{"id":"W4384499027","doi":"10.1097/icu.0000000000000986","title":"Toward safer ophthalmic artificial intelligence via distributed validation on real-world data","year":2023,"lang":"en","type":"review","venue":"Current Opinion in Ophthalmology","topic":"Retinal Imaging and Analysis","field":"Medicine","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Context (archaeology); Machine learning; SAFER; Computer science; Implementation; Process (computing); Medicine; Artificial intelligence; Pipeline (software); Data science; Risk analysis (engineering); Computer security; Software 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":[0.01411156,0.0009989409,0.002063608,0.00330649,0.0004249448,0.003189254,0.002579883,0.002578018,0.003202275],"category_scores_gemma":[0.03092628,0.0005083781,0.0018404,0.002061483,0.001871135,0.003477378,0.001922514,0.004230667,0.002332081],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001562631,"about_ca_system_score_gemma":0.005953673,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001672478,"about_ca_topic_score_gemma":0.001862731,"domain_scores_codex":[0.9948591,0.00212117,0.0006475199,0.0005634131,0.001653829,0.0001549207],"domain_scores_gemma":[0.9710972,0.02128643,0.001831051,0.001200114,0.004257738,0.0003273415],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00004374029,0.00003897977,0.0003820304,0.02339205,0.0003230564,0.00008911707,0.000107908,0.00161234,0.0005577402,0.01782545,0.01960193,0.9360256],"study_design_scores_gemma":[0.00004901231,0.000158286,0.001469885,0.04761029,0.000614293,0.0009916824,0.0001521602,0.002449207,0.00168497,0.03215197,0.9125769,0.00009133707],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"review","genre_gemma":"review","genre_scores_codex":[0.0001808658,0.9869995,0.007040771,0.00324647,0.0005773074,0.00005675551,0.00005255338,0.00008423039,0.001761606],"genre_scores_gemma":[0.003746583,0.982349,0.009897127,0.002309782,0.0008330112,0.0001109017,0.0001492278,0.00004180014,0.0005625339],"genre_candidate":"review","genre_consensus":"review","teacher_disagreement_score":0.01411156,"threshold_uncertainty_score":0.07462996,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.5447142235765434,"score_gpt":0.5332955114215067,"score_spread":0.01141871215503676,"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."}}