{"id":"W4302757635","doi":"10.21203/rs.3.rs-2093909/v1","title":"Semi-supervised learning to automate tumor bud detection in cytokeratin-stained whole-slide images of colorectal cancer","year":2022,"lang":"en","type":"preprint","venue":"Research Square","topic":"AI in cancer detection","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"Swiss Cancer Research Foundation; KWF Kankerbestrijding; European Commission","keywords":"Tumor budding; Artificial intelligence; Colorectal cancer; Computer science; Convolutional neural network; Machine learning; Cytokeratin; Deep learning; Pathology; Cancer; Medicine; Immunohistochemistry; Metastasis; 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":[0.001209455,0.0005431551,0.0005441969,0.000865209,0.0001782577,0.0005661371,0.0009563367,0.0006621924,0.0007121816],"category_scores_gemma":[0.002253291,0.0002654258,0.0005259209,0.0003916204,0.0003445839,0.0003803814,0.0005301784,0.0005226292,0.0006092582],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004950397,"about_ca_system_score_gemma":0.0005912196,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002587138,"about_ca_topic_score_gemma":0.00463821,"domain_scores_codex":[0.9994266,0.0001467521,0.00003843663,0.0001880116,0.0001275867,0.0000725604],"domain_scores_gemma":[0.998468,0.0006241795,0.0002040024,0.0002578238,0.0003590227,0.00008698133],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00109556,0.00055985,0.0188711,0.0005476028,0.0002214483,0.0004559663,0.000253397,0.3114553,0.1675788,0.001119233,0.007912163,0.4899297],"study_design_scores_gemma":[0.00000886136,0.00003599809,0.003392244,0.000007146998,0.000008037265,0.0000697603,0.00001628031,0.9801493,0.01547477,0.0004690881,0.0003614293,0.00000698532],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.4991862,0.0006096741,0.4905419,0.0002989155,0.00008994911,0.0001672381,0.0008666666,0.006868686,0.001370679],"genre_scores_gemma":[0.8447938,0.00013158,0.1514219,0.00008665511,0.00003028987,0.00008233986,0.001647734,0.000161844,0.001643702],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.002587138,"threshold_uncertainty_score":0.006396294,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03271908020060321,"score_gpt":0.3659890491056398,"score_spread":0.3332699689050366,"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."}}