{"id":"W3200840849","doi":"10.1038/s41591-021-01506-3","title":"Federated learning for predicting clinical outcomes in patients with COVID-19","year":2021,"lang":"en","type":"article","venue":"Nature Medicine","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":720,"is_retracted":false,"has_abstract":true,"ca_institutions":"Public Health Ontario; Toronto Public Health; University of Toronto; University Health Network; Schwartz/Reisman Emergency Medicine Institute; Sinai Health System; Lunenfeld-Tanenbaum Research Institute","funders":"National Cancer Institute; National Heart, Lung, and Blood Institute; NIHR Cambridge Biomedical Research Centre; Faculty of Medicine, Chulalongkorn University; Engineering and Physical Sciences Research Council; Genentech; National Health Insurance Administration; National Institutes of Health; Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital; Ministry of Science and Technology, Taiwan; Cambridge University Hospitals; Centre d'Imagerie BioMédicale; Colgate-Palmolive Company; Brigham and Women's Hospital; Chulalongkorn University; Massachusetts General Hospital; National Institute for Health and Care Research; University of California, San Francisco; U.S. National Library of Medicine; University of Cambridge; Foundation for the National Institutes of Health; Doris Duke Charitable Foundation; Cancer Research UK; National Taiwan University; National Center for Theoretical Sciences; American Association for Dental, Oral, and Craniofacial Research; Department of Health and Social Care; Children's National Hospital","keywords":"Generalizability theory; Coronavirus disease 2019 (COVID-19); Predictive modelling; Computer science; Artificial intelligence; Medicine; Emergency medicine; Medical emergency; Machine learning; Statistics; Internal 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":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch","open_science"],"consensus_categories":[],"category_scores_codex":[0.001135623,0.0001611158,0.0003987059,0.000138463,0.0001498634,0.00004493498,0.004405031,0.0003738036,0.00001254091],"category_scores_gemma":[0.2557735,0.0001109829,0.00003641174,0.0007330129,0.000118352,0.0002297606,0.008539372,0.001469008,0.000001696124],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001083349,"about_ca_system_score_gemma":0.0002318567,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002195972,"about_ca_topic_score_gemma":0.00008815026,"domain_scores_codex":[0.9980146,0.0001473637,0.0004210136,0.0006320535,0.0004612808,0.0003236957],"domain_scores_gemma":[0.9957207,0.00184208,0.0001782804,0.001883333,0.0002444044,0.0001311548],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00003670577,0.00008179017,0.9596091,0.00003114811,0.00003245462,0.00004377807,0.00005164629,0.00001165836,0.000005378274,0.0002212736,0.03067966,0.009195414],"study_design_scores_gemma":[0.01410221,0.001374785,0.8784614,0.0003684039,0.00003775841,0.00001448502,0.0002196394,0.06366412,0.0001978848,0.01500695,0.02608558,0.0004668593],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3980132,0.0006186349,0.4086292,0.1886375,0.001726693,0.0007228623,0.0000165243,0.001265553,0.000369867],"genre_scores_gemma":[0.9068879,0.00002009004,0.08793009,0.004804924,0.00009775978,0.0000223486,0.0001347336,0.00001565641,0.00008648269],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.5088747,"threshold_uncertainty_score":0.9994794,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0359123450077511,"score_gpt":0.363648147362923,"score_spread":0.3277358023551719,"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."}}