{"id":"W4408521274","doi":"10.1109/bhi62660.2024.10913641","title":"Improving Interpretability of Radiology Report-based Pediatric Brain Tumor Pathology Classification and Key-phrases Extraction Using Large Language Models","year":2024,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"SickKids Foundation; Hospital for Sick Children; University of Toronto","funders":"","keywords":"Interpretability; Computer science; Key (lock); Natural language processing; Artificial intelligence; Information extraction; Feature extraction; Pathology; 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":[0.007730014,0.001843609,0.0007051791,0.005804827,0.0004103526,0.002398534,0.001036522,0.001007973,0.001243256],"category_scores_gemma":[0.03082789,0.0003377625,0.001510381,0.002074961,0.0004346788,0.002392385,0.001593226,0.001321112,0.001311726],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006925783,"about_ca_system_score_gemma":0.001297861,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00388644,"about_ca_topic_score_gemma":0.003866178,"domain_scores_codex":[0.9951497,0.002152264,0.0007264983,0.0009447318,0.0008078361,0.0002190374],"domain_scores_gemma":[0.9732851,0.02017392,0.002437216,0.001619314,0.002192445,0.0002920976],"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":[0.001408537,0.0007017622,0.2260917,0.001206429,0.0009098052,0.002127449,0.001824209,0.06763332,0.03806437,0.002118296,0.0194843,0.6384298],"study_design_scores_gemma":[0.00009829872,0.0003550712,0.05557423,0.0002202977,0.0005401436,0.001402562,0.0009152073,0.899341,0.02497278,0.005941191,0.01052386,0.000115304],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5733164,0.003381876,0.395661,0.00221703,0.0003853789,0.0005827553,0.01037052,0.01061235,0.003472604],"genre_scores_gemma":[0.8685199,0.0007518904,0.1145559,0.0002146393,0.0002583654,0.0002813569,0.01425986,0.0003312028,0.0008268913],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.007730014,"threshold_uncertainty_score":0.04088068,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02897689657711457,"score_gpt":0.3028116212115767,"score_spread":0.2738347246344622,"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."}}