{"id":"W4223973033","doi":"10.1101/2022.03.29.22273138","title":"A proposed de-identification framework for a cohort of children presenting at a health facility in Uganda","year":2022,"lang":"en","type":"preprint","venue":"medRxiv","topic":"Ethics in Clinical Research","field":"Medicine","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Provincial Health Services Authority; BC Children's Hospital; University of British Columbia","funders":"","keywords":"Data anonymization; Identification (biology); Anonymity; Computer science; Identifier; Data sharing; Confidentiality; Context (archaeology); Raw data; Data science; Data mining; Computer security; Information privacy; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.08191715,0.0007362938,0.0008246149,0.005204189,0.002926644,0.00412558,0.003917033,0.002186133,0.002918541],"category_scores_gemma":[0.08198228,0.0005851408,0.001903693,0.003051468,0.003058956,0.005323331,0.007082498,0.00273027,0.0004135118],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.004422972,"about_ca_system_score_gemma":0.008106752,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01233702,"about_ca_topic_score_gemma":0.01142572,"domain_scores_codex":[0.9580633,0.02892069,0.003303052,0.00470338,0.003884849,0.001124709],"domain_scores_gemma":[0.9298283,0.0430203,0.008247569,0.005900683,0.01134887,0.001654216],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0003412355,0.0004429976,0.0899568,0.00116718,0.000298573,0.002536004,0.02372005,0.08376918,0.001669313,0.652835,0.007368617,0.1358951],"study_design_scores_gemma":[0.00007810278,0.0003871354,0.01791675,0.001114454,0.000167817,0.001884222,0.01350781,0.5245044,0.001741979,0.3946238,0.04383884,0.0002346917],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.04338636,0.0004987935,0.9420973,0.006515608,0.000111004,0.001435335,0.001116092,0.0002393518,0.004600139],"genre_scores_gemma":[0.4218272,0.0003626958,0.571174,0.0005959162,0.00008952671,0.002007831,0.00152979,0.00003457288,0.002378437],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.08191715,"threshold_uncertainty_score":0.4332245,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2847652170274104,"score_gpt":0.5475223049897544,"score_spread":0.262757087962344,"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."}}