{"id":"W4377002285","doi":"10.59275/j.melba.2022-db5c","title":"Label fusion and training methods for reliable representation of inter-rater uncertainty","year":2023,"lang":"en","type":"article","venue":"The Journal of Machine Learning for Biomedical Imaging","topic":"Medical Imaging and Analysis","field":"Engineering","cited_by":13,"is_retracted":false,"has_abstract":true,"ca_institutions":"Polytechnique Montréal; Mila - Quebec Artificial Intelligence Institute","funders":"Fonds de recherche du Québec – Nature et technologies; Canadian Institutes of Health Research; Natural Sciences and Engineering Research Council of Canada; Institut de Valorisation des Données; Craig H. Neilsen Foundation; Canada First Research Excellence Fund; Nvidia","keywords":"Segmentation; Artificial intelligence; Computer science; Ground truth; Machine learning; Task (project management); Pattern recognition (psychology)","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":[0.003733827,0.0001050755,0.0003195213,0.0002600659,0.000117287,0.00002448195,0.0001596145,0.00003464454,0.0000167951],"category_scores_gemma":[0.001579684,0.00006696787,0.0001104822,0.0003946293,0.0001361006,0.00009385141,0.00004485304,0.0003392487,6.719109e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000188618,"about_ca_system_score_gemma":0.00001766976,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003100351,"about_ca_topic_score_gemma":4.233245e-7,"domain_scores_codex":[0.998875,0.0001403295,0.0004932976,0.00008814868,0.0001789569,0.0002242326],"domain_scores_gemma":[0.9982421,0.001266203,0.0001891448,0.00008786358,0.0001092904,0.0001053818],"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":[0.00007750349,0.00001804577,0.0009422691,0.0002512965,0.0001751101,0.000003269419,0.002462288,0.01254739,0.08038973,0.0000226641,0.002779682,0.9003307],"study_design_scores_gemma":[0.0008884206,0.00008500248,0.00009494575,0.0001907426,0.0001571896,0.00003450492,0.0008126508,0.9812407,0.001258607,0.001050964,0.01411137,0.00007483627],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.07213165,0.001024597,0.9223544,0.003872256,0.0003617851,0.0001115405,0.000005345275,0.00007986374,0.00005855196],"genre_scores_gemma":[0.9303108,0.000583954,0.0682943,0.0001287708,0.0003241553,0.000008154031,0.00004482636,0.00004909734,0.0002559373],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9686934,"threshold_uncertainty_score":0.2730871,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03216285539838912,"score_gpt":0.365834025024052,"score_spread":0.3336711696256628,"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."}}