{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.004324294,0.0005729503,0.0008822834,0.0009322641,0.0007956821,0.001282112,0.001861224,0.0009576091,0.001060901],"category_scores_gemma":[0.01312786,0.0003995243,0.0007220856,0.0007176942,0.001343164,0.001886613,0.002667837,0.001555561,0.0007043926],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006489439,"about_ca_system_score_gemma":0.001520433,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001181724,"about_ca_topic_score_gemma":0.002182847,"domain_scores_codex":[0.9970275,0.001050932,0.0001786742,0.0007538071,0.00081004,0.0001791945],"domain_scores_gemma":[0.9898184,0.003356419,0.001103931,0.003727172,0.0017201,0.0002738771],"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":[0.0007010729,0.0006997487,0.01639298,0.0002296405,0.0002263715,0.0003417883,0.001001357,0.1726241,0.0387179,0.01188976,0.00714328,0.7500321],"study_design_scores_gemma":[0.00004728311,0.0001127518,0.002591336,0.00002179145,0.00002953669,0.0002446195,0.0001347295,0.9544533,0.02221977,0.01776445,0.002358745,0.00002165208],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02789582,0.0001077272,0.969725,0.0002042142,0.00002285075,0.00008141741,0.00004969922,0.00123585,0.000677434],"genre_scores_gemma":[0.6283448,0.00008741374,0.3681334,0.0002814609,0.00008091838,0.0002579555,0.000463675,0.000339361,0.002011099],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004324294,"threshold_uncertainty_score":0.02286929,"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."}}