{"id":"W4402140890","doi":"10.1101/2024.08.31.24312184","title":"Model uncertainty estimates for deep learning mammographic density prediction using ordinal and classification approaches","year":2024,"lang":"en","type":"preprint","venue":"medRxiv","topic":"Digital Radiography and Breast Imaging","field":"Medicine","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Sunnybrook Hospital; University of Toronto; Sunnybrook Health Science Centre","funders":"","keywords":"Ordinal data; Artificial intelligence; Ordinal regression; MAMMOGRAPHIC DENSITY; Computer science; Ordinal optimization; Machine learning; Econometrics; Pattern recognition (psychology); Statistics; Data mining; Mammography; Mathematics; 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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0004833734,0.0003194875,0.0004247534,0.0004779802,0.0001610422,0.0001892737,0.00008357214,0.0002402727,0.000001398789],"category_scores_gemma":[0.0001278346,0.0002884815,0.0002748759,0.0002704512,0.0001926111,0.0001009332,0.0002041178,0.000754016,0.000001192558],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007372966,"about_ca_system_score_gemma":0.0001010666,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002078888,"about_ca_topic_score_gemma":0.000005805316,"domain_scores_codex":[0.9983853,0.00003010736,0.000331587,0.0007203349,0.0002428643,0.0002898713],"domain_scores_gemma":[0.999234,0.00007790972,0.0001458778,0.0002628801,0.0001293173,0.0001499715],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000490555,0.0002109406,0.8249519,0.006642527,0.0008663235,0.00002613469,0.0006879736,0.07312664,0.005428453,0.002496365,0.00004301028,0.0850292],"study_design_scores_gemma":[0.0002960188,0.00006614258,0.05512431,0.000674336,0.0009759277,0.0001648907,0.0001492154,0.9265519,0.0001129135,0.01564959,0.00002675389,0.0002080126],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9090489,0.001630456,0.08717258,0.0003405761,0.0002244404,0.0006916164,0.00004202402,0.0002918447,0.0005575729],"genre_scores_gemma":[0.9847788,0.00006737284,0.01443284,0.00002386436,0.0001860267,0.0001017601,0.0002897376,0.00005581533,0.00006374809],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8534253,"threshold_uncertainty_score":0.9999567,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09068006128547482,"score_gpt":0.2926796802717823,"score_spread":0.2019996189863074,"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."}}