{"id":"W4282965457","doi":"10.1038/s41591-022-01854-8","title":"Reply to: ‘Potential sources of dataset bias complicate investigation of underdiagnosis by machine learning algorithms’ and ‘Confounding factors need to be accounted for in assessing bias by machine learning algorithms’","year":2022,"lang":"en","type":"letter","venue":"Nature Medicine","topic":"COVID-19 diagnosis using AI","field":"Medicine","cited_by":5,"is_retracted":false,"has_abstract":false,"ca_institutions":"Vector Institute; University of Toronto","funders":"","keywords":"Algorithm; Confounding; Machine learning; Computer science; Artificial intelligence; Mathematics; Statistics","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":["metaresearch","metaepi_narrow","research_integrity"],"consensus_categories":[],"category_scores_codex":[0.0026893,0.0009767155,0.002521491,0.002211118,0.0003784952,0.00009341442,0.0005731795,0.00112076,0.0001827368],"category_scores_gemma":[0.009083075,0.0008802003,0.0002022523,0.001950931,0.0003814185,0.0002313006,0.000481919,0.006603329,8.820547e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006630042,"about_ca_system_score_gemma":0.0002743921,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01049727,"about_ca_topic_score_gemma":0.00008370713,"domain_scores_codex":[0.9932184,0.0007028384,0.001761082,0.001556298,0.001931669,0.0008297339],"domain_scores_gemma":[0.9916501,0.005189203,0.001593869,0.0007555775,0.0004157131,0.0003955125],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0002552571,0.0001180078,0.04111914,0.001877471,0.0003970277,0.0001079111,0.001997662,0.0003723766,0.01410817,0.000001380902,0.937064,0.002581593],"study_design_scores_gemma":[0.004694318,0.0020186,0.003463871,0.003197598,0.001128576,0.00006560242,0.001137544,0.007122034,0.006925399,0.0000184869,0.9694154,0.0008125174],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"commentary","genre_gemma":"commentary","genre_scores_codex":[0.1700877,0.004893546,0.001748404,0.8141043,0.0008457065,0.002600473,0.00554165,0.0001702947,0.000007920409],"genre_scores_gemma":[0.1006388,0.0005093915,0.003415799,0.7706417,0.001581119,0.0002839554,0.1221798,0.0004051134,0.0003442777],"genre_candidate":"commentary","genre_consensus":"commentary","teacher_disagreement_score":0.1166381,"threshold_uncertainty_score":0.9993649,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06486493834896838,"score_gpt":0.355887271522923,"score_spread":0.2910223331739546,"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."}}