{"id":"W7116906457","doi":"10.1002/alz70861_108662","title":"Identifying Site‐Level Data Integrity Risks in Alzheimer’s Trials via Subject‐Level Heuristics and a Novel Machine Learning Framework","year":2025,"lang":"en","type":"article","venue":"Alzheimer s & Dementia","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Queen's University","funders":"","keywords":"Data integrity; Heuristics; Trustworthiness; Scalability; Clinical trial; Personal Integrity; Risk assessment; Component (thermodynamics)","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch","research_integrity"],"consensus_categories":[],"category_scores_codex":[0.02621337,0.0008734897,0.001948932,0.00342913,0.0006792794,0.002215415,0.002369483,0.001468572,0.001226504],"category_scores_gemma":[0.03959933,0.000508423,0.001631298,0.002057881,0.002106564,0.00208178,0.001677905,0.002124867,0.0002682014],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001946552,"about_ca_system_score_gemma":0.00409836,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004999228,"about_ca_topic_score_gemma":0.005712214,"domain_scores_codex":[0.9888375,0.007943985,0.0005200329,0.001738564,0.0006184354,0.0003415374],"domain_scores_gemma":[0.9521385,0.03769413,0.004781009,0.002585199,0.002061234,0.0007399568],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001406773,0.0009022763,0.08767747,0.000611292,0.0007747385,0.0004593259,0.0009629274,0.5680076,0.002635504,0.02398773,0.004920644,0.3076537],"study_design_scores_gemma":[0.00009011423,0.0002961695,0.005112614,0.00004245749,0.00008339054,0.0000664907,0.00007262065,0.9679436,0.0007502788,0.02461239,0.0009005028,0.00002939161],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1312971,0.0007192218,0.8627532,0.001311809,0.00006161615,0.0005804186,0.0006418658,0.001344059,0.001290605],"genre_scores_gemma":[0.7440486,0.0001250753,0.2531241,0.0004952099,0.00009980863,0.0006017337,0.0008838824,0.00006438336,0.0005572679],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9985314,"threshold_uncertainty_score":0.1386312,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3393788561223451,"score_gpt":0.4323196236877823,"score_spread":0.09294076756543718,"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."}}