{"id":"W4409258708","doi":"10.1016/j.jbi.2025.104795","title":"Low-cost algorithms for clinical notes phenotype classification to enhance epidemiological surveillance: A case study","year":2025,"lang":"en","type":"article","venue":"Journal of Biomedical Informatics","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":false,"ca_institutions":"","funders":"Fondo para la Investigación Científica y Tecnológica; Agencia Nacional de Promoción de la Investigación, el Desarrollo Tecnológico y la Innovación; Styrelsen för Internationellt Utvecklingssamarbete; Universidad Nacional de Quilmes; Consejo Nacional de Investigaciones Científicas y Técnicas; International Development Research Centre","keywords":"Computer science; Epidemiology; Phenotype; Algorithm; Medicine; Data mining; Artificial intelligence; Pathology; Biology; Genetics","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":[0.002860795,0.0005388208,0.0005407842,0.001225134,0.0004612555,0.001849801,0.001209143,0.001548039,0.001442896],"category_scores_gemma":[0.01490096,0.0002280187,0.000480019,0.000978935,0.0002451359,0.0009014798,0.0006864229,0.0009020297,0.0007292514],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005784442,"about_ca_system_score_gemma":0.001035226,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003198189,"about_ca_topic_score_gemma":0.003869352,"domain_scores_codex":[0.9982911,0.0007294975,0.0001874879,0.0002061097,0.0004778644,0.0001080444],"domain_scores_gemma":[0.9913467,0.005710545,0.0003252063,0.0006321675,0.001800059,0.0001853514],"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":[0.001568676,0.002282185,0.1008884,0.0002465574,0.0001501624,0.002741558,0.0003151013,0.04076132,0.01530006,0.002949072,0.007603754,0.8251932],"study_design_scores_gemma":[0.0002581926,0.0006800134,0.03034232,0.00007208094,0.0001927417,0.005944591,0.0004011957,0.9317824,0.01819934,0.005197084,0.006857878,0.00007210322],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5766855,0.001073328,0.4109911,0.004109018,0.0002252737,0.0006038233,0.001080587,0.001599522,0.003631761],"genre_scores_gemma":[0.6737697,0.0003062224,0.3229769,0.0003070437,0.00008149985,0.0001472742,0.000775949,0.00008087795,0.001554426],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.003198189,"threshold_uncertainty_score":0.01512951,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1203890973362984,"score_gpt":0.490809706567575,"score_spread":0.3704206092312766,"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."}}