{"id":"W2912848025","doi":"10.1002/hbm.24803","title":"Accurate and robust segmentation of neuroanatomy in T1‐weighted MRI by combining spatial priors with deep convolutional neural networks","year":2019,"lang":"en","type":"preprint","venue":"Human Brain Mapping","topic":"Medical Image Segmentation Techniques","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; Montreal Neurological Institute and Hospital","funders":"National Institute of Biomedical Imaging and Bioengineering; Fonds de Recherche du Québec - Santé; University of California, San Diego; Genentech; University of California, Los Angeles; U.S. Food and Drug Administration; National Institutes of Health; Canada First Research Excellence Fund; Eisai; Northern California Institute for Research and Education; F. Hoffmann-La Roche; Elan; Novartis; Medpace; GlaxoSmithKline; AstraZeneca; Eli Lilly and Company; Bristol-Myers Squibb; Pfizer; Synarc; National Institute on Aging; Alzheimer's Association; Foundation for the National Institutes of Health","keywords":"Sørensen–Dice coefficient; Segmentation; Artificial intelligence; Computer science; Convolutional neural network; Pattern recognition (psychology); Population; Deep learning; Prior probability; Robustness (evolution); Image segmentation; Computer vision; Bayesian probability","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0006090114,0.0003209127,0.0004516114,0.0003485967,0.0001243063,0.0002201724,0.0006836264,0.0001939128,0.00002747845],"category_scores_gemma":[0.00004531915,0.0003245959,0.00004967669,0.0002898143,0.0002071463,0.0004413536,0.0008412868,0.0007631329,9.359519e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001089636,"about_ca_system_score_gemma":0.00008821353,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002085659,"about_ca_topic_score_gemma":0.00004625812,"domain_scores_codex":[0.9973072,0.0003593746,0.0006972984,0.0007729678,0.0005220509,0.0003410965],"domain_scores_gemma":[0.9983281,0.0002903386,0.0006912149,0.0004483828,0.0001339322,0.0001080035],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0002759265,0.001163559,0.1252872,0.004460272,0.0007972784,0.0004134765,0.02984801,0.6485478,0.04969214,0.01545669,0.01883663,0.105221],"study_design_scores_gemma":[0.001073953,0.00009867551,0.02895111,0.0003996632,0.000009563201,0.000007778286,0.0001292241,0.9680721,0.0003337082,0.0005618039,0.00001201036,0.0003504194],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0556854,0.0001449392,0.942376,0.0004496757,0.0001926884,0.0008837172,0.000005738009,0.0001580273,0.0001038056],"genre_scores_gemma":[0.9156471,0.00002485317,0.08304347,0.0008108346,0.00005460919,0.00008097328,0.0002638874,0.00002976938,0.00004451475],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8599617,"threshold_uncertainty_score":0.9999206,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01972984430705215,"score_gpt":0.2644455397071143,"score_spread":0.2447156954000621,"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."}}