{"id":"W3198414654","doi":"10.1503/cmaj.202066","title":"Problems in the deployment of machine-learned models in health care","year":2021,"lang":"en","type":"letter","venue":"Canadian Medical Association Journal","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":64,"is_retracted":false,"has_abstract":true,"ca_institutions":"Mila - Quebec Artificial Intelligence Institute; Artificial Intelligence in Medicine (Canada); University of Toronto; University of Alberta","funders":"","keywords":"Software deployment; Key (lock); Psychology; Computer science; Operations research; Engineering; Computer security; Software engineering","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.04887248,0.001143384,0.000944904,0.001079056,0.001968605,0.005597949,0.005767827,0.01372187,0.008348559],"category_scores_gemma":[0.2299782,0.000980828,0.001059908,0.001450446,0.007417394,0.01270838,0.004857535,0.02285144,0.008035912],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00806121,"about_ca_system_score_gemma":0.00815545,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02161855,"about_ca_topic_score_gemma":0.01737041,"domain_scores_codex":[0.9573901,0.02987699,0.001804541,0.002511121,0.007405664,0.001011602],"domain_scores_gemma":[0.8011752,0.1679492,0.003204179,0.008080218,0.016296,0.003295251],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0002770382,0.0001480709,0.003100332,0.0007030466,0.000117163,0.0005724656,0.000937906,0.01726287,0.0002275624,0.1111189,0.5876557,0.2778791],"study_design_scores_gemma":[0.000261397,0.0002949165,0.001572432,0.002376955,0.00005955024,0.001034545,0.00155099,0.09148962,0.0005939486,0.4721404,0.4284358,0.0001894293],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"commentary","genre_gemma":"commentary","genre_scores_codex":[0.0007044026,0.002799771,0.01751781,0.9710793,0.002824353,0.00003569635,0.00009026995,0.0001949259,0.004753476],"genre_scores_gemma":[0.1587965,0.02049282,0.127697,0.6428267,0.03886309,0.0007794721,0.0006435145,0.0006448568,0.009256135],"genre_candidate":"commentary","genre_consensus":"commentary","teacher_disagreement_score":0.9511275,"threshold_uncertainty_score":0.2584655,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09475926949566549,"score_gpt":0.3684029874290428,"score_spread":0.2736437179333773,"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."}}