{"id":"W4282842066","doi":"10.1371/journal.pdig.0000067","title":"Comparing different versions of computer-aided detection products when reading chest X-rays for tuberculosis","year":2022,"lang":"en","type":"article","venue":"PLOS Digital Health","topic":"COVID-19 diagnosis using AI","field":"Medicine","cited_by":41,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Global Affairs Canada","keywords":"Triage; CAD; Receiver operating characteristic; Computer science; Product (mathematics); Computer-aided diagnosis; Medical physics; Medicine; Artificial intelligence; Machine learning; Engineering drawing; Medical emergency; Mathematics; Engineering","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001851613,0.0001622291,0.0005117153,0.0002197085,0.0003325925,0.0000346632,0.0001117221,0.00002857788,0.00001214037],"category_scores_gemma":[0.0002179017,0.0001646867,0.0001108059,0.0002569962,0.00003660688,0.0001384395,0.0001894556,0.0002131462,0.000003755221],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001128564,"about_ca_system_score_gemma":0.0002397965,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000177645,"about_ca_topic_score_gemma":0.00002223135,"domain_scores_codex":[0.9983783,0.00004758956,0.0004296472,0.0004026242,0.0003946295,0.0003472187],"domain_scores_gemma":[0.9987735,0.0003783185,0.0002053074,0.0003381087,0.0001159295,0.0001888388],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.005481357,0.03995882,0.4373667,0.0527245,0.002461206,0.0000627815,0.03058184,0.005416123,0.04988063,0.00151713,0.07358567,0.3009633],"study_design_scores_gemma":[0.02440987,0.03359854,0.450799,0.007203446,0.001327277,0.0002830387,0.00400684,0.2195919,0.1456747,0.001778754,0.1087343,0.002592255],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9817371,0.00008979841,0.001079165,0.01508222,0.0003194236,0.001390827,0.00008369457,0.0001653056,0.0000525086],"genre_scores_gemma":[0.9968311,0.00001164954,0.0006855871,0.001916007,0.0001750442,0.0001434988,0.0001518488,0.00003922999,0.00004605708],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.298371,"threshold_uncertainty_score":0.6715727,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07115879521134248,"score_gpt":0.3055661731602721,"score_spread":0.2344073779489296,"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."}}