{"id":"W4403659763","doi":"10.3389/fncom.2024.1487877","title":"Multi-stage semi-supervised learning enhances white matter hyperintensity segmentation","year":2024,"lang":"en","type":"article","venue":"Frontiers in Computational Neuroscience","topic":"Dementia and Cognitive Impairment Research","field":"Medicine","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Foothills Medical Centre; Women and Children’s Health Research Institute; University of Alberta; University of Calgary","funders":"Canadian Institutes of Health Research; National Institutes of Health; Genentech; IXICO; H. Lundbeck A/S; Servier; Eisai; Canada Foundation for Innovation; Northern California Institute for Research and Education; Pfizer; Novartis Pharmaceuticals Corporation; University of Southern California; Biogen; Eli Lilly and Company; Bristol-Myers Squibb; BioClinica; U.S. Department of Defense; Alzheimer's Disease Neuroimaging Initiative; Meso Scale Diagnostics; Alzheimer's Association","keywords":"Fluid-attenuated inversion recovery; Segmentation; Hyperintensity; Artificial intelligence; Computer science; Ground truth; Deep learning; Pattern recognition (psychology); Dementia; Gold standard (test); Cognitive decline; Magnetic resonance imaging; Machine learning; Medicine; Radiology; Disease; Pathology","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":[0.0002615212,0.0001099472,0.000133134,0.0002923422,0.0001109601,0.0001333516,0.00008603407,0.00002674209,0.0001385182],"category_scores_gemma":[0.00006852193,0.0001019738,0.00005082956,0.0005410059,0.0001550578,0.0002953309,0.00005853647,0.000267071,0.00006470671],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007530002,"about_ca_system_score_gemma":0.0001129762,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000005518856,"about_ca_topic_score_gemma":8.480496e-7,"domain_scores_codex":[0.9985409,0.00007940548,0.0001876686,0.0004345624,0.0005042597,0.0002531658],"domain_scores_gemma":[0.9996831,0.00004686771,0.00002386628,0.00006440361,0.00009531012,0.00008640827],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00004643739,0.00009786753,0.9655484,0.0001112911,0.000005503911,0.0001516213,0.0002689631,0.007676072,0.02236308,0.000008992797,0.00124231,0.00247949],"study_design_scores_gemma":[0.0003119188,0.00008221949,0.5779179,0.00005185443,0.000005490154,0.00002146074,0.0002385589,0.4199307,0.00072132,0.00004225224,0.0006114359,0.00006481984],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6726271,0.00009531485,0.3246836,0.001016231,0.0007429985,0.000262122,0.000006308128,0.00005908888,0.0005071742],"genre_scores_gemma":[0.967857,0.00002561718,0.0250044,0.001154426,0.00003410677,0.0000244914,0.00002959772,0.00001206066,0.005858272],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.4122547,"threshold_uncertainty_score":0.4158373,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0326673834892553,"score_gpt":0.3306069944613496,"score_spread":0.2979396109720943,"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."}}