{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001896834,0.0005386478,0.000569591,0.0004768603,0.0003030349,0.0006159677,0.0006516718,0.001015904,0.002436611],"category_scores_gemma":[0.008991324,0.0001341415,0.0004219448,0.0002324697,0.0001994311,0.001107581,0.0006353451,0.001006675,0.0005765153],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005981008,"about_ca_system_score_gemma":0.0008711739,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006223863,"about_ca_topic_score_gemma":0.007314392,"domain_scores_codex":[0.9994544,0.0001953976,0.00003145372,0.0001160786,0.00008944512,0.0001131997],"domain_scores_gemma":[0.9968963,0.001446943,0.0001966603,0.0001361156,0.000758545,0.0005654532],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"nonrandomized_trial","study_design_scores_codex":[0.01926315,0.00521722,0.3281245,0.0008341257,0.000790837,0.0004432183,0.000493341,0.03652153,0.006995411,0.0009884133,0.01530688,0.5850214],"study_design_scores_gemma":[0.0007426467,0.01256766,0.2552942,0.0003843586,0.0009606897,0.0008153929,0.00169376,0.7073405,0.01142122,0.003003755,0.005622671,0.0001530956],"study_design_candidate":"nonrandomized_trial","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9869812,0.001671618,0.005175449,0.001361599,0.0003979434,0.00006490879,0.0003846183,0.0002911183,0.003671653],"genre_scores_gemma":[0.9957424,0.0002797498,0.002333899,0.0002353421,0.00007225569,0.00001751365,0.0004144627,0.000007362255,0.0008968969],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.006223863,"threshold_uncertainty_score":0.01237524,"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."}}