{"id":"W2899847761","doi":"10.1016/j.radonc.2018.10.027","title":"Vulnerabilities of radiomic signature development: The need for safeguards","year":2018,"lang":"en","type":"article","venue":"Radiotherapy and Oncology","topic":"Radiomics and Machine Learning in Medical Imaging","field":"Medicine","cited_by":317,"is_retracted":false,"has_abstract":false,"ca_institutions":"Vector Institute; Princess Margaret Cancer Centre; Institute of Cancer Research; Ontario Institute for Cancer Research; University of Toronto; University Health Network","funders":"Canadian Institutes of Health Research; Terry Fox Research Institute; Ontario Institute for Cancer Research","keywords":"Voxel; Computer science; Concordance; Volume (thermodynamics); Artificial intelligence; Head and neck; Signature (topology); Concordance correlation coefficient; Pattern recognition (psychology); Index (typography); Data mining; Medicine; Medical physics; Statistics; Mathematics; Surgery","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":[],"consensus_categories":[],"category_scores_codex":[0.0006686129,0.0001068978,0.0003466404,0.00006866561,0.0001724698,0.00000780664,0.00009257536,0.0001310928,0.0001334323],"category_scores_gemma":[0.0001624462,0.00006568729,0.00006113059,0.00008396259,0.0006295299,0.00002416852,0.00001235796,0.0002255862,7.306131e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001177101,"about_ca_system_score_gemma":0.0002418462,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001346648,"about_ca_topic_score_gemma":0.000009046217,"domain_scores_codex":[0.9992242,0.0000741108,0.0002406727,0.000167519,0.00009516523,0.00019826],"domain_scores_gemma":[0.9991735,0.0004410347,0.00008715074,0.0001471791,0.00007988387,0.00007121592],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.002120214,0.0002137753,0.00578023,0.0002653415,0.0004719651,0.000008725527,0.01528759,0.00001125315,0.08167502,0.007159328,0.02382097,0.8631856],"study_design_scores_gemma":[0.003970989,0.001663019,0.004509893,0.00004943522,0.00005912434,0.0001819592,0.0003998184,0.006499417,0.006030632,0.0006854701,0.9758326,0.0001176902],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9420696,0.006645915,0.0251763,0.02144662,0.0008461648,0.0008199692,0.000003763148,0.00005187033,0.002939734],"genre_scores_gemma":[0.9566757,0.000446168,0.03471285,0.005651655,0.001059497,0.00004247903,0.000009179936,0.0000286836,0.001373779],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9520116,"threshold_uncertainty_score":0.267865,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01276414941444599,"score_gpt":0.321780288592158,"score_spread":0.309016139177712,"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."}}