{"id":"W4414702782","doi":"10.1038/s41746-025-02004-3","title":"Large language models forecast patient health trajectories enabling digital twins","year":2025,"lang":"en","type":"article","venue":"npj Digital Medicine","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":24,"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; Biogen; F. Hoffmann-La Roche; Roche; Eli Lilly and Company; BioClinica; European Commission; Alzheimer's Disease Neuroimaging Initiative; Meso Scale Diagnostics; National Institute on Aging; Alzheimer's Association; Foundation for the National Institutes of Health","keywords":"Benchmarking; Health care; Health records; Normalization (sociology); Imputation (statistics); Generative grammar; Database normalization; Missing data","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002258593,0.0005771597,0.00053012,0.0008074493,0.0003239522,0.001374401,0.001156998,0.0008367989,0.003067398],"category_scores_gemma":[0.0125152,0.0004814356,0.0008990788,0.0008193046,0.0005822864,0.001948623,0.002172663,0.001997431,0.001029029],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009173396,"about_ca_system_score_gemma":0.001484322,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005789552,"about_ca_topic_score_gemma":0.008216448,"domain_scores_codex":[0.9993118,0.0003182279,0.00004556028,0.0001655018,0.0001063555,0.00005257291],"domain_scores_gemma":[0.9969057,0.002174283,0.0001880726,0.0003800528,0.0002241226,0.0001276975],"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":[0.0004066565,0.0001376353,0.01328335,0.0001231284,0.0001920198,0.0004046407,0.0004558708,0.6270123,0.002523181,0.03460653,0.008810475,0.3120442],"study_design_scores_gemma":[0.00001251091,0.00002598342,0.0003380155,0.00001301859,0.00001680125,0.00005905027,0.00002746742,0.9711005,0.0009019458,0.02582418,0.001669384,0.00001117522],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04569129,0.0004716805,0.944964,0.002077076,0.0001593247,0.00007061285,0.001361773,0.002902891,0.002301373],"genre_scores_gemma":[0.7479178,0.0005617685,0.2428907,0.001086353,0.0001645459,0.0001663009,0.002862394,0.0003366886,0.004013514],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005789552,"threshold_uncertainty_score":0.01194477,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09066821201173551,"score_gpt":0.3957070671382421,"score_spread":0.3050388551265066,"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."}}