EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Med-KAG: Preliminary Results of a Medical Knowledge-Augmented Generation Approach

Edouard Haddag, Gabriel H A Medeiros, Lina F Soualmia

Studies in Health Technology and Informatics · 21.05.2026

DOI: 10.3233/shti260261 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Studies in Health Technology and Informatics, erschienen am 21.05.2026. Verfasst von Edouard Haddag und 2 weiteren Autor:innen. Beteiligte Einrichtung: Univ Rouen Normandie, INSA Rouen Normandie, Normandie Univ, LITIS UR 4108, FR-76000 Rouen, France. Hauptthemen laut MEDLINE: Decision Support Systems, Clinical, Unified Medical Language System, Artificial Intelligence, Information Storage and Retrieval. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

We propose in this paper Med-KAG, a novel Artificial Intelligence (AI) assistant architecture for clinical decision support. Med-KAG extends the Retrieval-Augmented Generation (RAG) paradigm by integrating a medical Knowledge Graph built from the Unified Medical Language System (UMLS) Metathesaurus. This approach grounds the model’s responses in verified biomedical relationships between diseases, symptoms, and treatments, reducing hallucinations and improving transparency. Our preliminary evaluation on MedQA-US compares a baseline large language model (Qwen3-235B-A22B) with its knowledge-enhanced variant, showing comparable accuracy while identifying the retriever as the primary source of error. These results highlight both the potential and current limitations of combining structured medical knowledge with generative AI to achieve more reliable and explainable clinical assistance.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Decision Support Systems, ClinicalUnified Medical Language SystemArtificial IntelligenceInformation Storage and RetrievalNatural Language ProcessingGenerative Artificial IntelligenceLarge Language ModelsHumans

Stichwörter der Autor:innen

Artificial Intelligence in MedicineKnowledge GraphRetrieval-Augmented Generation

Autor:innen und Institutionen

  • Univ Rouen Normandie, INSA Rouen Normandie, Normandie Univ, LITIS UR 4108, FR-76000 Rouen, France.

Zitieren

APA

Haddag, E., Medeiros, G. H. A., & Soualmia, L. F. (2026). Med-KAG: Preliminary Results of a Medical Knowledge-Augmented Generation Approach. Studies in Health Technology and Informatics. https://doi.org/10.3233/shti260261

Vancouver

Haddag E, Medeiros GHA, Soualmia LF. Med-KAG: Preliminary Results of a Medical Knowledge-Augmented Generation Approach. Studies in Health Technology and Informatics. 2026. doi:10.3233/shti260261

BibTeX
@article{Haddag2026,
  author = {Haddag, Edouard and Medeiros, Gabriel H A and Soualmia, Lina F},
  title = {Med-KAG: Preliminary Results of a Medical Knowledge-Augmented Generation Approach},
  journal = {Studies in Health Technology and Informatics},
  year = {2026},
  doi = {10.3233/shti260261}
}
Kostenlos starten →