{"id":"W4394824270","doi":"10.1016/j.jbi.2024.104638","title":"MixEHR-SurG: A joint proportional hazard and guided topic model for inferring mortality-associated topics from electronic health records","year":2024,"lang":"en","type":"article","venue":"Journal of Biomedical Informatics","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University Health Centre; McGill University; Mila - Quebec Artificial Intelligence Institute","funders":"Fonds de recherche du Québec – Nature et technologies; Canadian Institutes of Health Research; Canada Research Chairs; Heart And Stroke Foundation Of Quebec; Natural Sciences and Engineering Research Council of Canada; Heart and Stroke Foundation of Canada","keywords":"Medicine; Inference; Proportional hazards model; Health records; Hazard ratio; Medical record; Scale (ratio); Artificial intelligence; Survival analysis; Machine learning; Topic model; Computer science; Data mining; Internal medicine; Health care; Cartography","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001956712,0.0001386523,0.0003807601,0.0002258562,0.0001172509,0.0002066533,0.0003477791,0.0001322678,0.000007904561],"category_scores_gemma":[0.0004635621,0.0001087392,0.0001303508,0.0002643456,0.00006370798,0.0005284387,0.0001398284,0.0006727609,0.000001697437],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004681054,"about_ca_system_score_gemma":0.001743947,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00009964342,"about_ca_topic_score_gemma":0.0000374274,"domain_scores_codex":[0.9971926,0.00006794652,0.001557648,0.0001137217,0.0006755403,0.0003925034],"domain_scores_gemma":[0.9985202,0.0001508833,0.0006254226,0.0001854528,0.0002190158,0.000299022],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00004037088,0.0004161167,0.01475621,0.004504011,0.001129544,0.00007962841,0.03406812,0.004259998,0.00005361962,0.1220565,0.04743063,0.7712052],"study_design_scores_gemma":[0.0003499001,0.0003610578,0.002169316,0.0003449236,0.00001558314,0.00006180898,0.00005181852,0.9628241,0.000006593769,0.02313332,0.01057468,0.0001069061],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1639808,0.000597713,0.8175161,0.01682314,0.0007732961,0.000198207,0.00001806258,0.00006322702,0.00002937845],"genre_scores_gemma":[0.7847522,0.0004816268,0.2105658,0.00311151,0.0008104198,0.00001490951,0.00005066379,0.00002099122,0.0001918896],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9585641,"threshold_uncertainty_score":0.4434255,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05841647606246331,"score_gpt":0.3577413703770494,"score_spread":0.2993248943145861,"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."}}