{"id":"W4413776629","doi":"10.2196/71137","title":"Development of a Large-Scale Dataset of Chest Computed Tomography Reports in Japanese and a High-Performance Finding Classification Model: Dataset Development and Validation Study","year":2025,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Japan Society for the Promotion of Science","keywords":"Computer science; Artificial intelligence; Machine translation; Natural language processing; Computed tomography; Scale (ratio); Machine learning; Radiology; Medicine; Cartography","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"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":[0.006482177,0.002213814,0.001098445,0.002752851,0.001174987,0.001252204,0.003206058,0.001497932,0.002494968],"category_scores_gemma":[0.01173004,0.0004711515,0.002063404,0.002609645,0.001169587,0.001470991,0.002380515,0.001958324,0.002687708],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002068514,"about_ca_system_score_gemma":0.003742386,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02643948,"about_ca_topic_score_gemma":0.02456818,"domain_scores_codex":[0.9962624,0.001153442,0.0005449789,0.001063574,0.0007291983,0.0002464181],"domain_scores_gemma":[0.9913716,0.002391462,0.0006242158,0.002180244,0.002909311,0.0005231711],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.003117529,0.006560131,0.1238479,0.004296416,0.00176473,0.004324581,0.001658044,0.08705154,0.04516632,0.002170768,0.1847866,0.5352554],"study_design_scores_gemma":[0.002610251,0.004822469,0.2247187,0.0009821648,0.001919934,0.005073595,0.003522964,0.5331422,0.08710533,0.003924141,0.1314802,0.0006980676],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8097718,0.002849886,0.07207155,0.001532689,0.0009007318,0.004423824,0.09157236,0.01083752,0.006039678],"genre_scores_gemma":[0.4329927,0.001002453,0.1370311,0.0007457063,0.0002396556,0.003788913,0.4196061,0.000572202,0.004021056],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02643948,"threshold_uncertainty_score":0.05257118,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09631670015043926,"score_gpt":0.4066055629076236,"score_spread":0.3102888627571844,"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."}}