{"id":"W4415404211","doi":"10.2196/70708","title":"Development and Validation of a Generative Artificial Intelligence-Based Pipeline for Automated Clinical Data Extraction From Electronic Health Records: Technical Implementation Study","year":2025,"lang":"en","type":"article","venue":"JMIR Bioinformatics and Biotechnology","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Pipeline (software); Data extraction; Data integration; Timeline; Information extraction; Extraction (chemistry); Health data; Data modeling","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001015625,0.0001167389,0.000315441,0.0002782485,0.0001444954,0.00002144236,0.00009446921,0.0002477416,0.000006456889],"category_scores_gemma":[0.0001968172,0.0001015395,0.00002204898,0.000288642,0.0001061168,0.00009946928,0.00006936606,0.0002134403,0.000001327687],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001080857,"about_ca_system_score_gemma":0.0008364625,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002586582,"about_ca_topic_score_gemma":0.001077461,"domain_scores_codex":[0.9980175,0.00004452701,0.001376105,0.0002571819,0.0001006849,0.0002040021],"domain_scores_gemma":[0.9989309,0.0001894219,0.0003852117,0.000301949,0.0001430628,0.00004949502],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0002485713,0.0004552431,0.002180989,0.0001414136,0.00004465543,1.689552e-7,0.0006284423,0.000003825452,0.0005861261,0.001363472,0.0003849395,0.9939622],"study_design_scores_gemma":[0.0007182014,0.004886345,0.00660089,0.0003019942,0.0002155315,0.00001019644,0.04255957,0.6919808,0.2376149,0.006062598,0.008693864,0.0003551915],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7311754,0.0001147941,0.2574246,0.008629858,0.0001653464,0.002309155,0.00004017744,0.0001349239,0.000005678565],"genre_scores_gemma":[0.9183895,0.0002501062,0.07983849,0.0003320263,0.00004934056,0.0001302413,0.001000348,0.000006082346,0.000003887967],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.993607,"threshold_uncertainty_score":0.4140659,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2218310790735947,"score_gpt":0.5287346243886004,"score_spread":0.3069035453150057,"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."}}