{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.004010581,0.002196656,0.00122705,0.004397784,0.001077167,0.001528258,0.002657196,0.002358151,0.002751966],"category_scores_gemma":[0.008902098,0.0003472909,0.001758915,0.003357508,0.001045253,0.001296614,0.002205725,0.001395767,0.002921675],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001933492,"about_ca_system_score_gemma":0.001777045,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02374699,"about_ca_topic_score_gemma":0.03622068,"domain_scores_codex":[0.9957758,0.0008571385,0.0005593158,0.0009092127,0.00151849,0.0003800344],"domain_scores_gemma":[0.994661,0.001249554,0.0004743365,0.00140002,0.00159129,0.0006238566],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.005990904,0.006975633,0.07393286,0.006450661,0.001841168,0.001944083,0.0004524214,0.02474833,0.02855426,0.002080819,0.4250802,0.4219487],"study_design_scores_gemma":[0.003945872,0.006005577,0.362311,0.001503543,0.001880939,0.01158863,0.002472789,0.1998327,0.06559453,0.004154926,0.3398255,0.0008838915],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5478432,0.01092348,0.02838986,0.003760722,0.002213156,0.004614234,0.3729352,0.01362791,0.01569229],"genre_scores_gemma":[0.1894099,0.002173094,0.04416424,0.000854703,0.0003496375,0.001329455,0.7567821,0.0004512052,0.004485635],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02374699,"threshold_uncertainty_score":0.04721755,"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."}}