{"id":"W4415077526","doi":"10.1007/978-3-032-07502-4_2","title":"SCOPE: Label Extraction of Stroke Diagnosis from Unstructured Medical Reports Using Retrieval-Augmented Generation","year":2025,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"Foothills Medical Centre; University of Calgary","funders":"","keywords":"Parsing; Scope (computer science); Information extraction; Data extraction; Medical imaging; Code (set theory); Language model; Deep learning","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":[0.001089869,0.001469571,0.0008748339,0.003800253,0.000644464,0.001701009,0.001185763,0.001625719,0.007045868],"category_scores_gemma":[0.002449863,0.0004539045,0.001616871,0.00183235,0.0004670646,0.001448363,0.001649503,0.0008922954,0.006947071],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004850527,"about_ca_system_score_gemma":0.001338641,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003573503,"about_ca_topic_score_gemma":0.005721258,"domain_scores_codex":[0.9991397,0.0001678575,0.00007526136,0.0002855065,0.0002454745,0.00008620352],"domain_scores_gemma":[0.9986873,0.0005331497,0.00009163043,0.000252694,0.0003705753,0.00006459493],"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":[0.0005217559,0.0002428791,0.002753386,0.0007174144,0.000154251,0.0006210679,0.0002941972,0.004287087,0.0571051,0.001824413,0.06513822,0.8663402],"study_design_scores_gemma":[0.000427157,0.0009446195,0.02755125,0.0004939755,0.0009665475,0.004188569,0.000789074,0.6413876,0.165835,0.02409816,0.1330462,0.0002718393],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04917448,0.003533118,0.876469,0.001125892,0.0009802668,0.00121568,0.01506269,0.04318258,0.009256165],"genre_scores_gemma":[0.1629881,0.001567344,0.7750977,0.0005221043,0.0007365044,0.0006426946,0.04127369,0.001926592,0.01524516],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007045868,"threshold_uncertainty_score":0.02357078,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03351766075588018,"score_gpt":0.2913383535571634,"score_spread":0.2578206928012832,"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."}}