{"id":"W4210630844","doi":"10.1101/2022.01.24.22269730","title":"Diagnostic accuracy of chest X-ray interpretation for tuberculosis by three artificial intelligence-based software in a screening use-case: an individual patient meta-analysis of global data","year":2022,"lang":"en","type":"preprint","venue":"medRxiv","topic":"COVID-19 diagnosis using AI","field":"Medicine","cited_by":25,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; McGill University Health Centre","funders":"National Institutes of Health; Lunit; Rijksdienst voor Ondernemend Nederland","keywords":"Receiver operating characteristic; CAD; Medicine; Medical physics; Software; Tuberculosis; Meta-analysis; Artificial intelligence; Computer science; Pathology; Internal medicine","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":["metaresearch","metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.002253689,0.0005781303,0.002582576,0.0009168973,0.00009075227,0.00009782056,0.001138488,0.0003267384,0.0005381942],"category_scores_gemma":[0.02601451,0.0005630719,0.001245007,0.001934433,0.000191819,0.0002858186,0.001575028,0.0006754848,8.448551e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003183827,"about_ca_system_score_gemma":0.0006420545,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005414728,"about_ca_topic_score_gemma":0.00698864,"domain_scores_codex":[0.9940351,0.000593578,0.002158226,0.00156088,0.001233658,0.0004185816],"domain_scores_gemma":[0.9841445,0.01081562,0.001543551,0.002832279,0.0004652109,0.0001988542],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"meta_analysis","study_design_scores_codex":[0.001026752,0.003687185,0.5681452,0.001933865,0.07768175,0.0003020555,0.002202521,0.2787618,0.00008929402,0.00003482987,0.000602613,0.06553213],"study_design_scores_gemma":[0.0007276388,0.001369325,0.133818,0.0006217811,0.4665386,0.00001282166,0.001046642,0.3922072,0.001648661,0.0004912983,0.0004439973,0.001074002],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7541771,0.001066232,0.2195348,0.002225245,0.0001909703,0.002368426,0.02036585,0.00007084566,5.994274e-7],"genre_scores_gemma":[0.9567112,0.00002444471,0.02806832,0.00103595,0.00003667329,0.0009189259,0.01313954,0.00006432457,5.758482e-7],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.4343272,"threshold_uncertainty_score":0.9996821,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2514316592668995,"score_gpt":0.4037660877401339,"score_spread":0.1523344284732344,"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."}}