{"id":"W4392654434","doi":"10.1136/heartjnl-2024-bscmr.9","title":"11 Visual quality control of assessment of AI-assisted high-volume CMR segmentation in the UK Biobank","year":2024,"lang":"en","type":"article","venue":"Abstracts","topic":"Radiomics and Machine Learning in Medical Imaging","field":"Medicine","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Circle Cardiovascular Imaging","funders":"","keywords":"Computer science; Visual inspection; Outlier; Artificial intelligence; Image quality; Quality assurance; Segmentation; Biobank; Reliability (semiconductor); Visualization; Image processing; Computer vision; Data mining; Medicine; Image (mathematics); Pathology","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":[0.02306976,0.0004342398,0.0003754612,0.002264466,0.0007429312,0.002341605,0.001565396,0.0008432965,0.004182224],"category_scores_gemma":[0.06373294,0.0004346411,0.0004835337,0.001584117,0.001268791,0.0006624297,0.001326556,0.0004011935,0.001420636],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0014894,"about_ca_system_score_gemma":0.001021824,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008742722,"about_ca_topic_score_gemma":0.01042087,"domain_scores_codex":[0.9863934,0.005489404,0.002144533,0.001348221,0.004302201,0.0003222344],"domain_scores_gemma":[0.9351377,0.02072026,0.01100062,0.007733002,0.02466203,0.0007462737],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.005538564,0.0003751179,0.3428033,0.001801323,0.0003563234,0.001013264,0.007119223,0.007770174,0.08492087,0.00162813,0.02356001,0.5231136],"study_design_scores_gemma":[0.0002152054,0.00162412,0.8615792,0.000394797,0.0002190087,0.002822419,0.001309684,0.03746023,0.07457523,0.001056684,0.01845887,0.0002846305],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8029794,0.002392175,0.1710114,0.0008215962,0.0003814773,0.002075633,0.003505435,0.00391004,0.01292297],"genre_scores_gemma":[0.8854845,0.0003417653,0.1078922,0.0002082684,0.00009421186,0.00073839,0.001788531,0.0005994372,0.002852611],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02306976,"threshold_uncertainty_score":0.1220061,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01662010893048931,"score_gpt":0.3900369164801031,"score_spread":0.3734168075496138,"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."}}