{"id":"W4205096618","doi":"10.2196/25157","title":"Assessment of Natural Language Processing Methods for Ascertaining the Expanded Disability Status Scale Score From the Electronic Health Records of Patients With Multiple Sclerosis: Algorithm Development and Validation Study","year":2022,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Multiple Sclerosis Research Studies","field":"Medicine","cited_by":15,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto; Regent Park Community Health Centre; St. Michael's Hospital","funders":"St. Michael's Hospital Foundation; Li Ka Shing Foundation","keywords":"Expanded Disability Status Scale; F1 score; Recall; Artificial intelligence; Health records; Computer science; Convolutional neural network; Precision and recall; Standard score; Machine learning; Data mining; Medicine; Natural language processing; Algorithm; Multiple sclerosis; Health care; Psychology","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":true,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002896152,0.0001431557,0.0004244412,0.00003858968,0.0004635857,0.00002090436,0.0001914745,0.00002920673,0.00002028213],"category_scores_gemma":[0.0006066694,0.00007283083,0.00004163857,0.0002691389,0.0002639959,0.0001098161,0.0003075621,0.0005021369,4.014563e-8],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003659771,"about_ca_system_score_gemma":0.0009066524,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002181938,"about_ca_topic_score_gemma":0.0001917466,"domain_scores_codex":[0.996979,0.0003756629,0.0007968319,0.000141308,0.001333454,0.0003737468],"domain_scores_gemma":[0.9978436,0.001110662,0.000451642,0.0002614761,0.0002093758,0.0001232598],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.0001471645,0.0005148891,0.3594552,0.0002312418,0.0001147791,3.629568e-8,0.05113341,0.000005153518,0.000009170936,7.305633e-7,0.00002190661,0.5883663],"study_design_scores_gemma":[0.003923908,0.001021418,0.8445719,0.0002041541,0.00003579859,3.480078e-7,0.09811164,0.05178891,0.0001336157,0.000003001667,0.0001316783,0.00007365279],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9860631,0.0003233741,0.00996394,0.0002648805,0.00003976109,0.003256793,0.00006296697,0.0000183332,0.00000684629],"genre_scores_gemma":[0.9364352,0.00003077726,0.06224407,0.0001464689,0.00002132425,0.0008939272,0.0002141493,0.00001143777,0.000002649834],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5882927,"threshold_uncertainty_score":0.3565573,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05208524982829123,"score_gpt":0.4109838042878863,"score_spread":0.3588985544595951,"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."}}