{"id":"W3156931326","doi":"10.2196/22461","title":"Neural Machine Translation–Based Automated Current Procedural Terminology Classification System Using Procedure Text: Development and Validation Study","year":2021,"lang":"en","type":"article","venue":"JMIR Formative Research","topic":"Medical Coding and Health Information","field":"Health Professions","cited_by":16,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"National Institute of General Medical Sciences; Medical School, University of Michigan; National Institutes of Health","keywords":"Computer science; Artificial intelligence; Machine learning; Medical diagnosis; Support vector machine; Preprocessor; Machine translation; Data mining; Data pre-processing; Terminology; Medical classification; Medicine","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["sts"],"consensus_categories":[],"category_scores_codex":[0.00267438,0.0001722074,0.0002850937,0.0004205051,0.002154535,0.00004923368,0.0001300254,0.0001875571,0.00004945299],"category_scores_gemma":[0.000282218,0.0001415862,0.00002427046,0.0007431608,0.00008346845,0.0004857773,0.0000841841,0.001211772,0.00008528153],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006706283,"about_ca_system_score_gemma":0.002029504,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002178508,"about_ca_topic_score_gemma":0.00004249671,"domain_scores_codex":[0.995358,0.001720123,0.001071733,0.000295074,0.0008898852,0.0006652153],"domain_scores_gemma":[0.9977369,0.0004693396,0.0003138241,0.0002066423,0.00103776,0.0002354857],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0009659468,0.002326248,0.1726831,0.05028209,0.0001342542,0.00002596562,0.379559,0.0002930375,0.001219328,0.002736675,0.004862795,0.3849115],"study_design_scores_gemma":[0.001825969,0.000178585,0.1348503,0.000802081,0.00001061991,0.00001108567,0.02664684,0.8334208,0.0002061257,0.000009442943,0.001881712,0.0001563954],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9923413,0.000246855,0.001592439,0.001317134,0.0003265981,0.002995572,0.00001782261,0.0003747595,0.000787511],"genre_scores_gemma":[0.997992,0.000006410432,0.0005407093,0.00008985614,0.00008249497,0.0009398837,0.0002953557,0.0000169543,0.00003632193],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8331278,"threshold_uncertainty_score":0.9991446,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4995261141940527,"score_gpt":0.5734256507261446,"score_spread":0.07389953653209186,"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."}}