{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0009090864,0.000153904,0.0002193307,0.0001682028,0.0001530842,0.00005265504,0.0012266,0.0001329572,0.0000241109],"category_scores_gemma":[0.00133595,0.0001247218,0.00005463947,0.001011492,0.000218683,0.0003677709,0.0007420311,0.0002297717,0.0001418519],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003586074,"about_ca_system_score_gemma":0.00005881132,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000463045,"about_ca_topic_score_gemma":0.00003055304,"domain_scores_codex":[0.9981105,0.0003647717,0.0003249794,0.0005656687,0.0002500706,0.0003839754],"domain_scores_gemma":[0.9942196,0.003796395,0.0001835007,0.001638828,0.0001027786,0.00005891014],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001573107,0.0005068739,0.1369058,0.0001955137,0.0002174465,0.003256167,0.00429258,0.2620103,0.01404544,0.004047996,0.1868643,0.3875003],"study_design_scores_gemma":[0.0005532153,0.0002516976,0.3247913,0.00005228701,0.00004571431,0.0000856695,0.00003510179,0.6696893,0.001749297,0.001848428,0.0007012191,0.0001968197],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0767277,0.000449571,0.9176879,0.002419238,0.001740748,0.0003446212,0.0001181934,0.0004062716,0.000105744],"genre_scores_gemma":[0.9964461,0.000326942,0.001417097,0.001166141,0.0002985712,0.0000917026,0.0002184272,0.00001710081,0.00001787153],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9197184,"threshold_uncertainty_score":0.5086009,"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."}}