{"id":"W2901038558","doi":"10.1016/j.neuroimage.2019.03.026","title":"Unsupervised domain adaptation for medical imaging segmentation with self-ensembling","year":2019,"lang":"en","type":"preprint","venue":"NeuroImage","topic":"Domain Adaptation and Few-Shot Learning","field":"Computer Science","cited_by":14,"is_retracted":false,"has_abstract":false,"ca_institutions":"Université de Montréal; Polytechnique Montréal","funders":"Fonds de Recherche du Québec - Santé; Natural Sciences and Engineering Research Council of Canada; Institut de Valorisation des Données; Canada First Research Excellence Fund; Canadian Institutes of Health Research; Coordenação de Aperfeiçoamento de Pessoal de Nível Superior; Canada Foundation for Innovation; Réseau en Bio-Imagerie du Quebec","keywords":"Computer science; Segmentation; Generalization; Artificial intelligence; Domain adaptation; Domain (mathematical analysis); Task (project management); Adaptation (eye); Medical imaging; Machine learning; Modality (human–computer interaction); Deep learning; Image (mathematics); Image segmentation; Pattern recognition (psychology); Psychology; Mathematics","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":[0.0008434004,0.0003641372,0.0003460474,0.0002760566,0.0002075651,0.0006881619,0.0009989114,0.000158305,0.00002700756],"category_scores_gemma":[0.0001459066,0.0003492637,0.0001399367,0.0002664695,0.0000428955,0.0005794227,0.0004846213,0.0007009941,0.00005783193],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001019217,"about_ca_system_score_gemma":0.0005401691,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001895162,"about_ca_topic_score_gemma":0.000007406169,"domain_scores_codex":[0.9967605,0.0002644295,0.0004778664,0.001058587,0.001009081,0.0004294704],"domain_scores_gemma":[0.9980261,0.0004059237,0.0003647598,0.000759833,0.0002524797,0.0001909347],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000869023,0.001327548,0.009506942,0.004346627,0.0006361988,0.00125973,0.069539,0.2665583,0.02062291,0.05315362,0.005219823,0.5669603],"study_design_scores_gemma":[0.001884554,0.00008523072,0.001878826,0.0001745189,0.00002949073,0.00004238775,0.0004302474,0.9896351,0.0002047397,0.001389115,0.003794899,0.0004509297],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01544599,0.0001000003,0.977544,0.002379404,0.00102572,0.001125012,0.000008272525,0.0006383376,0.001733341],"genre_scores_gemma":[0.3995712,0.00005868868,0.5966442,0.002727589,0.0002829792,0.000208335,0.0001739164,0.00009731127,0.0002357521],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.7230768,"threshold_uncertainty_score":0.9998959,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02208971409744873,"score_gpt":0.2715188129530055,"score_spread":0.2494290988555568,"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."}}