{"id":"W4289130391","doi":"10.1007/s00234-022-03029-1","title":"Rule-based natural language processing for automation of stroke data extraction: a validation study","year":2022,"lang":"en","type":"article","venue":"Neuroradiology","topic":"Acute Ischemic Stroke Management","field":"Medicine","cited_by":18,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Toronto; Sunnybrook Health Science Centre; University of Ottawa; Health Sciences Centre; Ottawa Hospital","funders":"","keywords":"Medicine; Neuroradiology; Occlusion; Collateral circulation; Stroke (engine); Radiology; Angiography; Computed tomography angiography; Data extraction; Artificial intelligence; Natural language processing; Neurology; MEDLINE; Internal medicine; Computer science","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":[0.0134718,0.001095918,0.0007695023,0.001963828,0.0007054983,0.00149661,0.001709017,0.00100215,0.002681412],"category_scores_gemma":[0.03520728,0.0004349789,0.001502503,0.001107806,0.0007559053,0.001435958,0.001497819,0.001033007,0.001439276],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006463548,"about_ca_system_score_gemma":0.002016426,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005565896,"about_ca_topic_score_gemma":0.004234271,"domain_scores_codex":[0.9926026,0.003745918,0.001202903,0.001167102,0.001027054,0.0002543203],"domain_scores_gemma":[0.9434918,0.04326954,0.001140766,0.005012992,0.006709456,0.0003753664],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.01154526,0.01429556,0.1199957,0.0041367,0.002563401,0.003041265,0.005440951,0.03509761,0.06767245,0.001459714,0.008041703,0.7267097],"study_design_scores_gemma":[0.005843865,0.02140601,0.2538245,0.001082414,0.004826486,0.008739486,0.004825096,0.4615557,0.1927283,0.005480628,0.03921564,0.0004719183],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9123212,0.001275797,0.07265364,0.0002316765,0.00009328931,0.002520722,0.005704767,0.002358924,0.002839971],"genre_scores_gemma":[0.8536214,0.000604792,0.1283454,0.0002139297,0.00003993115,0.001118244,0.01404488,0.0002310385,0.001780511],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.0134718,"threshold_uncertainty_score":0.0712465,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03996540448498499,"score_gpt":0.3494512144862339,"score_spread":0.3094858100012489,"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."}}