{"id":"W4229451765","doi":"10.3390/jimaging8050131","title":"BI-RADS BERT and Using Section Segmentation to Understand Radiology Reports","year":2022,"lang":"en","type":"article","venue":"Journal of Imaging","topic":"Topic Modeling","field":"Computer Science","cited_by":42,"is_retracted":false,"has_abstract":true,"ca_institutions":"Sunnybrook Health Science Centre; University of Toronto","funders":"Simon Fraser University; Compute Canada; Canadian Institutes of Health Research; Sunnybrook Research Institute","keywords":"Computer science; Segmentation; Artificial intelligence; Lexicon; Sentence; Natural language processing; Breast imaging; Section (typography); Classifier (UML); Mammography; Breast cancer; Pattern recognition (psychology); Medicine; Cancer","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006561595,0.00004460917,0.00008958937,0.0001938487,0.0001650079,0.00006707761,0.00009727612,0.000006908654,0.000005027484],"category_scores_gemma":[0.00002358613,0.00004491069,0.0000238897,0.0001468846,0.000007833857,0.0003934186,0.0001244799,0.000133144,6.20094e-8],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002236216,"about_ca_system_score_gemma":0.00006073052,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002300598,"about_ca_topic_score_gemma":9.0601e-7,"domain_scores_codex":[0.9992678,0.00006820646,0.0002615367,0.0001185709,0.0001866728,0.00009719048],"domain_scores_gemma":[0.9995545,0.00002050326,0.000232241,0.00009833138,0.00004501425,0.00004946122],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00006158646,0.00008659344,0.1004307,0.00004308173,0.0001092267,0.002437511,0.03266846,0.4037128,0.2959317,0.001585498,0.002473727,0.1604591],"study_design_scores_gemma":[0.0008229173,0.0002435492,0.005013311,0.00006255107,0.00003922228,0.06360272,0.005725955,0.9107049,0.002801128,0.009223302,0.001464072,0.0002963729],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3863797,0.0001775117,0.6116014,0.0009783575,0.0007886394,0.00002701783,5.935279e-8,0.000006335179,0.00004100839],"genre_scores_gemma":[0.921766,0.000004059039,0.07767795,0.0004053984,0.0001320765,3.393289e-7,6.715622e-8,0.00000352461,0.00001063196],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.5353863,"threshold_uncertainty_score":0.1831405,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02553176897951776,"score_gpt":0.2778159636013181,"score_spread":0.2522841946218003,"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."}}