{"id":"W4410842945","doi":"10.1038/s41746-025-01729-5","title":"Comparative analysis of natural language processing methodologies for classifying computed tomography enterography reports in Crohn’s disease patients","year":2025,"lang":"en","type":"article","venue":"npj Digital Medicine","topic":"Radiomics and Machine Learning in Medical Imaging","field":"Medicine","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"Alberta Innovates; Canadian Institutes of Health Research; Mitacs; University of Alberta; CIHR Skin Research Training Centre; Alberta Machine Intelligence Institute; Government of Canada; Canadian Institute for Advanced Research; Natural Sciences and Engineering Research Council of Canada; Alberta Health Services","keywords":"Crohn's disease; Computed tomography; Medicine; Disease; Radiology; Crohn disease; Artificial intelligence; Natural language processing; Computer science; Pathology","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":[0.0004035317,0.0002000219,0.0009914596,0.00160461,0.00005153935,0.00003020141,0.0001058518,0.00005229399,0.000005785471],"category_scores_gemma":[0.001735318,0.0001461094,0.0002727086,0.002232481,0.0003640049,0.0001447877,0.00005589646,0.0002824739,7.93713e-8],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004492663,"about_ca_system_score_gemma":0.00006171882,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003063858,"about_ca_topic_score_gemma":0.000006760029,"domain_scores_codex":[0.9982657,0.00005584627,0.0007020161,0.000386048,0.0003273211,0.0002631148],"domain_scores_gemma":[0.9985783,0.0005151487,0.0002930898,0.0002420199,0.0002314944,0.0001399536],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0004977821,0.0002672922,0.8978922,0.000532412,0.0008676613,0.0001267244,0.001404118,0.0001099481,0.0002778381,0.00005530257,0.0004919939,0.09747674],"study_design_scores_gemma":[0.001957715,0.0001330884,0.9618769,0.001202847,0.001134294,0.000002184359,0.0009341257,0.03204891,0.00005089185,0.0001925291,0.0003510211,0.0001154638],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9781195,0.003820266,0.01429688,0.001016999,0.0003221008,0.0006696486,0.00002015321,0.00008308195,0.001651401],"genre_scores_gemma":[0.995105,0.000007813321,0.003671309,0.0005567467,0.00004539567,0.00003398156,0.0004988615,0.000009591183,0.00007127541],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.09736127,"threshold_uncertainty_score":0.5958169,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02703396727880505,"score_gpt":0.3803729274664553,"score_spread":0.3533389601876503,"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."}}