{"id":"W4386461358","doi":"10.36227/techrxiv.24078639.v1","title":"Predicting Progression from Mild Cognitive Impairment to Alzheimer's Dementia with Adversarial Attacks","year":2023,"lang":"en","type":"preprint","venue":"","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"National Institute on Aging; National Institute of Biomedical Imaging and Bioengineering; Canadian Institutes of Health Research; Genentech; IXICO; H. Lundbeck A/S; Servier; National Institutes of Health; F. Hoffmann-La Roche; Türkiye Bilimsel ve Teknolojik Araştırma Kurumu; Eisai; Northern California Institute for Research and Education; Pfizer; Biogen; BioClinica; Eli Lilly and Company; U.S. Department of Defense; Alzheimer's Disease Neuroimaging Initiative; Meso Scale Diagnostics; University of Southern California; Novartis Pharmaceuticals Corporation; Bristol-Myers Squibb; Alzheimer's Drug Discovery Foundation; Foundation for the National Institutes of Health","keywords":"Dementia; Adversarial system; Subtyping; Artificial intelligence; Computer science; Machine learning; Binary classification; Cognitive impairment; Deep learning; Decision boundary; Disease; Artificial neural network; Alzheimer's disease; Cognition; Set (abstract data type); Psychology; Cognitive psychology; Psychiatry; Medicine; Pathology","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0006176422,0.0005215039,0.0004785391,0.0002503383,0.0002955268,0.0003949874,0.001546115,0.0003323319,0.00009669326],"category_scores_gemma":[0.0001102139,0.0004247895,0.0001217461,0.0003871723,0.00004209497,0.0001833948,0.005667042,0.001301178,0.0002504813],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001043133,"about_ca_system_score_gemma":0.0005087568,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006572271,"about_ca_topic_score_gemma":0.0008684379,"domain_scores_codex":[0.9954536,0.0003612374,0.0005753955,0.001810585,0.001108287,0.0006908788],"domain_scores_gemma":[0.9973678,0.0004051893,0.0004176856,0.001064073,0.0003228869,0.0004223557],"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":[0.0006486446,0.0003646468,0.8736574,0.0003638816,0.005384924,0.0003530021,0.0149989,0.02111837,0.000008251167,0.0004690526,0.007279691,0.07535323],"study_design_scores_gemma":[0.00360925,0.002881096,0.4272791,0.009332854,0.001330348,0.00001441231,0.0009213773,0.5484105,0.001051783,0.001712417,0.0008794001,0.002577531],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2512691,0.001287977,0.7100817,0.01325325,0.008020105,0.00826689,0.0002886752,0.005498084,0.002034184],"genre_scores_gemma":[0.8074695,0.00001113595,0.190013,0.0006624318,0.0008007204,0.0006087649,0.0002748274,0.00008145052,0.00007812782],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.5562004,"threshold_uncertainty_score":0.9998204,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05297670813844756,"score_gpt":0.3549371046384401,"score_spread":0.3019603964999926,"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."}}