{"id":"W4287958519","doi":"10.1007/s11548-022-02707-y","title":"Training deep neural networks with noisy clinical labels: toward accurate detection of prostate cancer in US data","year":2022,"lang":"en","type":"article","venue":"International Journal of Computer Assisted Radiology and Surgery","topic":"AI in cancer detection","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":false,"ca_institutions":"Queen's University; University of British Columbia; University of British Columbia Hospital","funders":"","keywords":"Overfitting; Computer science; Artificial intelligence; Generalizability theory; Prostate cancer; Machine learning; Artificial neural network; Pattern recognition (psychology); Cancer; Medicine; Mathematics; Statistics","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":[0.002326238,0.001157089,0.0009325311,0.001023174,0.0003989335,0.0009989367,0.001570484,0.001960849,0.0008264718],"category_scores_gemma":[0.009713112,0.0006330051,0.0006261545,0.0008676474,0.0006886162,0.001098281,0.001321553,0.002202949,0.0007343961],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008158156,"about_ca_system_score_gemma":0.001413078,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01222708,"about_ca_topic_score_gemma":0.01541332,"domain_scores_codex":[0.9990933,0.0002931151,0.00005049718,0.0002573987,0.0001611702,0.0001445347],"domain_scores_gemma":[0.9966912,0.002076531,0.0002614678,0.00035686,0.0004923954,0.0001214602],"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":[0.001044522,0.0006895895,0.04210218,0.0002519006,0.0002763544,0.0003054187,0.0002190505,0.3469596,0.01384327,0.0022378,0.01449836,0.5775719],"study_design_scores_gemma":[0.000009992343,0.00003062139,0.001196845,0.0000147402,0.00001523197,0.00002499863,0.00001851475,0.9949369,0.001975712,0.001382157,0.0003883593,0.000005884669],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5387095,0.003727643,0.4452472,0.003404242,0.0003972703,0.0001107253,0.001625469,0.003917856,0.00286018],"genre_scores_gemma":[0.9350222,0.000453119,0.05869016,0.0006904596,0.0001801993,0.00006007677,0.002049167,0.0001077878,0.002746784],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01222708,"threshold_uncertainty_score":0.02431184,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08352760268747464,"score_gpt":0.3355123175108106,"score_spread":0.251984714823336,"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."}}