{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.01324194,0.001157511,0.002279149,0.001495791,0.005829215,0.006948196,0.00281099,0.09119614,0.008800512],"category_scores_gemma":[0.1077052,0.001621756,0.001812238,0.001792484,0.005299303,0.00441802,0.003070825,0.05738476,0.01143685],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.006157,"about_ca_system_score_gemma":0.008546565,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01258488,"about_ca_topic_score_gemma":0.02009386,"domain_scores_codex":[0.986573,0.004113,0.002425664,0.001759431,0.003245022,0.00188389],"domain_scores_gemma":[0.9410685,0.03876164,0.004173238,0.001579697,0.01026487,0.00415217],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00003019998,0.000008952726,0.0006847444,0.00004728342,0.00001852456,0.0003587914,0.000129517,0.00002596221,0.00005340976,0.001107485,0.9948556,0.002679451],"study_design_scores_gemma":[0.000308668,0.00009942069,0.004495674,0.001039816,0.0001193859,0.001987288,0.0008525393,0.0007534841,0.0004107984,0.01328627,0.9764383,0.000208328],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"commentary","genre_gemma":"commentary","genre_scores_codex":[0.0001383774,0.0003027226,0.00007306498,0.9888777,0.009814151,0.000010519,0.0000886862,0.00001644998,0.0006782861],"genre_scores_gemma":[0.000722899,0.0001413719,0.00009654223,0.985358,0.01245051,0.0000252821,0.00002529747,0.00001160351,0.001168436],"genre_candidate":"commentary","genre_consensus":"commentary","teacher_disagreement_score":0.9867581,"threshold_uncertainty_score":0.07003093,"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."}}