{"id":"W3011720109","doi":"10.1038/s41598-020-61297-4","title":"Privacy-preserving distributed learning of radiomics to predict overall survival and HPV status in head and neck cancer","year":2020,"lang":"en","type":"article","venue":"Scientific Reports","topic":"Radiomics and Machine Learning in Medical Imaging","field":"Medicine","cited_by":80,"is_retracted":false,"has_abstract":true,"ca_institutions":"Princess Margaret Cancer Centre; University of Toronto","funders":"National Institute of Biomedical Imaging and Bioengineering; Stichting voor de Technische Wetenschappen; National Institute of Dental and Craniofacial Research; Center for Translational Molecular Medicine; National Cancer Institute; KWF Kankerbestrijding; National Institutes of Health; Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung; Nederlandse Organisatie voor Wetenschappelijk Onderzoek; Health Foundation Limburg; National Science Foundation; Eurostars; Horizon 2020 Framework Programme; Cancer Research UK; Andrew Sabin Family Foundation; Division of Mathematical Sciences; University of Texas MD Anderson Cancer Center; Interreg; Elekta","keywords":"Radiomics; Feature selection; Logistic regression; Computer science; Workflow; Feature (linguistics); Artificial intelligence; Cluster analysis; Machine learning; Receiver operating characteristic; Head and neck cancer; Data mining; Cancer; Medicine; Database; 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":[],"consensus_categories":[],"category_scores_codex":[0.0009062895,0.0001315366,0.0003920156,0.0001268422,0.0001000518,0.00009570921,0.00007695366,0.00005167912,0.00005453054],"category_scores_gemma":[0.002801213,0.0001193236,0.0000368469,0.0004263765,0.0001740139,0.0001039866,0.000273358,0.0003705837,4.61515e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005571391,"about_ca_system_score_gemma":0.0001808247,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005898404,"about_ca_topic_score_gemma":0.00002351598,"domain_scores_codex":[0.9980863,0.00005762597,0.0004880429,0.0005982284,0.0004220639,0.0003477596],"domain_scores_gemma":[0.9988782,0.00006258334,0.0001815345,0.0002838534,0.00008502779,0.0005087435],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00008766112,0.00003292417,0.967342,0.0002370475,0.00002564374,0.0003283704,0.001846794,0.001788833,0.01827269,0.0000210474,0.002365182,0.007651844],"study_design_scores_gemma":[0.001855883,0.0002343662,0.8104051,0.0005229705,0.00006441648,0.0001586596,0.0003534996,0.1202737,0.001294391,0.0003333336,0.06424458,0.0002591445],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9930893,0.000661549,0.0009482949,0.003905085,0.0007439101,0.0003161092,0.000007056644,0.00003960308,0.0002891537],"genre_scores_gemma":[0.9978401,0.0001081162,0.001550757,0.0001480942,0.00008306437,0.000007428434,0.00005888794,0.00001908698,0.0001844541],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1569369,"threshold_uncertainty_score":0.4865876,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01776262357648538,"score_gpt":0.2961112534293367,"score_spread":0.2783486298528513,"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."}}