{"id":"W4301396488","doi":"10.5334/dsj-2022-017","title":"A Survey on Publicly Available Open Datasets Derived From Electronic Health Records (EHRs) of Patients with Neuroblastoma","year":2022,"lang":"en","type":"article","venue":"Data Science Journal","topic":"Neuroblastoma Research and Treatments","field":"Medicine","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Neuroblastoma; Computer science; Health records; Open science; License; The Internet; Electronic health record; Open data; World Wide Web; Internet privacy; Medicine; Information retrieval; Data mining; Political science; Mathematics; Statistics; Health care","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[{"model":"gemma","categories":["open_science"],"domain":null,"study_design":"observational","genre":"empirical","about_ca_system":false,"about_ca_topic":false,"confidence":"low","status":"direct model label, unvalidated"},{"model":"gpt","categories":[],"domain":null,"study_design":"design_other","genre":"review","about_ca_system":false,"about_ca_topic":false,"confidence":"low","status":"direct model label, unvalidated"}],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["open_science"],"consensus_categories":[],"category_scores_codex":[0.01021409,0.0006292441,0.0009444067,0.008501466,0.000736537,0.002241986,0.001486131,0.001364635,0.003123239],"category_scores_gemma":[0.06579199,0.0003146983,0.001036238,0.01566376,0.000608556,0.003039402,0.003758419,0.0009651149,0.001658581],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00127497,"about_ca_system_score_gemma":0.003022275,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006003798,"about_ca_topic_score_gemma":0.008588018,"domain_scores_codex":[0.9803949,0.005603066,0.004827285,0.002252272,0.006014243,0.0009082492],"domain_scores_gemma":[0.9168954,0.04896139,0.01248089,0.006739492,0.01279014,0.002132711],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"observational","study_design_scores_codex":[0.0009726149,0.0002032,0.3698104,0.0102693,0.0007011017,0.0004307862,0.001907235,0.001310113,0.001968987,0.006593823,0.4383466,0.1674857],"study_design_scores_gemma":[0.00008956951,0.000116908,0.4225538,0.004722445,0.0003544003,0.001405545,0.003630559,0.001526093,0.00203295,0.003519495,0.5598889,0.0001593129],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"empirical","genre_scores_codex":[0.102447,0.01776408,0.005774887,0.01040243,0.0005151989,0.0003713639,0.8513907,0.0009353296,0.01039908],"genre_scores_gemma":[0.08552543,0.008559856,0.007349293,0.003592142,0.0002725078,0.0008958807,0.8925357,0.0002217718,0.001047476],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9985139,"threshold_uncertainty_score":0.0540179,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07305559528067233,"score_gpt":0.3504970068041415,"score_spread":0.2774414115234691,"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."}}