{"id":"W4391345880","doi":"10.1038/s41598-023-50382-z","title":"Edge roughness quantifies impact of physician variation on training and performance of deep learning auto-segmentation models for the esophagus","year":2024,"lang":"en","type":"article","venue":"Scientific Reports","topic":"Radiomics and Machine Learning in Medical Imaging","field":"Medicine","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"National Cancer Institute; National Institutes of Health","keywords":"Segmentation; Esophagus; Artificial intelligence; Computer science; Deep learning; Variation (astronomy); Surface roughness; Medicine; Computer vision; Pattern recognition (psychology); Anatomy; Materials science; Composite material","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":[],"consensus_categories":[],"category_scores_codex":[0.001581321,0.00009015392,0.0001982295,0.0001386147,0.0001697507,0.00008840156,0.00003934376,0.00003101184,0.000009238599],"category_scores_gemma":[0.0002082017,0.00005753951,0.0001052068,0.0002594028,0.0001698273,0.000175362,0.00001484376,0.0001627232,3.678434e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003672338,"about_ca_system_score_gemma":0.0001091255,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002920621,"about_ca_topic_score_gemma":5.436387e-7,"domain_scores_codex":[0.9988957,0.00002591165,0.0003427616,0.0002887694,0.0003038798,0.0001429715],"domain_scores_gemma":[0.9992598,0.0001919131,0.0002108039,0.000192524,0.0001041189,0.00004088888],"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":[0.0001617555,0.0001170136,0.009680097,0.001179105,0.0003089476,0.00003718453,0.02877141,0.1983724,0.07852384,0.001168752,0.000325462,0.681354],"study_design_scores_gemma":[0.0001676636,0.0002481855,0.0138375,0.0003001075,0.00009958822,0.00004847741,0.0003856138,0.9794447,0.004174478,0.001176663,0.00006109464,0.0000559101],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9555057,0.001738645,0.04051555,0.0001501324,0.001540563,0.0003469981,0.000001250438,0.00003551293,0.0001656443],"genre_scores_gemma":[0.998684,0.00002409845,0.0008347435,0.00001314306,0.00008571653,0.00001151406,0.00002997961,0.00001479852,0.0003019325],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7810723,"threshold_uncertainty_score":0.2346393,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02435172172744584,"score_gpt":0.3212167419278237,"score_spread":0.2968650202003779,"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."}}