{"id":"W4223430324","doi":"10.1038/s41746-022-00592-y","title":"Machine learning for medical imaging: methodological failures and recommendations for the future","year":2022,"lang":"en","type":"review","venue":"npj Digital Medicine","topic":"Radiomics and Machine Learning in Medical Imaging","field":"Medicine","cited_by":593,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; Mila - Quebec Artificial Intelligence Institute; Institut National de la Recherche Scientifique","funders":"Novo Nordisk Fonden; Agence Nationale de la Recherche","keywords":"Medical imaging; Psychology; Computer science; Data science; Engineering ethics; Artificial intelligence; Medical physics; Medicine; Engineering","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":[0.07100752,0.00179267,0.004469038,0.00658793,0.001324966,0.007338403,0.004430864,0.007990878,0.006985371],"category_scores_gemma":[0.1007452,0.0009965712,0.002535483,0.006217635,0.006535329,0.02281191,0.004184483,0.01632131,0.003998844],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00519385,"about_ca_system_score_gemma":0.01748333,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004968684,"about_ca_topic_score_gemma":0.009531471,"domain_scores_codex":[0.9848187,0.008697296,0.001995043,0.0009743439,0.003109624,0.0004048585],"domain_scores_gemma":[0.830795,0.1317327,0.005868307,0.005311588,0.02340265,0.00288972],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0001301855,0.00009085742,0.0005165815,0.02999918,0.0003849722,0.0002062559,0.0004127741,0.000931376,0.0002685087,0.07921099,0.1258525,0.7619958],"study_design_scores_gemma":[0.00009148468,0.0001388997,0.001126763,0.09773624,0.0004811798,0.0004531492,0.001097697,0.001153104,0.000251897,0.2014179,0.6958946,0.0001570932],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"review","genre_gemma":"review","genre_scores_codex":[0.0000654709,0.9165362,0.001941815,0.07861148,0.002233511,0.00002333284,0.00003748711,0.00003705027,0.0005134882],"genre_scores_gemma":[0.001373544,0.9734806,0.007403235,0.01390731,0.003165621,0.0001066986,0.00005435769,0.00003049388,0.0004781933],"genre_candidate":"review","genre_consensus":"review","teacher_disagreement_score":0.9289925,"threshold_uncertainty_score":0.3755282,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1037579113986433,"score_gpt":0.4339280365929593,"score_spread":0.330170125194316,"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."}}