Improving large language models accuracy for aortic stenosis treatment via Heart Team simulation: a prompt design analysis
Prompt design significantly impacts LLM performance in clinical decision-making for severe aortic stenosis. Tree-of-Thought prompting markedly improved accuracy and aligned recommendations with expert decisions, though LLMs tended toward conservative treatment approaches.
Authors
- Pascal Meier (ORCID: https://orcid.org/0000-0001-8909-2180)
- Mario Togni (ORCID: https://orcid.org/0009-0008-4673-3599)
- D. Garín (ORCID: https://orcid.org/0000-0002-3872-3496)
- Diego Fernández Arroyo (ORCID: https://orcid.org/0000-0001-6693-2706)
- Stéphane Cook (ORCID: https://orcid.org/0000-0003-1221-2978)
- Charlie Ferry (ORCID: https://orcid.org/0009-0000-8997-7042)
- Peter Wenaweser
- Serban Puricel
- Wesley Bennar
Institutions
- University of Fribourg (CH)
Publication Details
- Journal
- European Heart Journal - Digital Health
- Published
- 2025-06-16
- DOI
- https://doi.org/10.1093/ehjdh/ztaf068
- Citations
- 2
- Primary Topic
- Cardiac Valve Diseases and Treatments
- Type
- article
- Field-Weighted Citation Impact
- 2.78