{"id":"W4407578265","doi":"10.1038/s41598-025-86110-y","title":"Reducing inference cost of Alzheimer’s disease identification using an uncertainty-aware ensemble of uni-modal and multi-modal learners","year":2025,"lang":"en","type":"article","venue":"Scientific Reports","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Canadian Institutes of Health Research; National Institutes of Health; Genentech; IXICO; H. Lundbeck A/S; Servier; Eisai; Northern California Institute for Research and Education; Pfizer; Novartis Pharmaceuticals Corporation; University of Southern California; Biogen; Science Foundation Ireland; Eli Lilly and Company; Bristol-Myers Squibb; BioClinica; U.S. Department of Defense; Alzheimer's Disease Neuroimaging Initiative; Meso Scale Diagnostics; National Institute on Aging; Alzheimer's Association","keywords":"Positron emission tomography; Computer science; Artificial intelligence; Inference; Deep learning; Magnetic resonance imaging; Neuroimaging; Identification (biology); Machine learning; Modal; Medicine; Nuclear medicine; Radiology","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002191292,0.0001741113,0.0002942328,0.0004764489,0.0003953033,0.000278357,0.0004685025,0.00007604893,0.000006495133],"category_scores_gemma":[0.0007256355,0.000177848,0.00006840831,0.001154199,0.0004087685,0.0006425667,0.0003521091,0.0002055642,7.979154e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007192974,"about_ca_system_score_gemma":0.001037078,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001551964,"about_ca_topic_score_gemma":0.00008136794,"domain_scores_codex":[0.9968271,0.0002639867,0.0009291081,0.001092485,0.0005658524,0.0003214221],"domain_scores_gemma":[0.9965815,0.00009612198,0.0008178905,0.001676997,0.0006029107,0.0002245944],"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":[0.00005366134,0.00064333,0.3196961,0.001148051,0.0001024211,0.0002194409,0.008103947,0.3650256,0.1715178,0.007103745,0.0002384065,0.1261476],"study_design_scores_gemma":[0.0001723407,0.00002867927,0.04086222,0.000334322,0.00004850689,0.00002369769,0.0002900461,0.9392938,0.01555579,0.003053944,0.000123992,0.0002126331],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8188196,0.0003385127,0.1775256,0.0001941638,0.002491859,0.0004902704,0.000005000531,0.00008093111,0.00005407119],"genre_scores_gemma":[0.9905772,0.000004491243,0.009114768,0.00001673417,0.0000146513,0.00001226107,0.00003281909,0.000009665321,0.0002174006],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5742682,"threshold_uncertainty_score":0.725243,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05866277394009835,"score_gpt":0.3703622250799805,"score_spread":0.3116994511398821,"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."}}