{"id":"W3209826882","doi":"10.1148/ryai.2021210027","title":"Deep Learning for Lung Cancer Detection on Screening CT Scans: Results of a Large-Scale Public Competition and an Observer Study with 11 Radiologists","year":2021,"lang":"en","type":"article","venue":"Radiology Artificial Intelligence","topic":"Lung Cancer Diagnosis and Treatment","field":"Medicine","cited_by":50,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia; BC Cancer Agency","funders":"","keywords":"Medicine; Receiver operating characteristic; Radiology; Lung cancer; Lung cancer screening; Test set; Cancer; Nuclear medicine; Retrospective cohort study; Computed tomography; Artificial intelligence; Surgery; Internal medicine; Computer science","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":[],"consensus_categories":[],"category_scores_codex":[0.0004504466,0.0001682946,0.0004067376,0.00009346127,0.0002188801,0.00002127725,0.00005684834,0.00007345241,0.00003899767],"category_scores_gemma":[0.0002250956,0.0001356154,0.00005439755,0.0002487324,0.0001418069,0.00009636302,0.00002171702,0.0001801148,7.501731e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000150102,"about_ca_system_score_gemma":0.00008994908,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002992688,"about_ca_topic_score_gemma":0.01444496,"domain_scores_codex":[0.9984108,0.0002451351,0.0003770778,0.0005460129,0.0001295239,0.0002914366],"domain_scores_gemma":[0.9989243,0.0003035544,0.0001670964,0.000236568,0.0002623034,0.0001062514],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.006578715,0.002955768,0.8070715,0.0002319847,0.0008475962,0.0002571384,0.003541999,0.02868107,0.002642228,0.002844481,0.000036502,0.144311],"study_design_scores_gemma":[0.005539198,0.0263352,0.421985,0.001012201,0.001469642,0.000642653,0.01851169,0.3623149,0.1599941,0.0003458389,0.001002179,0.0008474037],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9224019,0.0007648946,0.07525795,0.0005522037,0.0001541357,0.0007675122,0.0000215591,0.00004344842,0.00003637678],"genre_scores_gemma":[0.9978812,0.0002610776,0.001175705,0.0001061175,0.0001775999,0.0002710393,0.00007834092,0.00001869225,0.00003028077],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3850866,"threshold_uncertainty_score":0.8060627,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05653896743861862,"score_gpt":0.3564196165415092,"score_spread":0.2998806491028906,"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."}}