{"id":"W3041695882","doi":"10.1007/s10140-020-01767-4","title":"Can AI outperform a junior resident? Comparison of deep neural network to first-year radiology residents for identification of pneumothorax","year":2020,"lang":"en","type":"article","venue":"Emergency Radiology","topic":"Ultrasound in Clinical Applications","field":"Medicine","cited_by":40,"is_retracted":false,"has_abstract":false,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"RSNA Research and Education Foundation","keywords":"Medicine; Pneumothorax; Radiology; Receiver operating characteristic; Convolutional neural network; Radiography; Identification (biology); Artificial intelligence; Nuclear medicine; Internal medicine; Computer science","routes":{"ca_aff":true,"ca_fund":false,"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":[0.0004742776,0.000203331,0.000945628,0.0001039031,0.0001121111,0.000002980713,0.0004112914,0.0002543607,0.0003812613],"category_scores_gemma":[0.002095272,0.0001981871,0.0002426015,0.0005527072,0.0001783044,0.00004514785,0.00007351933,0.0002884607,0.00003803648],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005228443,"about_ca_system_score_gemma":0.00006331605,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001012766,"about_ca_topic_score_gemma":0.0003109711,"domain_scores_codex":[0.9968783,0.0001371589,0.001828132,0.0005473117,0.0001901992,0.000418935],"domain_scores_gemma":[0.9977648,0.0005280866,0.0005562879,0.0005853073,0.0002793047,0.0002861611],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.001279574,0.0002154162,0.8634178,0.0002704057,0.0002381223,0.000001212762,0.002235922,0.00399982,0.02747317,0.002987021,0.09712044,0.0007611008],"study_design_scores_gemma":[0.002331509,0.004448815,0.95205,0.00004240113,0.0005442946,0.00003947821,0.0006151838,0.01933284,0.004690287,0.005500223,0.009979381,0.0004255999],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9391177,0.0006014068,0.01656536,0.04040953,0.001087353,0.001900335,0.00009714374,0.00006719635,0.0001539813],"genre_scores_gemma":[0.9943655,0.0001482913,0.003239349,0.0009478634,0.0006510433,0.0002319186,0.0002003422,0.00003866414,0.0001770357],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.08863219,"threshold_uncertainty_score":0.8081835,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05407042811151464,"score_gpt":0.378600799569545,"score_spread":0.3245303714580303,"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."}}