{"id":"W4416781075","doi":"10.1016/j.esmorw.2025.100597","title":"401P Persistent bias in head and neck segmentation: Evidence from MedSAM2","year":2025,"lang":"en","type":"article","venue":"ESMO Real World Data and Digital Oncology","topic":"Advanced Radiotherapy Techniques","field":"Physics and Astronomy","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Princess Margaret Cancer Centre","funders":"","keywords":"Head and neck; Segmentation; Head (geology); Radiation therapy; Foundation (evidence)","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":[0.05133538,0.001209197,0.001659809,0.001541314,0.001357729,0.003409519,0.003319751,0.002982922,0.009147909],"category_scores_gemma":[0.2177016,0.001098876,0.00322997,0.002249092,0.002897741,0.002899504,0.002757228,0.002042048,0.005489076],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00141348,"about_ca_system_score_gemma":0.002054136,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01524246,"about_ca_topic_score_gemma":0.01687605,"domain_scores_codex":[0.9715328,0.01801354,0.001216313,0.005271078,0.003560621,0.0004055918],"domain_scores_gemma":[0.816328,0.1348164,0.01401973,0.02488959,0.008574332,0.0013719],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0116356,0.0003179387,0.4039675,0.005653782,0.01871313,0.0006459708,0.001719823,0.05285694,0.003964582,0.008811486,0.0884715,0.4032417],"study_design_scores_gemma":[0.002303791,0.001966354,0.456589,0.004580733,0.01079182,0.008669276,0.001209314,0.1617292,0.01853097,0.1336533,0.1989165,0.001059812],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.611419,0.06461646,0.1674121,0.02788853,0.003667271,0.0007996676,0.06385718,0.01140643,0.04893337],"genre_scores_gemma":[0.9379574,0.003313836,0.02549453,0.005940259,0.0010439,0.0002295217,0.01944639,0.003477479,0.003096808],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.05133538,"threshold_uncertainty_score":0.2714907,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07141697942490335,"score_gpt":0.3852661163662907,"score_spread":0.3138491369413874,"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."}}