{"id":"W4323030608","doi":"10.1016/j.jfop.2023.100005","title":"Conversational AI Models for ophthalmic diagnosis: Comparison of ChatGPT and the Isabel Pro Differential Diagnosis Generator","year":2023,"lang":"en","type":"article","venue":"JFO Open Ophthalmology","topic":"Clinical Reasoning and Diagnostic Skills","field":"Medicine","cited_by":116,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta; University of Toronto","funders":"","keywords":"Medical diagnosis; Differential diagnosis; Generator (circuit theory); Medicine; Set (abstract data type); Differential (mechanical device); Computer science; Artificial intelligence; Medical physics; Radiology; Pathology; Engineering","routes":{"ca_aff":true,"ca_fund":false,"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":[0.01084097,0.001140457,0.0006541656,0.002214192,0.0005110801,0.002352323,0.001773003,0.001114898,0.004264598],"category_scores_gemma":[0.06243784,0.0003834135,0.0009388895,0.0008540188,0.0006563763,0.00252314,0.002405724,0.001649694,0.001428189],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002103682,"about_ca_system_score_gemma":0.002364135,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006823334,"about_ca_topic_score_gemma":0.006323814,"domain_scores_codex":[0.9920585,0.005580413,0.0005165605,0.0006796228,0.001016628,0.0001483348],"domain_scores_gemma":[0.9177024,0.07282837,0.00199242,0.002534569,0.003803627,0.001138633],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.009245403,0.001590673,0.05934625,0.002378838,0.0009460563,0.001374152,0.004997965,0.1703987,0.007398972,0.01122189,0.02019052,0.7109106],"study_design_scores_gemma":[0.0003107022,0.00129262,0.006796034,0.0003523813,0.0002877982,0.0009819199,0.001422311,0.964022,0.004661228,0.009385071,0.01034822,0.0001397632],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5577056,0.003099552,0.3886977,0.005649608,0.0007764919,0.003156667,0.005115959,0.01585103,0.01994728],"genre_scores_gemma":[0.7689989,0.0009188429,0.2204299,0.0009644605,0.0001711982,0.001440367,0.004171459,0.0003498904,0.002554998],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01084097,"threshold_uncertainty_score":0.05733317,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1180761907008215,"score_gpt":0.4254146746001916,"score_spread":0.3073384838993701,"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."}}