{"id":"W4386781106","doi":"10.1016/j.cgh.2023.08.033","title":"Accuracy, Reliability, and Comprehensibility of ChatGPT-Generated Medical Responses for Patients With Nonalcoholic Fatty Liver Disease","year":2023,"lang":"en","type":"article","venue":"Clinical Gastroenterology and Hepatology","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":84,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University Health Centre","funders":"Universitätsmedizin der Johannes Gutenberg-Universität Mainz; Humanitas Research Hospital; Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas; Ipsen; Assistance publique-Hôpitaux de Paris; Università Cattolica del Sacro Cuore; Shionogi; Chinese University of Hong Kong; Siemens Healthineers; Fonds de Recherche du Québec - Santé; Intercept Pharmaceuticals; Institut National de la Santé et de la Recherche Médicale; Università degli Studi di Palermo; Imperial College Healthcare NHS Trust; Imperial College London; Swedish Orphan Biovitrum; Novo Nordisk; Humanitas University; McGill University; McGill University Health Centre; Ionis Pharmaceuticals; Julius Clinical; Inova Health System; Pfizer; Inventiva Pharma; Gilead Sciences; Universidad de Sevilla; Università degli Studi di Milano; Sanofi; Allergan; AstraZeneca","keywords":"Medicine; Nonalcoholic fatty liver disease; Disease; Reliability (semiconductor); Fatty liver; MEDLINE; Internal medicine; Gastroenterology; Biochemistry","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"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":[0.01379116,0.0006303678,0.0006473825,0.001612036,0.0004114214,0.001428552,0.0006909812,0.001250617,0.004876623],"category_scores_gemma":[0.1407381,0.0002346418,0.001209784,0.000665265,0.0006973137,0.001633073,0.002138953,0.001283726,0.002320084],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004829471,"about_ca_system_score_gemma":0.0003795012,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001120384,"about_ca_topic_score_gemma":0.001171946,"domain_scores_codex":[0.9898393,0.005760008,0.001282447,0.001337592,0.001503974,0.0002766452],"domain_scores_gemma":[0.8618733,0.1147589,0.009248031,0.005147587,0.007498911,0.00147325],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.004185961,0.000652514,0.8462433,0.00112269,0.0006760193,0.0003049939,0.0110635,0.001697394,0.001520157,0.0004645493,0.010894,0.1211749],"study_design_scores_gemma":[0.0002826719,0.0015131,0.9574519,0.0009922568,0.0004623843,0.0008688709,0.004805757,0.02053134,0.002463204,0.0021683,0.008253691,0.0002064452],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9883726,0.0007938354,0.003128552,0.0006055664,0.0002642229,0.0002357333,0.002640505,0.0003243544,0.003634511],"genre_scores_gemma":[0.9940132,0.0002607507,0.002151675,0.0002982763,0.00009323988,0.0003178919,0.002075729,0.00005163934,0.000737571],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01379116,"threshold_uncertainty_score":0.07293546,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1581993732780403,"score_gpt":0.4515020842461751,"score_spread":0.2933027109681348,"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."}}