{"id":"W3195814479","doi":"10.1371/journal.pone.0255809","title":"Colonoscopy Polyp Detection and Classification: Dataset Creation and Comparative Evaluations","year":2021,"lang":"en","type":"article","venue":"arXiv (Cornell University)","topic":"Colorectal Cancer Screening and Detection","field":"Medicine","cited_by":124,"is_retracted":false,"has_abstract":true,"ca_institutions":"Toronto Metropolitan University","funders":"National Cancer Institute","keywords":"Colonoscopy; Benchmark (surveying); Computer science; Artificial intelligence; Ground truth; Colorectal cancer; Deep learning; Object detection; Machine learning; Cancer; Pattern recognition (psychology); Medicine; Internal medicine","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00007980824,0.00006978887,0.0001121532,0.00007419547,0.0001711792,0.00002201621,0.00001818908,0.00005434216,0.00004416141],"category_scores_gemma":[0.00004182137,0.00008034841,0.00001754446,0.0003407947,0.00008221874,0.0001479906,0.00002707544,0.00009212414,0.00001140385],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00008484011,"about_ca_system_score_gemma":0.00006672057,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00008441967,"about_ca_topic_score_gemma":0.000458307,"domain_scores_codex":[0.9994575,0.00006475431,0.00007144336,0.0002875498,0.00004294219,0.00007579259],"domain_scores_gemma":[0.9995267,0.00006431148,0.00004781116,0.0001497161,0.0001275301,0.00008393926],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.04996036,0.001870582,0.3771098,0.0007181929,0.001823416,0.0008152582,0.006194265,0.007640607,0.3624789,0.08456051,0.008874306,0.09795385],"study_design_scores_gemma":[0.003197754,0.001945168,0.6378835,0.00006684134,0.0005818005,0.0002030734,0.002342356,0.3230221,0.02193401,0.0008888256,0.007656837,0.0002777323],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9823493,0.000140245,0.01514292,0.0001653699,0.00005778075,0.0001478021,0.00004503194,0.00004215486,0.001909369],"genre_scores_gemma":[0.9988685,0.000112377,0.000152413,0.00006403631,0.00003552949,0.00000141625,0.0001865686,0.000003357888,0.000575761],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3405448,"threshold_uncertainty_score":0.3276513,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1749586662926574,"score_gpt":0.2827564298839285,"score_spread":0.1077977635912712,"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."}}