{"id":"W3198929365","doi":"10.1016/j.cpet.2021.06.013","title":"Objective Task-Based Evaluation of Artificial Intelligence-Based Medical Imaging Methods","year":2021,"lang":"en","type":"review","venue":"PET Clinics","topic":"Medical Imaging Techniques and Applications","field":"Medicine","cited_by":42,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of British Columbia","funders":"National Institutes of Health","keywords":"Medicine; Artificial intelligence; Task (project management); Medical physics; Machine learning; Systems engineering; Computer science; Engineering","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":["metaresearch","metaepi_narrow","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.01033734,0.000360372,0.002172763,0.0002787056,0.0000616998,0.00002151528,0.000351651,0.0003281439,0.001464459],"category_scores_gemma":[0.01466496,0.0002950841,0.0009448005,0.0008215305,0.000370512,0.00002228194,0.00009406696,0.001200125,0.00002728414],"about_ca_system_candidate":true,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002355699,"about_ca_system_score_gemma":0.009963785,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001773481,"about_ca_topic_score_gemma":0.000003329016,"domain_scores_codex":[0.9940381,0.001439143,0.001927784,0.0006711748,0.001639656,0.0002841459],"domain_scores_gemma":[0.993601,0.003129365,0.0008539413,0.0009789291,0.001104036,0.0003326823],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00000582099,0.0003880641,0.000003781953,0.003305107,0.00008026569,0.00002715959,0.000009729244,0.000002766554,0.000001180416,0.0007316596,0.0007749636,0.9946695],"study_design_scores_gemma":[0.0002384614,0.00007290635,0.000001060487,0.02438555,0.006056084,0.00006644429,0.0000368134,0.2688003,0.000103491,0.002254447,0.6976587,0.0003258389],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"review","genre_gemma":"review","genre_scores_codex":[0.00000130232,0.6161596,0.3806663,0.001342305,0.0002506923,0.001100293,0.00003936496,0.00009900195,0.0003411021],"genre_scores_gemma":[0.00003251429,0.7390231,0.2584242,0.0006562933,0.0003846791,0.0005176669,0.0008793153,0.00006520599,0.00001699602],"genre_candidate":"review","genre_consensus":"review","teacher_disagreement_score":0.9943436,"threshold_uncertainty_score":0.9999501,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2486077522918173,"score_gpt":0.594007935415539,"score_spread":0.3454001831237217,"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."}}