{"id":"W4210506170","doi":"10.1002/mp.15514","title":"On the proper use of structural similarity for the robust evaluation of medical image synthesis models","year":2022,"lang":"en","type":"article","venue":"Medical Physics","topic":"Image and Video Quality Assessment","field":"Computer Science","cited_by":27,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université Laval","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Normalization (sociology); Voxel; Image quality; Artificial intelligence; Metric (unit); Pattern recognition (psychology); Computer science; Medical imaging; Similarity (geometry); Mathematics; Computation; Image processing; Image (mathematics); Algorithm","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":[0.004916996,0.00008811663,0.0001687416,0.00001396415,0.0002452881,0.00003866157,0.001301717,0.00003954039,0.0002646984],"category_scores_gemma":[0.002061356,0.00004517637,0.0001133334,0.0001950455,0.0002672654,0.0002434769,0.0005136569,0.0003095595,6.64347e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005217028,"about_ca_system_score_gemma":0.0005411714,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000933377,"about_ca_topic_score_gemma":0.000004614843,"domain_scores_codex":[0.9944797,0.0006229113,0.0002857357,0.0001968244,0.004259309,0.0001555229],"domain_scores_gemma":[0.9961145,0.002891107,0.0001408502,0.0005595886,0.0002357179,0.00005824746],"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":[0.00006807376,0.0006801435,0.00003417497,0.0001224021,0.0001898518,0.000003922743,0.003072999,0.0502968,0.0001508196,0.5880013,0.01528323,0.3420963],"study_design_scores_gemma":[0.0002131387,0.00006052963,0.00009688251,0.00002151688,0.00004217739,0.00000133138,0.00007515414,0.9374565,0.00226778,0.0596298,0.00007930492,0.0000558069],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01931825,0.00003241253,0.9573768,0.02242363,0.0002021528,0.0005334221,0.00002599655,0.0000150309,0.00007234549],"genre_scores_gemma":[0.9966339,0.000004129728,0.001587294,0.001489441,0.00008251479,0.0001845538,0.000003521722,0.000006257249,0.000008391969],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9773157,"threshold_uncertainty_score":0.2898262,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2245681303530576,"score_gpt":0.3715333982732142,"score_spread":0.1469652679201567,"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."}}