{"id":"W3156251590","doi":"10.1088/1361-6560/ac3e0e","title":"Understanding machine learning classifier decisions in automated radiotherapy quality assurance","year":2021,"lang":"en","type":"article","venue":"Physics in Medicine and Biology","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":10,"is_retracted":false,"has_abstract":true,"ca_institutions":"Princess Margaret Cancer Centre; University of Toronto","funders":"Canadian Institutes of Health Research","keywords":"Computer science; Classifier (UML); Artificial intelligence; Machine learning; Quality assurance; Medicine","routes":{"ca_aff":true,"ca_fund":true,"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":[0.000811398,0.0001104037,0.0003094824,0.00009926836,0.00008342171,0.00002064816,0.0002228702,0.00007584129,0.00001223453],"category_scores_gemma":[0.0005621567,0.00008993884,0.00002124252,0.0008294639,0.0001584291,0.0001458636,0.00009240583,0.0003017653,0.000002855539],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00009508599,"about_ca_system_score_gemma":0.00006507262,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003115635,"about_ca_topic_score_gemma":0.0006857447,"domain_scores_codex":[0.9985422,0.0003988422,0.000328365,0.0003704398,0.0000913009,0.0002687767],"domain_scores_gemma":[0.9983659,0.001236667,0.00007672382,0.0002268488,0.00004586973,0.00004799654],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00002533544,0.0001433559,0.1027489,0.00002004505,0.00002459324,0.0001195589,0.003608472,0.002536254,0.060837,0.7761009,0.0002875997,0.05354798],"study_design_scores_gemma":[0.0008699605,0.0002050195,0.009157265,0.0001913254,0.000003225969,0.00001724092,0.00157575,0.6959122,0.002827448,0.2865565,0.002412779,0.0002712794],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0961013,0.001613546,0.8948093,0.005791955,0.0002822757,0.00008350034,0.000001019336,0.0001117331,0.001205393],"genre_scores_gemma":[0.9923588,0.001704041,0.00511127,0.0006936739,0.00008354284,0.00000557221,0.000007695206,0.000005397467,0.000030022],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8962575,"threshold_uncertainty_score":0.3667599,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.5385558467910203,"score_gpt":0.4658012825977877,"score_spread":0.07275456419323262,"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."}}