{"id":"W4409148359","doi":"10.1038/s41598-025-89210-x","title":"Investigation on potential bias factors in histopathology datasets","year":2025,"lang":"en","type":"article","venue":"Scientific Reports","topic":"AI in cancer detection","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"Wilfrid Laurier University; Brock University; Ontario Tech University","funders":"","keywords":"Computer science; Artificial intelligence; Digital pathology; Metric (unit); Machine learning; Precision medicine; Deep learning; Pattern recognition (psychology); Artificial neural network; Data mining; Pathology; Medicine","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":[0.01301677,0.0007750907,0.0006489488,0.001606361,0.0008280271,0.001160064,0.001253731,0.001169304,0.0006417603],"category_scores_gemma":[0.04048891,0.0002662356,0.0006088449,0.001927023,0.0008525143,0.00114687,0.001172046,0.0008823348,0.0004075296],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001380539,"about_ca_system_score_gemma":0.001144003,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005423905,"about_ca_topic_score_gemma":0.008887743,"domain_scores_codex":[0.9932246,0.002042143,0.0007560352,0.00158672,0.001963198,0.0004273861],"domain_scores_gemma":[0.974036,0.01447395,0.002238964,0.003758704,0.005172096,0.0003203856],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.001863155,0.0006076822,0.5949966,0.00116237,0.001080116,0.001436679,0.0008754689,0.1058754,0.03674186,0.007777699,0.0270169,0.2205661],"study_design_scores_gemma":[0.0002338363,0.0006552945,0.2974591,0.0004137564,0.0004595239,0.004175933,0.00121874,0.52983,0.09594623,0.02098659,0.04842527,0.0001958324],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9423087,0.003263701,0.04207186,0.001451288,0.0003341186,0.0002090746,0.006432301,0.001422878,0.002506089],"genre_scores_gemma":[0.9534711,0.0005372165,0.0254555,0.0005912945,0.00008256609,0.0001299381,0.01826278,0.0002351627,0.001234469],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9869832,"threshold_uncertainty_score":0.06884009,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04264991251651127,"score_gpt":0.267237179877191,"score_spread":0.2245872673606797,"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."}}