EUVIMEDEUVIMED entdecken
Crossref

Expert-Guided Large Language Models for Clinical Decision Support in Precision Oncology

Jacqueline Lammert iD, Tobias Dreyer iD, Sonja Mathes, Leonid Kuligin, Kai J Borm, Ulrich A Schatz iD, Marion Kiechle iD, Alisa M Lörsch, Johannes Jung iD, Sebastian Lange iD, Nicole Pfarr iD, Anna Durner iD und 7 weitere

JCO Precision Oncology · 2024-10 · (8)

DOI: 10.1200/po-24-00478 ↗

Kurz gefasst

Fachartikel in der Zeitschrift JCO Precision Oncology, 2024-10. Verfasst von Jacqueline Lammert und 18 weiteren Autor:innen. Beteiligte Einrichtung: Department of Gynecology and Center for Hereditary Breast and Ovarian Cancer, Technical University of Munich (TUM), School of Medicine and Health, Kl…. Hauptthemen laut MEDLINE: Precision Medicine, Decision Support Systems, Clinical, Medical Oncology. Die Arbeit wurde bisher 71-mal zitiert (Crossref) und nennt 38 Literaturangaben.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

PURPOSE Rapidly expanding medical literature challenges oncologists seeking targeted cancer therapies. General-purpose large language models (LLMs) lack domain-specific knowledge, limiting their clinical utility. This study introduces the LLM system Medical Evidence Retrieval and Data Integration for Tailored Healthcare (MEREDITH), designed to support treatment recommendations in precision oncology. Built on Google's Gemini Pro LLM, MEREDITH uses retrieval-augmented generation and chain of thought . METHODS We evaluated MEREDITH on 10 publicly available fictional oncology cases with iterative feedback from a molecular tumor board (MTB) at a major German cancer center. Initially limited to PubMed -indexed literature (draft system), MEREDITH was enhanced to incorporate clinical studies on drug response within the specific tumor type, trial databases, drug approval status, and oncologic guidelines. The MTB provided a benchmark with manually curated treatment recommendations and assessed the clinical relevance of LLM-generated options (qualitative assessment). We measured semantic cosine similarity between LLM suggestions and clinician responses (quantitative assessment). RESULTS MEREDITH identified a broader range of treatment options (median 4) compared with MTB experts (median 2). These options included therapies on the basis of preclinical data and combination treatments, expanding the treatment possibilities for consideration by the MTB. This broader approach was achieved by incorporating a curated medical data set that contextualized molecular targetability. Mirroring the approach MTB experts use to evaluate MTB cases improved the LLM's ability to generate relevant suggestions. This is supported by high concordance between LLM suggestions and expert recommendations (94.7% for the enhanced system) and a significant increase in semantic similarity from the draft to the enhanced system (from 0.71 to 0.76, P = .01). CONCLUSION Expert feedback and domain-specific data augment LLM performance. Future research should investigate responsible LLM integration into real-world clinical workflows.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Precision MedicineDecision Support Systems, ClinicalMedical OncologyHumansNeoplasms

Autor:innen und Institutionen

  • Department of Gynecology and Center for Hereditary Breast and Ovarian Cancer, Technical University of Munich (TUM), School of Medicine and Health, Klinikum rechts der Isar, TUM University Hospital, Munich, Germany.
  • Center for Personalized Medicine (ZPM), Technical University of Munich (TUM), School of Medicine and Health, Klinikum rechts der Isar, TUM University Hospital, Munich, Germany.
  • German Cancer Consortium (DKTK), partner site Munich, a partnership between DKFZ and TUM University Hospital, Munich, Germany.
  • EUropean Network for RAre CANcers (EURACAN) Initiative, partner site Munich, Munich, Germany.
  • Department of Dermatology and Allergy Biederstein, Technical University of Munich (TUM), School of Medicine and Health, Klinikum rechts der Isar, TUM University Hospital, Munich, Germany.
  • Institute for History, Theory and Ethics of Medicine, University of Mainz Medical Center, Mainz, Germany.

ORCID-Profile: Jacqueline Lammert, Tobias Dreyer, Ulrich A Schatz, Marion Kiechle, Johannes Jung, Sebastian Lange, Nicole Pfarr, Anna Durner, Christof Winter, Dyke Ferber, Jakob Nikolas Kather, Maximilian Tschochohei

Zitieren

APA

Lammert, J., Dreyer, T., Mathes, S., Kuligin, L., Borm, K. J., Schatz, U. A., Kiechle, M., Lörsch, A. M., Jung, J., Lange, S., Pfarr, N., Durner, A., Schwamborn, K., Winter, C., Ferber, D., Kather, J. N., Mogler, C., Illert, A. L., & Tschochohei, M. (2024). Expert-Guided Large Language Models for Clinical Decision Support in Precision Oncology. JCO Precision Oncology. https://doi.org/10.1200/po-24-00478

Vancouver

Lammert J, Dreyer T, Mathes S, Kuligin L, Borm KJ, Schatz UA, et al. Expert-Guided Large Language Models for Clinical Decision Support in Precision Oncology. JCO Precision Oncology. 2024. doi:10.1200/po-24-00478

BibTeX
@article{Lammert2024,
  author = {Lammert, Jacqueline and Dreyer, Tobias and Mathes, Sonja and Kuligin, Leonid and Borm, Kai J and Schatz, Ulrich A and Kiechle, Marion and Lörsch, Alisa M and Jung, Johannes and Lange, Sebastian and Pfarr, Nicole and Durner, Anna and Schwamborn, Kristina and Winter, Christof and Ferber, Dyke and Kather, Jakob Nikolas and Mogler, Carolin and Illert, Anna L and Tschochohei, Maximilian},
  title = {Expert-Guided Large Language Models for Clinical Decision Support in Precision Oncology},
  journal = {JCO Precision Oncology},
  year = {2024},
  number = {8},
  doi = {10.1200/po-24-00478}
}
Kostenlos starten →