{"id":"W2801459697","doi":"10.1016/j.compmedimag.2019.04.005","title":"Combo loss: Handling input and output imbalance in multi-organ segmentation","year":2019,"lang":"en","type":"article","venue":"Computerized Medical Imaging and Graphics","topic":"Radiomics and Machine Learning in Medical Imaging","field":"Medicine","cited_by":23,"is_retracted":false,"has_abstract":false,"ca_institutions":"Simon Fraser University","funders":"","keywords":"Artificial intelligence; Computer science; Segmentation; Deep learning; Voxel; False positive paradox; Pattern recognition (psychology); False positives and false negatives; Image segmentation; Artificial neural network; Inference; Medical imaging; Leverage (statistics); Machine learning; Autoencoder; Cross entropy; Computer vision","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":[],"consensus_categories":[],"category_scores_codex":[0.0007291216,0.0001955565,0.0004583652,0.0002372166,0.00007881365,0.00007053096,0.0001012986,0.0001014691,0.00001648573],"category_scores_gemma":[0.0002684037,0.0001699276,0.00004820757,0.0002332434,0.0003324354,0.0001064052,0.0001348366,0.0007633358,0.000004596654],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002227028,"about_ca_system_score_gemma":0.00007457442,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00008344313,"about_ca_topic_score_gemma":0.000003724625,"domain_scores_codex":[0.9983985,0.00008485233,0.00036603,0.0004236751,0.0004010907,0.000325829],"domain_scores_gemma":[0.9990854,0.0001997831,0.00008311022,0.0001833972,0.0000498602,0.0003985011],"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":[0.00006540472,0.0001405172,0.88535,0.000347334,0.00003778203,0.0002608512,0.0007399158,0.000006805975,0.001948884,0.0003093075,0.0001750975,0.1106181],"study_design_scores_gemma":[0.01060896,0.00007991639,0.2388928,0.001247996,0.00004281099,0.0005722212,0.000142732,0.7440141,0.00007670326,0.0002533804,0.003793243,0.0002750916],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9341645,0.001641792,0.05488313,0.008356346,0.0004652457,0.0002938804,0.000001475743,0.00009659835,0.00009706609],"genre_scores_gemma":[0.9806845,0.001576343,0.01177863,0.005674955,0.0001193091,0.000006279981,0.00003219107,0.00002835234,0.00009949508],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7440073,"threshold_uncertainty_score":0.6929445,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01069846836706176,"score_gpt":0.2887381605245697,"score_spread":0.2780396921575079,"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."}}