{"id":"W2899974865","doi":"10.1109/trpms.2018.2880617","title":"An Empirical Approach for Avoiding False Discoveries When Applying High-Dimensional Radiomics to Small Datasets","year":2018,"lang":"en","type":"article","venue":"IEEE Transactions on Radiation and Plasma Medical Sciences","topic":"Radiomics and Machine Learning in Medical Imaging","field":"Medicine","cited_by":25,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University Health Centre","funders":"Natural Sciences and Engineering Research Council of Canada; Canadian Institutes of Health Research; Terry Fox Research Institute","keywords":"Overfitting; Feature selection; Artificial intelligence; Radiomics; Logistic regression; Computer science; Support vector machine; Feature (linguistics); Machine learning; Receiver operating characteristic; Pattern recognition (psychology)","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.1643381,0.001686574,0.001878246,0.004181436,0.002088431,0.003506687,0.003750833,0.002751661,0.001340947],"category_scores_gemma":[0.4598012,0.00102739,0.002022326,0.003112986,0.005803616,0.003063596,0.003568735,0.003475376,0.0003953989],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001664213,"about_ca_system_score_gemma":0.003017287,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00115004,"about_ca_topic_score_gemma":0.001451754,"domain_scores_codex":[0.8811923,0.08991619,0.007917984,0.008571129,0.01191356,0.0004888894],"domain_scores_gemma":[0.5545957,0.346798,0.02709302,0.05227303,0.01810789,0.001132341],"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":[0.001781735,0.0008748016,0.1540241,0.002882463,0.004814696,0.001832478,0.003287216,0.09002985,0.01518037,0.05132441,0.008374849,0.6655931],"study_design_scores_gemma":[0.000772954,0.002549803,0.04763909,0.001500744,0.001583248,0.002752984,0.001121269,0.7493818,0.03389855,0.1328686,0.02555785,0.000372986],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01752067,0.0003498571,0.9790031,0.0007879164,0.00007299795,0.0007081615,0.0001143606,0.0006179299,0.0008248608],"genre_scores_gemma":[0.2383111,0.0001511564,0.7576272,0.0006184973,0.0001395415,0.002397503,0.0002233468,0.0001780402,0.0003536963],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.1643381,"threshold_uncertainty_score":0.8691131,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03991037103679596,"score_gpt":0.3378253655991552,"score_spread":0.2979149945623593,"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."}}