{"id":"W4407147701","doi":"10.1007/978-3-031-81101-2_3","title":"Bayesian Uncertainty Estimation Improves nnU-Net Generalization to Unseen Sites for Stroke Lesion Segmentation","year":2025,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":false,"ca_institutions":"Vancouver General Hospital; University of British Columbia","funders":"","keywords":"Generalization; Computer science; Segmentation; Bayesian probability; Artificial intelligence; Estimation; Pattern recognition (psychology); Machine learning; Mathematics; Management","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":[0.000885203,0.000493578,0.0004253719,0.001172162,0.0004658938,0.0008704846,0.001989656,0.0002898251,0.000008533721],"category_scores_gemma":[0.0003088171,0.0004916036,0.0001226556,0.0008837515,0.0001933056,0.001018259,0.0006603504,0.0002819365,0.00002150164],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006546542,"about_ca_system_score_gemma":0.0005641937,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001136972,"about_ca_topic_score_gemma":0.0004785515,"domain_scores_codex":[0.9963002,0.00005006095,0.0006814527,0.001556914,0.0007771963,0.0006341987],"domain_scores_gemma":[0.9973738,0.0005818913,0.0003176636,0.0009907747,0.0005657127,0.0001701348],"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":[0.00001095997,0.00001515067,0.00001162154,0.00005111014,0.000006182309,0.00000321149,0.0006055608,0.4259495,0.005669056,0.01877879,0.00008363902,0.5488153],"study_design_scores_gemma":[0.0001105215,0.0002489844,0.00002058211,0.0002146925,0.00001095383,0.000003191565,0.000001052145,0.8573618,0.04031307,0.1008488,0.0004328008,0.0004335031],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0002013489,0.00008723555,0.9947363,0.001567785,0.001212899,0.001509404,0.00002381476,0.0001918902,0.0004693903],"genre_scores_gemma":[0.05028764,0.00002875045,0.9455119,0.002325492,0.0003227367,0.00008935834,0.000102067,0.00003293859,0.001299123],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.5483817,"threshold_uncertainty_score":0.9997535,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02351366951640395,"score_gpt":0.2948137183357528,"score_spread":0.2713000488193488,"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."}}