{"id":"W4416868033","doi":"10.1101/2025.11.26.25341040","title":"Temporal deep learning with clinically engineered biomarkers for the early prediction of type 2 diabetes","year":2025,"lang":"","type":"preprint","venue":"medRxiv","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Toronto Metropolitan University; Institute of Health Services and Policy Research","funders":"","keywords":"Deep learning; Type 2 diabetes; Prediabetes; Glycated hemoglobin; Convolutional neural network; Artificial neural network; Support vector machine; Feature engineering","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001050971,0.0007594567,0.0004471694,0.0005527481,0.00018674,0.000505886,0.0005440657,0.0004278992,0.0009833904],"category_scores_gemma":[0.002363732,0.0001995647,0.0005912808,0.0004656523,0.0001558174,0.0005679906,0.0006439302,0.001069696,0.0002725082],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006881049,"about_ca_system_score_gemma":0.0009241504,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01267789,"about_ca_topic_score_gemma":0.0139284,"domain_scores_codex":[0.9997073,0.0001039717,0.00002065465,0.00007431601,0.00004984699,0.00004391923],"domain_scores_gemma":[0.9993469,0.0003294791,0.00008707317,0.00005609485,0.0001362427,0.00004417706],"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":[0.001234051,0.0009976294,0.06761672,0.0001532334,0.0004767278,0.0002722207,0.00008724786,0.4801964,0.008322022,0.002926855,0.008856871,0.42886],"study_design_scores_gemma":[0.000007745461,0.00004097001,0.001329468,0.00001065028,0.0000231026,0.00001285776,0.000006926162,0.9966131,0.0007402985,0.0009800723,0.0002308753,0.000003977474],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.621991,0.004205423,0.3615308,0.002409795,0.0002975856,0.0001063227,0.002167054,0.002157779,0.005134213],"genre_scores_gemma":[0.9712098,0.0004166476,0.02534536,0.0002573124,0.00006999684,0.0000384107,0.001117336,0.00002171317,0.001523356],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01267789,"threshold_uncertainty_score":0.02520818,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0264367787554617,"score_gpt":0.2950899944634397,"score_spread":0.268653215707978,"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."}}