{"id":"W4200191097","doi":"10.1038/s41591-021-01595-0","title":"Underdiagnosis bias of artificial intelligence algorithms applied to chest radiographs in under-served patient populations","year":2021,"lang":"en","type":"article","venue":"Nature Medicine","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":784,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute; University of Toronto","funders":"Natural Sciences and Engineering Research Council of Canada; Canadian Institute for Advanced Research; Microsoft Research","keywords":"Context (archaeology); Socioeconomic status; Medicine; Artificial intelligence; Exacerbation; Computer science; Machine learning; Population; Internal medicine","routes":{"ca_aff":true,"ca_fund":true,"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":[0.0004570464,0.0002004967,0.0005394999,0.0005977442,0.00007667761,0.000008731452,0.0001087803,0.0003647101,0.0003060751],"category_scores_gemma":[0.001296525,0.0001732505,0.00009434181,0.002610167,0.000129752,0.00004466619,0.00003812291,0.0006970843,0.00002275317],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001850932,"about_ca_system_score_gemma":0.0003017265,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001213461,"about_ca_topic_score_gemma":0.002278818,"domain_scores_codex":[0.9975612,0.00007821926,0.001036318,0.0004415326,0.0005465026,0.0003362122],"domain_scores_gemma":[0.998219,0.0003990184,0.0001664704,0.0004524612,0.0004834562,0.0002795846],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0004008515,0.001596906,0.05851477,0.0004270811,0.00009663678,0.0001015752,0.01353594,0.00136677,0.01414612,0.03783616,0.002701532,0.8692756],"study_design_scores_gemma":[0.0003661321,0.00229495,0.2813138,0.002794728,0.0004134694,0.000115974,0.05754814,0.002058147,0.564863,0.08232947,0.005058393,0.0008437886],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9435288,0.002390518,0.008591427,0.04171922,0.001773923,0.0009182831,0.000009718652,0.00005378665,0.001014269],"genre_scores_gemma":[0.9896603,0.0002154589,0.005453867,0.003872192,0.0005136766,0.00007102556,0.0001571591,0.00002414519,0.00003215307],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8684319,"threshold_uncertainty_score":0.7064952,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2655884020617279,"score_gpt":0.4424643733409022,"score_spread":0.1768759712791743,"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."}}