{"id":"W4392247492","doi":"10.1016/j.ctro.2024.100747","title":"Determining risk and predictors of head and neck cancer treatment-related lymphedema: A clinicopathologic and dosimetric data mining approach using interpretable machine learning and ensemble feature selection","year":2024,"lang":"en","type":"article","venue":"Clinical and Translational Radiation Oncology","topic":"Lymphatic System and Diseases","field":"Medicine","cited_by":14,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Canadian Institutes of Health Research; National Institutes of Health; Northwestern University; National Cancer Institute; American Lung Association","keywords":"Feature selection; Lymphedema; Medicine; Head and neck cancer; Feature (linguistics); Radiation therapy; Artificial intelligence; Multicollinearity; Ensemble learning; Machine learning; Head and neck; Cancer; Pattern recognition (psychology); Computer science; Radiology; Breast cancer; Regression analysis; Surgery; Internal medicine","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001870467,0.0005290681,0.000690762,0.001761919,0.0002034334,0.0005217658,0.0004634848,0.0003324226,0.0003000531],"category_scores_gemma":[0.003978559,0.0001492382,0.000737992,0.0008684245,0.000150143,0.000434288,0.0003477829,0.0005726563,0.00008385223],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004377256,"about_ca_system_score_gemma":0.0004522212,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002609267,"about_ca_topic_score_gemma":0.003197443,"domain_scores_codex":[0.999544,0.0001667148,0.00004347909,0.000101066,0.0001055646,0.00003912907],"domain_scores_gemma":[0.9979595,0.001099808,0.0003849608,0.0001813261,0.0003047085,0.00006971385],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0003831391,0.0004057428,0.6370379,0.00006906253,0.0003220354,0.0002783507,0.0001265457,0.1477877,0.005703501,0.000279822,0.001357388,0.2062488],"study_design_scores_gemma":[0.00001581505,0.0002487998,0.1275607,0.00002050267,0.0001338141,0.000339976,0.00008520224,0.8674976,0.002495141,0.001202606,0.0003764997,0.00002332301],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8560653,0.0005476741,0.1416104,0.0002910204,0.00002120446,0.00007634287,0.0006782591,0.0003205049,0.000389258],"genre_scores_gemma":[0.979317,0.00008581205,0.01991956,0.00001972516,0.00001642419,0.00003659642,0.0004987737,0.000007298108,0.00009871145],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.002609267,"threshold_uncertainty_score":0.009892106,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06688330504666729,"score_gpt":0.40447750023779,"score_spread":0.3375941951911227,"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."}}