{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.02073533,0.0007787584,0.0006970034,0.004962693,0.0007795406,0.004579975,0.001585005,0.00171617,0.001438297],"category_scores_gemma":[0.09681331,0.0004237452,0.001331557,0.001745475,0.0007234621,0.002574887,0.001192028,0.001487384,0.0007435983],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001864631,"about_ca_system_score_gemma":0.005286001,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01019987,"about_ca_topic_score_gemma":0.00759444,"domain_scores_codex":[0.9837908,0.007169995,0.003913469,0.002514308,0.002119543,0.0004918986],"domain_scores_gemma":[0.8723677,0.1009452,0.007134875,0.003323003,0.015535,0.0006942506],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.002530402,0.003012907,0.5273751,0.001464592,0.001193461,0.0009535427,0.002714158,0.04497112,0.03248287,0.00462213,0.008601308,0.3700784],"study_design_scores_gemma":[0.0006332358,0.001529417,0.1326974,0.0006081731,0.0007729587,0.001582468,0.003638195,0.7874885,0.05696384,0.006982644,0.00694741,0.0001557466],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8346665,0.0007269516,0.1497262,0.001293363,0.0002093546,0.002248338,0.005673931,0.003376196,0.002079012],"genre_scores_gemma":[0.667604,0.0002648687,0.3209006,0.0003501058,0.00006912708,0.0008787015,0.009423064,0.0001416891,0.0003677937],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02073533,"threshold_uncertainty_score":0.1096603,"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."}}