002614287,"domain_scores_codex":[0.9975397,0.0001359621,0.0009310861,0.0002995753,0.0007609829,0.0003327188],"domain_scores_gemma":[0.9986729,0.0002275699,0.0004113831,0.0003152263,0.0002145866,0.0001583102],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0003251243,0.0004527091,0.975356,0.00006814831,0.00013914,0.00001330825,0.001278618,0.00001181298,0.006247986,0.000003408276,0.0007956148,0.01530815],"study_design_scores_gemma":[0.003597919,0.0009463029,0.9783555,0.0001413818,0.00004061642,0.00005789391,0.002280506,0.01080078,0.002081668,0.00001372016,0.00121767,0.0004660088],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9974731,0.0004519087,0.0004855011,0.00002821279,0.0006576647,0.0004128294,0.00003003152,0.00009098537,0.0003697659],"genre_scores_gemma":[0.994766,0.00001329931,0.003729769,0.0004640797,0.0004756225,0.00001495973,0.0003065876,0.00002103392,0.0002086484],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01484214,"threshold_uncertainty_score":0.7564272,"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."}}