{"id":"W3215649600","doi":"10.1101/2021.12.01.470730","title":"Benchmarking Geometric Deep Learning for Cortical Segmentation and Neurodevelopmental Phenotype Prediction","year":2021,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Neonatal and fetal brain pathology","field":"Medicine","cited_by":23,"is_retracted":false,"has_abstract":true,"ca_institutions":"École de Technologie Supérieure","funders":"National Health and Medical Research Council; Engineering and Physical Sciences Research Council; Menzies Centre for Australian Studies, King's College London, University of London; Medical Research Council; Medical Research Council Centre for Neurodevelopmental Disorders; European Commission; King's College London; Commonwealth Scholarship Commission; National Institute for Health and Care Research; Maudsley Charity; Centre For Medical Engineering, King’s College London; British Heart Foundation; Wellcome Trust","keywords":"Deep learning; Human Connectome Project; Artificial intelligence; Segmentation; Benchmarking; Computer science; Convolutional neural network; Machine learning; Pattern recognition (psychology); Neuroscience; Functional connectivity; Psychology","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0003826132,0.0003403795,0.0004875405,0.0003490418,0.0002226078,0.0001027907,0.00008741724,0.0003799209,0.00004231119],"category_scores_gemma":[0.000550022,0.0003692002,0.000104865,0.0004461209,0.0000978296,0.0001149568,0.0003087287,0.0008165026,0.00000564224],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001887129,"about_ca_system_score_gemma":0.0002321514,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000006299816,"about_ca_topic_score_gemma":6.120258e-7,"domain_scores_codex":[0.9979223,0.00009170146,0.0004436869,0.0008714526,0.0002861592,0.0003847215],"domain_scores_gemma":[0.998812,0.0001468722,0.0002125226,0.0002823343,0.0003160914,0.0002301915],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"observational","study_design_scores_codex":[0.0002348636,0.0001864341,0.06783433,0.001377254,0.000230668,0.0001998413,0.00005765912,0.00009537196,0.9291825,0.00006237174,0.00002913929,0.0005095618],"study_design_scores_gemma":[0.001494327,0.0003145628,0.9460567,0.0003049373,0.0003409135,6.205279e-7,0.00002636929,0.005733204,0.04469347,9.082337e-7,0.0006070549,0.0004269681],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9675983,0.001556413,0.02848303,0.00009350476,0.001111338,0.0009136375,0.000038748,0.000196925,0.000008134677],"genre_scores_gemma":[0.9760818,0.0003663526,0.02257601,0.0002552024,0.0004321496,0.000194269,0.00001094399,0.00007642611,0.000006844594],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8844891,"threshold_uncertainty_score":0.999876,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01344681618870399,"score_gpt":0.2286054861470394,"score_spread":0.2151586699583354,"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."}}