{"id":"W3217643131","doi":"10.1016/j.thromres.2021.11.020","title":"Developing and validating natural language processing algorithms for radiology reports compared to ICD-10 codes for identifying venous thromboembolism in hospitalized medical patients","year":2021,"lang":"en","type":"article","venue":"Thrombosis Research","topic":"Venous Thromboembolism Diagnosis and Management","field":"Medicine","cited_by":40,"is_retracted":false,"has_abstract":false,"ca_institutions":"Sinai Health System; Toronto Rehabilitation Institute; University of Toronto; St. Michael's Hospital","funders":"","keywords":"Medicine; Pulmonary embolism; Algorithm; Receiver operating characteristic; Venous thromboembolism; Venous thrombosis; Area under the curve; Radiology; Gold standard (test); Diagnosis code; Machine learning; Thrombosis; Internal medicine; Population","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.002800428,0.0002692907,0.001284406,0.0004392022,0.0004741655,0.0001809668,0.0002119886,0.0001988013,0.00008976352],"category_scores_gemma":[0.0038101,0.0002561676,0.0001157066,0.0007193598,0.0001727612,0.000171516,0.0004664181,0.0003938724,0.000005337075],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003825533,"about_ca_system_score_gemma":0.0005886518,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003192512,"about_ca_topic_score_gemma":0.0002065207,"domain_scores_codex":[0.9959274,0.0002357094,0.0007982489,0.0009235008,0.001044466,0.001070695],"domain_scores_gemma":[0.9975442,0.0006319437,0.0001433811,0.0003890545,0.0009791324,0.0003123485],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.001308899,0.0117888,0.1281984,0.0249089,0.002651533,0.002362507,0.09703419,0.00009832205,0.04660075,0.003181546,0.04238981,0.6394764],"study_design_scores_gemma":[0.01037104,0.001485842,0.9389317,0.004893652,0.0003351408,0.0001728583,0.01316485,0.007513534,0.0165944,0.001665698,0.003731807,0.001139483],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9898649,0.001856598,0.002306195,0.002253857,0.0003541734,0.002975622,0.00002016213,0.0000740114,0.0002944916],"genre_scores_gemma":[0.9727944,0.0003099297,0.02459557,0.0005943017,0.0002513651,0.0009100113,0.0003799901,0.00007296356,0.00009154258],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8107333,"threshold_uncertainty_score":0.999989,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1064606714904495,"score_gpt":0.45222072072763,"score_spread":0.3457600492371806,"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."}}