{"id":"W4385074164","doi":"10.1007/s11042-023-16029-x","title":"Artificial intelligence bias in medical system designs: a systematic review","year":2023,"lang":"en","type":"review","venue":"Multimedia Tools and Applications","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":58,"is_retracted":false,"has_abstract":false,"ca_institutions":"Queen's University","funders":"","keywords":"Computer science; Artificial intelligence; Machine learning; Selection bias; Ranking (information retrieval); Process (computing); Protocol (science); Data mining; 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":["metaepi_narrow","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.001994112,0.0003888613,0.002676078,0.0003409613,0.000128966,0.00006091185,0.0002882296,0.0005141586,0.00006542993],"category_scores_gemma":[0.002698631,0.0002933404,0.0002879284,0.001368638,0.0001278863,0.00006522387,0.00007035953,0.0006734325,0.001614466],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002809832,"about_ca_system_score_gemma":0.0009804723,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001518144,"about_ca_topic_score_gemma":0.00007644914,"domain_scores_codex":[0.9954244,0.0003072421,0.002716529,0.0006253879,0.0005370999,0.0003893097],"domain_scores_gemma":[0.9955421,0.002645862,0.0005355152,0.0006919514,0.0001621177,0.0004224739],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"systematic_review","study_design_gemma":"systematic_review","study_design_scores_codex":[9.182286e-7,0.00004896436,9.184883e-7,0.5015118,0.00002716392,0.00001234728,0.00004884736,5.931873e-8,2.34722e-8,0.0008144107,0.00005855396,0.497476],"study_design_scores_gemma":[0.000008713399,0.00003875317,8.775278e-7,0.9505286,0.001640412,0.000208875,0.0005392961,0.0009137802,0.000002156395,0.0002330747,0.04557569,0.0003097227],"study_design_candidate":"systematic_review","study_design_consensus":"systematic_review","genre_codex":"review","genre_gemma":"review","genre_scores_codex":[6.711883e-7,0.9852358,0.002618729,0.0006591438,0.0001645401,0.01097227,0.00005940183,0.0001755551,0.0001139568],"genre_scores_gemma":[0.00001610884,0.9821869,0.0004011237,0.0002729194,0.0005438453,0.0159749,0.0004319817,0.00006483317,0.0001074089],"genre_candidate":"review","genre_consensus":"review","teacher_disagreement_score":0.4971663,"threshold_uncertainty_score":0.9999519,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.6323140650078787,"score_gpt":0.5271705183621161,"score_spread":0.1051435466457625,"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."}}