{"id":"W4385707210","doi":"10.1016/j.jcjo.2023.07.016","title":"Artificial intelligence chatbot performance in triage of ophthalmic conditions","year":2023,"lang":"en","type":"article","venue":"Canadian Journal of Ophthalmology","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":91,"is_retracted":false,"has_abstract":false,"ca_institutions":"","funders":"Foundation Fighting Blindness; Research to Prevent Blindness","keywords":"Triage; Medical diagnosis; Medicine; Medical emergency; Pathology","routes":{"ca_aff":false,"ca_fund":true,"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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.000719856,0.00009237153,0.0003419195,0.001121574,0.00007136896,0.000006941819,0.000148793,0.0001401038,0.001295977],"category_scores_gemma":[0.0006252073,0.00009120069,0.00008612493,0.0008883242,0.0002090984,0.0001024086,0.000006871141,0.0003256469,0.0001124697],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001653923,"about_ca_system_score_gemma":0.002168098,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002943431,"about_ca_topic_score_gemma":0.000338863,"domain_scores_codex":[0.9984205,0.00007239736,0.0009106867,0.0001175888,0.0001237238,0.0003551021],"domain_scores_gemma":[0.9987156,0.0002035523,0.0002484081,0.0001615236,0.0002990117,0.0003719318],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0004656531,0.0002287719,0.9280094,0.0003734579,0.00008647676,0.02538686,0.006865428,0.002089011,0.002488175,0.004225304,0.0006843046,0.02909715],"study_design_scores_gemma":[0.0001472847,0.004037555,0.7532685,0.0008992294,0.00008456372,0.1537703,0.01046436,0.002175311,0.04860639,0.02579671,0.0003730308,0.0003767193],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9937554,0.0001742027,0.00001567392,0.001586375,0.001115238,0.0001681801,0.00001049142,0.000003395855,0.003171015],"genre_scores_gemma":[0.9993687,0.00003203802,0.00009012923,0.00002989639,0.0002261951,0.000006531715,0.00001389892,0.00001215533,0.0002204927],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1747409,"threshold_uncertainty_score":0.999617,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2504959227755974,"score_gpt":0.4359478780541811,"score_spread":0.1854519552785837,"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."}}