{"id":"W4205679507","doi":"10.1007/s10140-022-02019-3","title":"Deep learning prediction of sex on chest radiographs: a potential contributor to biased algorithms","year":2022,"lang":"en","type":"article","venue":"Emergency Radiology","topic":"COVID-19 diagnosis using AI","field":"Medicine","cited_by":17,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Ottawa","funders":"Johns Hopkins University","keywords":"Convolutional neural network; Artificial intelligence; Receiver operating characteristic; Medicine; Test set; Undersampling; Machine learning; Deep learning; Radiography; Set (abstract data type); Pattern recognition (psychology); F1 score; Computer science; Radiology","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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0005869112,0.0001956685,0.0005167847,0.0005763251,0.0002465004,0.000002592862,0.0001676691,0.0001089942,0.002055738],"category_scores_gemma":[0.0005428573,0.0002072455,0.0002458016,0.0007976058,0.00006221911,0.00002678549,0.00008452975,0.0004772026,0.00002246803],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002223242,"about_ca_system_score_gemma":0.0000921169,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001885723,"about_ca_topic_score_gemma":0.000005460762,"domain_scores_codex":[0.9978114,0.0003962276,0.0005351024,0.0004977705,0.0003501399,0.0004093376],"domain_scores_gemma":[0.9990615,0.0001550812,0.0001727332,0.0003464979,0.00007953167,0.0001846258],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"not_applicable","study_design_scores_codex":[0.003827574,0.003470524,0.310251,0.0003238463,0.001865265,0.0006795295,0.008425049,0.1862038,0.1898461,0.0002641538,0.2339983,0.0608449],"study_design_scores_gemma":[0.008200783,0.01675306,0.3099603,0.00007466594,0.0006767769,0.0005773895,0.001225518,0.06358714,0.004316794,0.0001366436,0.593838,0.0006529152],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9728398,0.001220681,0.006884489,0.01324889,0.004102149,0.001133669,0.0001143549,0.0002896788,0.0001662246],"genre_scores_gemma":[0.9956489,0.0002268424,0.0004292796,0.002187853,0.0005648122,0.000261645,0.0002340044,0.00004323084,0.0004033978],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3598397,"threshold_uncertainty_score":0.9988565,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02010145942009194,"score_gpt":0.2892124169157456,"score_spread":0.2691109574956536,"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."}}