{"id":"W4377010966","doi":"10.1186/s13000-023-01355-3","title":"Biased data, biased AI: deep networks predict the acquisition site of TCGA images","year":2023,"lang":"en","type":"article","venue":"Diagnostic Pathology","topic":"AI in cancer detection","field":"Computer Science","cited_by":57,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo; Ontario Tech University; Brock University; William Osler Health System; Laurentian University; McMaster University; University Health Network","funders":"Ontario Research Foundation","keywords":"Artificial intelligence; Computer science; Pathology; Computational biology; Medicine; Biology","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.0019751,0.0008896688,0.0004521335,0.001027864,0.0003159106,0.001078598,0.000987235,0.001270914,0.00137891],"category_scores_gemma":[0.01040538,0.0003005831,0.0005226727,0.000744655,0.0006865744,0.001237821,0.001050295,0.001423802,0.0004015557],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001131478,"about_ca_system_score_gemma":0.000908388,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005128816,"about_ca_topic_score_gemma":0.005794426,"domain_scores_codex":[0.9993038,0.0001555645,0.00004457134,0.0002188672,0.0001607042,0.0001165345],"domain_scores_gemma":[0.9963601,0.002039738,0.0004984741,0.0004199379,0.0005205658,0.0001611168],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.002079702,0.0004191117,0.2083962,0.0004846313,0.0004747526,0.0007741469,0.0002510379,0.4016684,0.01448641,0.006830895,0.02191178,0.3422229],"study_design_scores_gemma":[0.00005033194,0.0001472848,0.01866932,0.00004902909,0.00007850122,0.0002286363,0.00005474783,0.9643255,0.007201246,0.007683658,0.001484497,0.00002713261],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8672609,0.003714554,0.1165261,0.001999789,0.000354789,0.0001269902,0.002867485,0.001399835,0.005749621],"genre_scores_gemma":[0.9848762,0.0002740365,0.01073096,0.0002714358,0.00009309016,0.00004317315,0.002078508,0.00004837456,0.001584222],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9980249,"threshold_uncertainty_score":0.01044548,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02352614117836099,"score_gpt":0.2762786573897046,"score_spread":0.2527525162113436,"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."}}