{"id":"W2148059418","doi":"10.1186/1472-6947-10-35","title":"De-identification of primary care electronic medical records free-text data in Ontario, Canada","year":2010,"lang":"en","type":"article","venue":"BMC Medical Informatics and Decision Making","topic":"Electronic Health Records Systems","field":"Health Professions","cited_by":28,"is_retracted":false,"has_abstract":true,"ca_institutions":"National Research Council Canada; Institute for Clinical Evaluative Sciences","funders":"York University","keywords":"Medical record; Text messaging; Protected health information; Identification (biology); Health informatics; Computer science; Context (archaeology); Set (abstract data type); Information retrieval; Medicine; World Wide Web; Public health; Nursing; Health policy","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"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":[0.007615834,0.0004900962,0.0006670674,0.004304981,0.002261661,0.001719949,0.002624939,0.0005195031,0.001841045],"category_scores_gemma":[0.04788559,0.0004490692,0.0005376762,0.009290979,0.0009066291,0.0008027962,0.001628763,0.000587881,0.0003805074],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.06225653,"about_ca_system_score_gemma":0.07222282,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.9941897,"about_ca_topic_score_gemma":0.9950048,"domain_scores_codex":[0.9858236,0.001494385,0.001717103,0.001426603,0.008532591,0.001005717],"domain_scores_gemma":[0.9360934,0.01131488,0.006194974,0.002612337,0.04195906,0.001825409],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0004858061,0.0001623641,0.7678033,0.002766871,0.0003224051,0.0008708937,0.009755068,0.003782703,0.003470354,0.001704574,0.05289656,0.1559792],"study_design_scores_gemma":[0.00004018877,0.00004803195,0.9678963,0.000350921,0.00007239493,0.0001530017,0.001301206,0.004262591,0.00147607,0.0001282215,0.02421073,0.0000602916],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8568684,0.006828999,0.01183958,0.005520693,0.0002454862,0.002614223,0.09579256,0.0009183561,0.01937173],"genre_scores_gemma":[0.9186829,0.003295791,0.01777537,0.001377978,0.00008758194,0.0007835726,0.04630779,0.0001290599,0.01155988],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9941897,"threshold_uncertainty_score":0.4517049,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05042003103964154,"score_gpt":0.39875889996687,"score_spread":0.3483388689272285,"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."}}