{"id":"W3013605954","doi":"10.2196/17984","title":"Clinical Text Data in Machine Learning: Systematic Review","year":2020,"lang":"en","type":"review","venue":"JMIR Medical Informatics","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":367,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Engineering and Physical Sciences Research Council; Medical Research Council","keywords":"Machine learning; Computer science; Artificial intelligence; Natural language processing; Bottleneck; Information retrieval","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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.05967357,0.00202138,0.01092943,0.02417236,0.001078597,0.004510407,0.004265131,0.003326547,0.007681015],"category_scores_gemma":[0.3079533,0.001481128,0.008695208,0.02310254,0.002923258,0.007302394,0.003296672,0.003313207,0.0008845891],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.005675223,"about_ca_system_score_gemma":0.02431548,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005426588,"about_ca_topic_score_gemma":0.01162474,"domain_scores_codex":[0.9263734,0.0416485,0.01967826,0.003269091,0.008444362,0.0005863397],"domain_scores_gemma":[0.4928944,0.4584301,0.02875779,0.006347869,0.01266152,0.0009083095],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"systematic_review","study_design_gemma":"systematic_review","study_design_scores_codex":[0.0001058877,0.00001817831,0.0006199081,0.9334777,0.007581423,0.00007662541,0.0002100537,0.000327576,0.00005066896,0.0007061054,0.00347226,0.05335358],"study_design_scores_gemma":[0.0001619011,0.0001113812,0.001366155,0.9576055,0.01973278,0.0001575859,0.0002202327,0.0003660689,0.0001799262,0.002276975,0.01776934,0.00005216907],"study_design_candidate":"systematic_review","study_design_consensus":"systematic_review","genre_codex":"review","genre_gemma":"review","genre_scores_codex":[0.0005650302,0.9930172,0.002097965,0.001546089,0.0002800988,0.0009421079,0.001077667,0.00004913128,0.0004246494],"genre_scores_gemma":[0.02296181,0.9599071,0.008885507,0.002345235,0.0003974547,0.004025235,0.001278671,0.00004781226,0.0001511628],"genre_candidate":"review","genre_consensus":"review","teacher_disagreement_score":0.9403265,"threshold_uncertainty_score":0.3155878,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4199594948799194,"score_gpt":0.57648193156965,"score_spread":0.1565224366897306,"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."}}