{"id":"W3128830162","doi":"10.1007/s12021-021-09510-1","title":"Improved Segmentation of the Intracranial and Ventricular Volumes in Populations with Cerebrovascular Lesions and Atrophy Using 3D CNNs","year":2021,"lang":"en","type":"article","venue":"Neuroinformatics","topic":"Acute Ischemic Stroke Management","field":"Medicine","cited_by":33,"is_retracted":false,"has_abstract":false,"ca_institutions":"Ontario Brain Institute; Université de Montréal; University of Toronto; Western University; Heart and Stroke Foundation; York University; Montreal Heart Institute; Toronto Western Hospital; Ottawa Hospital; University of Ottawa; Thunder Bay Regional Research Institute; Sunnybrook Health Science Centre","funders":"Canadian Institutes of Health Research; Canadian Network for Research and Innovation in Machining Technology, Natural Sciences and Engineering Research Council of Canada; Réseau en Bio-Imagerie du Quebec; Ontario Brain Institute; Heart and Stroke Foundation of Canada","keywords":"Segmentation; Sørensen–Dice coefficient; Artificial intelligence; Convolutional neural network; Neuroimaging; Pattern recognition (psychology); Computer science; Atrophy; Conditional random field; Hyperintensity; Robustness (evolution); Magnetic resonance imaging; Image segmentation; Medicine; Pathology; Radiology; Psychology; Neuroscience","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":[0.00004748748,0.00007727845,0.0001409291,0.00005603712,0.00005517033,0.00001605639,0.0000295967,0.00002754177,0.000005319491],"category_scores_gemma":[0.00005997737,0.00005507158,0.00002827807,0.0002179305,0.00004929854,0.0001063004,0.00008756069,0.0001022856,2.121755e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002762329,"about_ca_system_score_gemma":0.00004443596,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002816769,"about_ca_topic_score_gemma":0.00001521774,"domain_scores_codex":[0.9993988,0.00001736158,0.0002415213,0.00007752769,0.0001685778,0.00009616381],"domain_scores_gemma":[0.999612,0.00001661178,0.0001006245,0.0001885633,0.00004901343,0.00003320576],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001468375,0.0003720917,0.8846077,0.002372327,0.0003677642,0.0002083531,0.004692433,0.004491976,0.07304722,0.0004068334,0.000491883,0.02879461],"study_design_scores_gemma":[0.004629794,0.0002138628,0.4297506,0.0002772602,0.0007181387,0.001017966,0.002494351,0.5538492,0.006347724,0.00002153551,0.0004793092,0.000200275],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9948637,0.0000589937,0.004300492,0.0002186773,0.0000435672,0.0003330657,0.000005160718,0.000007586105,0.0001687264],"genre_scores_gemma":[0.9580752,0.00004082754,0.04162605,0.000181412,0.00001757733,0.00000323161,0.00001270415,0.000009172647,0.00003380828],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5493572,"threshold_uncertainty_score":0.2245754,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01501105714295613,"score_gpt":0.2398766414263236,"score_spread":0.2248655842833675,"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."}}