{"id":"W3000524228","doi":"10.1016/j.media.2020.101636","title":"Improving cardiac MRI convolutional neural network segmentation on small training datasets and dataset shift: A continuous kernel cut approach","year":2020,"lang":"en","type":"article","venue":"Medical Image Analysis","topic":"Radiomics and Machine Learning in Medical Imaging","field":"Medicine","cited_by":64,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Toronto; Sunnybrook Health Science Centre","funders":"Engineering and Physical Sciences Research Council; Canadian Institutes of Health Research; British Heart Foundation; Medical Research Council; National Institute for Health and Care Research","keywords":"Segmentation; Convolutional neural network; Computer science; Artificial intelligence; Pattern recognition (psychology); Deep learning; Kernel (algebra); Market segmentation; Image segmentation; Scale-space segmentation; Machine learning; 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":[],"consensus_categories":[],"category_scores_codex":[0.001111543,0.000273032,0.0008517423,0.0001445082,0.0001923666,0.0001105138,0.0002161476,0.0001301712,0.0002910737],"category_scores_gemma":[0.00139272,0.0002296605,0.0002285837,0.0006954712,0.0003390982,0.0001645117,0.000174732,0.0009243909,0.00001641416],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004241976,"about_ca_system_score_gemma":0.0001211912,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002908823,"about_ca_topic_score_gemma":0.000009709991,"domain_scores_codex":[0.9970945,0.0002444622,0.0005237873,0.0007419116,0.0008993943,0.0004959598],"domain_scores_gemma":[0.9981671,0.0002830682,0.0001784091,0.0003266565,0.00003807357,0.001006715],"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":[0.001780207,0.001260613,0.1854486,0.001461759,0.01719297,0.004374256,0.007894929,0.01082852,0.004636277,0.0009145208,0.2973647,0.4668427],"study_design_scores_gemma":[0.001855382,0.00018572,0.009304385,0.00004845737,0.003329467,0.0000410517,0.0004424689,0.9792479,0.0000228272,0.00001901835,0.00521603,0.0002873138],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2104686,0.00156101,0.7370141,0.0448611,0.0003704427,0.001056339,0.003502074,0.000324845,0.0008415798],"genre_scores_gemma":[0.8799027,0.0001645032,0.03235389,0.02615561,0.002088389,0.00005047893,0.05917456,0.00006510387,0.00004471323],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9684194,"threshold_uncertainty_score":0.9365281,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01914023953641023,"score_gpt":0.2829352904255403,"score_spread":0.2637950508891301,"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."}}