{"id":"W4406486231","doi":"10.1007/978-3-031-88711-6_10","title":"Evaluating LLM Abilities to Understand Tabular Electronic Health Records: A Comprehensive Study of Patient Data Extraction and Retrieval","year":2025,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"","funders":"Fonds de Recherche du Québec - Santé; Grand Équipement National De Calcul Intensif","keywords":"Computer science; Task (project management); Context (archaeology); Data extraction; Information retrieval; Health records; Selection (genetic algorithm); Serialization; Data mining; Machine learning; Feature selection; Data science; Artificial intelligence; MEDLINE; Health care; Engineering","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.008516475,0.0004760027,0.0004248801,0.001979808,0.0003687479,0.00149309,0.0006782379,0.0006257211,0.001866271],"category_scores_gemma":[0.05962696,0.0001984313,0.0005567048,0.001358238,0.000496762,0.003343163,0.001596943,0.0004931279,0.0008145412],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000657863,"about_ca_system_score_gemma":0.0008229708,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002750353,"about_ca_topic_score_gemma":0.003325789,"domain_scores_codex":[0.9961434,0.001911388,0.0006438098,0.0003533539,0.0007918645,0.0001561705],"domain_scores_gemma":[0.8866972,0.1026131,0.004075921,0.002134165,0.003755749,0.0007239036],"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":[0.002181417,0.001066431,0.4622095,0.001156481,0.0004101662,0.0003970671,0.01517938,0.002087375,0.01000237,0.0006402647,0.005089576,0.49958],"study_design_scores_gemma":[0.0001913541,0.004209312,0.8966311,0.0006471233,0.00073971,0.001816448,0.01670441,0.04563121,0.02020321,0.002351392,0.01073703,0.0001377349],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9881312,0.001062826,0.006374262,0.0003288055,0.00001096529,0.0001741028,0.0008482184,0.0002653968,0.002804306],"genre_scores_gemma":[0.9857914,0.0005838479,0.01040184,0.0001346253,0.00001141117,0.0001034838,0.001578722,0.0000496486,0.001344935],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.008516475,"threshold_uncertainty_score":0.04503995,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07135074493848703,"score_gpt":0.3813718943281653,"score_spread":0.3100211493896782,"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."}}