{"id":"W4413149186","doi":"10.1016/j.compmedimag.2025.102630","title":"Inference time correction based on confidence and uncertainty for improved deep-learning model performance and explainability in medical image classification","year":2025,"lang":"en","type":"article","venue":"Computerized Medical Imaging and Graphics","topic":"Radiomics and Machine Learning in Medical Imaging","field":"Medicine","cited_by":2,"is_retracted":false,"has_abstract":false,"ca_institutions":"ASTER","funders":"Qatar National Research Fund; Qatar Foundation","keywords":"Inference; Artificial intelligence; Computer science; Deep learning; Machine learning; Image (mathematics); Confidence interval; Medical imaging; Pattern recognition (psychology); Computer vision; Statistics; Mathematics","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":[],"consensus_categories":[],"category_scores_codex":[0.00181949,0.0002058793,0.0004259713,0.000278776,0.0002192748,0.00007296334,0.000106731,0.0001705086,0.000009398172],"category_scores_gemma":[0.004448576,0.0001792768,0.00004437551,0.0002516555,0.0007445078,0.0001037381,0.0000903635,0.001012104,2.528318e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004636151,"about_ca_system_score_gemma":0.0002937151,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004659212,"about_ca_topic_score_gemma":0.000004122169,"domain_scores_codex":[0.9982719,0.0001338389,0.0004061991,0.0005384872,0.0003479728,0.0003016448],"domain_scores_gemma":[0.9979594,0.001252109,0.0000898049,0.0001824896,0.0001393795,0.0003768546],"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":[0.0006875917,0.0002048114,0.03888505,0.0008267827,0.00002586177,0.00002089838,0.0002909755,0.0004045309,0.001550737,0.001472885,0.0002852938,0.9553446],"study_design_scores_gemma":[0.003546219,0.0001299933,0.02802318,0.00111048,0.00003520488,0.00003478277,0.00004345912,0.9656033,0.0000167339,0.0008293161,0.0004781883,0.0001491654],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3395839,0.0002125687,0.641314,0.01786306,0.0002319828,0.0004586252,0.000001418347,0.0001068114,0.0002276053],"genre_scores_gemma":[0.9884627,0.0008031987,0.005959882,0.004551958,0.00005528668,0.00005342884,0.00003521496,0.00001528348,0.00006306724],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9651988,"threshold_uncertainty_score":0.7310695,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.008764051783965858,"score_gpt":0.3013993181054609,"score_spread":0.2926352663214951,"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."}}