{"id":"W3092643853","doi":"10.2196/21252","title":"Patient Triage by Topic Modeling of Referral Letters: Feasibility Study","year":2020,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Healthcare Systems and Technology","field":"Business, Management and Accounting","cited_by":25,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Health and Care Research Wales","keywords":"Triage; Referral; Topic model; Latent Dirichlet allocation; Artificial intelligence; Context (archaeology); Computer science; Set (abstract data type); Relevance (law); Medicine; Machine learning; Family medicine; Medical emergency","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.000376319,0.000116085,0.0003197327,0.00006693094,0.00005148003,0.00003502002,0.0002419239,0.0001297273,0.0001098482],"category_scores_gemma":[0.0003045604,0.0000911943,0.00005298313,0.000269041,0.00003940436,0.000403612,0.0002110655,0.000291483,0.00005004988],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002315464,"about_ca_system_score_gemma":0.00003107825,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003220261,"about_ca_topic_score_gemma":0.00003375483,"domain_scores_codex":[0.9981297,0.0000158747,0.0009595741,0.0001020953,0.0005777919,0.0002150168],"domain_scores_gemma":[0.9993671,0.00001876471,0.0002618897,0.000221657,0.00008201344,0.00004860599],"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":[0.0006203252,0.004949375,0.2043126,0.01884458,0.000430726,0.0001299408,0.05687879,0.0003970145,0.0003334669,0.01614164,0.1788257,0.5181358],"study_design_scores_gemma":[0.006034904,0.0006959119,0.002004109,0.0003259915,0.00006068496,0.000004447817,0.04583326,0.8380924,0.00003163237,0.0004599038,0.1057475,0.0007093072],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9904638,0.00001656849,0.001865335,0.004603285,0.0001543833,0.0007680242,0.000002573484,0.0001275855,0.001998439],"genre_scores_gemma":[0.992224,0.000001466133,0.0000910192,0.007401036,0.0002266053,0.00003136603,0.00001420864,0.000007647762,0.000002679859],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8376954,"threshold_uncertainty_score":0.3718795,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06571533772948462,"score_gpt":0.310388216746979,"score_spread":0.2446728790174944,"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."}}