{"id":"W4407599026","doi":"10.1038/s41746-025-01499-0","title":"Self-supervised identification and elimination of harmful datasets in distributed machine learning for medical image analysis","year":2025,"lang":"en","type":"article","venue":"npj Digital Medicine","topic":"COVID-19 diagnosis using AI","field":"Medicine","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Institut Universitaire de Gériatrie de Montréal; Hotchkiss Brain Institute; Alberta Children's Hospital; Women and Children’s Health Research Institute; Université de Montréal; University of Alberta; University of Calgary","funders":"Parkinson Association of Alberta; Parkinson Vereniging; Consortium canadien en neurodégénérescence associée au vieillissement; Alzheimer's Disease Neuroimaging Initiative","keywords":"Identification (biology); Computer science; Artificial intelligence; Machine learning; Pattern recognition (psychology); Image (mathematics); 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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.0007617918,0.0001522789,0.0005516469,0.0007592788,0.00004525008,0.00002852982,0.0001178028,0.0001068488,0.00005948403],"category_scores_gemma":[0.008353689,0.0001307559,0.00007170853,0.001410558,0.0001618014,0.0002187126,0.00007358444,0.0001894169,0.000001686942],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001225933,"about_ca_system_score_gemma":0.0001191528,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001138474,"about_ca_topic_score_gemma":0.0000590821,"domain_scores_codex":[0.9982594,0.00004273914,0.0006647216,0.0003742005,0.0004774063,0.0001815539],"domain_scores_gemma":[0.9982768,0.0009341853,0.0001523981,0.0002867408,0.0002165811,0.0001333176],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.002046252,0.004003637,0.5742792,0.008927044,0.003385105,0.0002106385,0.002678243,0.0001568105,0.02006448,0.001986312,0.03305411,0.3492081],"study_design_scores_gemma":[0.01348259,0.0007368512,0.518168,0.001705552,0.003699475,0.00001579694,0.0005511821,0.4201865,0.003346506,0.0003944203,0.03739542,0.0003176658],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6523144,0.001860826,0.1111667,0.2291819,0.0003173791,0.002482052,0.001837364,0.0003271542,0.0005122946],"genre_scores_gemma":[0.9872491,0.00015045,0.0003730879,0.001060298,0.00004659131,0.00004631213,0.01099024,0.00001236444,0.00007151961],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.4200297,"threshold_uncertainty_score":0.9999993,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01368227694001169,"score_gpt":0.3329284566835312,"score_spread":0.3192461797435195,"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."}}