Improving Meditron for Medical Coding Through Fine-Tuning: A Comparative Evaluation Against GPT-4
Studies in Health Technology and Informatics · 21.05.2026
DOI: 10.3233/shti260356 ↗Kurz gefasst
Vergleichsstudie in der Zeitschrift Studies in Health Technology and Informatics, erschienen am 21.05.2026. Verfasst von Coralie Galland-Decker und 7 weiteren Autor:innen. Beteiligte Einrichtung: Medical informatics, Lausanne University Hospital, Switzerland. Hauptthemen laut MEDLINE: Systematized Nomenclature of Medicine, Clinical Coding, Electronic Health Records. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).
Abstract (Originalsprache)
Medical coding is essential but burdensome. In a prior study, Meditron3-70B showed poor coding accuracy on clinical vignettes. We compared GPT-4 with Meditron-CHUV, a version fine-tuned on diverse clinical tasks at Lausanne University Hospital, for generating ICD-10 (billing) and SNOMED-CT (clinical) admission diagnoses from anamnesis alone. Five blinded clinicians rated 40 synthetic vignettes. GPT-4 was preferred overall, but Meditron-CHUV performed better in adult medicine. Model choice varied by specialty (p < 0.001), not profession. SNOMED-CT codes were systematically invalid despite plausible labels. Results confirm LLM limitations and support hybrid methods that combine retrieval and ontologies. Real-data validation is planned.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Medical informatics, Lausanne University Hospital, Switzerland.
- Biomedical Data Science Center, Lausanne University Hospital, Switzerland.
Zitieren
APA
Galland-Decker, C., Ursenbacher, M., Nunes, C., Bouche, F., Blanchard, M., Carra, G., Raisaro, J. L., & Bastardot, F. (2026). Improving Meditron for Medical Coding Through Fine-Tuning: A Comparative Evaluation Against GPT-4. Studies in Health Technology and Informatics. https://doi.org/10.3233/shti260356
Vancouver
Galland-Decker C, Ursenbacher M, Nunes C, Bouche F, Blanchard M, Carra G, et al. Improving Meditron for Medical Coding Through Fine-Tuning: A Comparative Evaluation Against GPT-4. Studies in Health Technology and Informatics. 2026. doi:10.3233/shti260356
BibTeX
@article{GallandDecker2026,
author = {Galland-Decker, Coralie and Ursenbacher, Muaziza and Nunes, Christophe and Bouche, François and Blanchard, Marie and Carra, Giorgia and Raisaro, Jean Louis and Bastardot, François},
title = {Improving Meditron for Medical Coding Through Fine-Tuning: A Comparative Evaluation Against GPT-4},
journal = {Studies in Health Technology and Informatics},
year = {2026},
doi = {10.3233/shti260356}
}
EUVIMED