{"id":"W4406694526","doi":"10.1093/jamiaopen/ooae152","title":"pyDeid: an improved, fast, flexible, and generalizable rule-based approach for deidentification of free-text medical records","year":2024,"lang":"en","type":"article","venue":"JAMIA Open","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto; Hospital for Sick Children; St. Michael's Hospital","funders":"Natural Sciences and Engineering Research Council of Canada; Canadian Institutes of Health Research; Canadian Frailty Network; Canadian Cancer Society","keywords":"Computer science; Python (programming language); Software; Precision and recall; Identification (biology); Benchmark (surveying); Data mining; Artificial intelligence; Extant taxon; Flexibility (engineering); Machine learning; Recall; Information retrieval; Natural language processing; Programming language","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001683001,0.0001447002,0.0002399539,0.0001418647,0.0001430268,0.0006338421,0.002137402,0.0001512421,0.00005382363],"category_scores_gemma":[0.000245976,0.000130306,0.00004797158,0.0003915381,0.00005860475,0.0006879665,0.0006207425,0.000233184,0.000005484435],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000497753,"about_ca_system_score_gemma":0.0004857267,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003085462,"about_ca_topic_score_gemma":0.0001401028,"domain_scores_codex":[0.9981124,0.0001973224,0.0003881083,0.0006900372,0.0003448371,0.0002672453],"domain_scores_gemma":[0.9984302,0.000183462,0.0001091574,0.0009470065,0.0001290343,0.0002011555],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001347027,0.0003255332,0.005620212,0.003217364,0.00007415866,0.00001202377,0.001463236,0.00292363,0.001970823,0.1063785,0.0214723,0.8564075],"study_design_scores_gemma":[0.0005119529,0.0001812391,0.001269193,0.00007837242,0.0000090793,0.0000106863,0.00002309978,0.9836561,0.0008189189,0.002768553,0.01051836,0.0001544568],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.008819719,0.0004796029,0.9837967,0.003409967,0.00048887,0.001062086,0.00002983442,0.000190942,0.001722334],"genre_scores_gemma":[0.1656938,0.00003248439,0.828832,0.0008689172,0.0003260632,0.0005526819,0.0001988949,0.00005536356,0.003439738],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9807324,"threshold_uncertainty_score":0.6112154,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03624260455257254,"score_gpt":0.3462986527902708,"score_spread":0.3100560482376983,"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."}}