{"id":"W4205334579","doi":"10.1016/j.compbiomed.2022.105230","title":"Impact of feature harmonization on radiogenomics analysis: Prediction of EGFR and KRAS mutations from non-small cell lung cancer PET/CT images","year":2022,"lang":"en","type":"article","venue":"Computers in Biology and Medicine","topic":"Radiomics and Machine Learning in Medical Imaging","field":"Medicine","cited_by":88,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia; BC Cancer Agency","funders":"BC Cancer Foundation; Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung; National Science Foundation","keywords":"Radiogenomics; KRAS; Feature (linguistics); Lung cancer; Cancer; Harmonization; Computer science; Computational biology; Medicine; Artificial intelligence; Radiomics; Oncology; Biology; Internal medicine; Colorectal cancer","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":[0.0002595965,0.0001101872,0.0004444946,0.0003854337,0.00006800204,0.000001954383,0.00006099919,0.00002023253,0.00003983577],"category_scores_gemma":[0.00005601265,0.0000849457,0.00005560078,0.0003495821,0.0002210572,0.0000161989,0.00004833399,0.0003250229,3.154952e-8],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007841958,"about_ca_system_score_gemma":0.00006178382,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0008734161,"about_ca_topic_score_gemma":0.00000601901,"domain_scores_codex":[0.9992289,0.00009469049,0.0002512911,0.0002336965,0.00007750058,0.0001138996],"domain_scores_gemma":[0.999392,0.0002039405,0.0001681835,0.000121301,0.00003922297,0.00007530054],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0001407157,0.00009094195,0.9712985,0.00006710229,0.0005165854,0.00004332936,0.0008999091,0.007730294,0.01419032,0.00003149701,0.0009573765,0.004033384],"study_design_scores_gemma":[0.002871643,0.0007819758,0.7657509,0.0001362862,0.0008588169,0.00006158056,0.0002034777,0.2288679,0.0002160756,0.0001012979,0.00008210022,0.00006783996],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9842637,0.003584869,0.01027879,0.001251831,0.0002441099,0.0001967512,0.00009776797,0.000009804555,0.00007241077],"genre_scores_gemma":[0.9953913,0.00133044,0.002373526,0.0002601083,0.0001044435,0.00001081898,0.000498247,0.000008241612,0.00002284787],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2211377,"threshold_uncertainty_score":0.3463985,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.007789998847016184,"score_gpt":0.3044521610248692,"score_spread":0.296662162177853,"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."}}