Abstract
Large language models (LLMs) show considerable potential for atrial fibrillation (AF) management, yet current clinical applications frequently remain suboptimal due to accuracy limitations. To address these limitations, this study developed PULSE (Potentiated User-friendly LLM-driven Search Engine), a novel knowledge-enhanced, domain-aware LLM agent specifically designed to improve AF patient self-management across the entire care continuum. The proposed framework integrates multimodal inputs, meticulously curated clinical knowledge bases, optimized prompt engineering, and retrieval-augmented generation within an agent-based architecture. Performance was rigorously evaluated against four leading base LLMs using response quality (clinical accuracy, content integrity, practical utility, and patient safety) and readability (clarity, conciseness, and empathy). Comprehensive clinical validation was subsequently conducted through blinded expert assessment of 75 real-world AF-related patient queries. The results demonstrated that PULSE improved clinical accuracy, content integrity, utility, and safety (P < 0.05) across all tested models. Furthermore, it substantially enhanced empathy and clarity while maintaining comparable conciseness. Overall, PULSE improves both the factual accuracy and readability of patient-facing medical outputs, highlighting the immense clinical potential of agent-driven LLM systems to advance chronic disease self-management and improve long-term patient outcomes.
Acknowledgements
Clinical and Translational Research Project of Anhui Province (202427b10020086, 202427b10020089, 202427b10020097), Research Funds of Joint Research Center for Regional Diseases of IHM (2024bydjk001, 2024bydjk002, 2024bydjk005), Anhui Provincial Health and Health Commission Scientific Research Project (AHWJ2024Aa10053), The First Affiliated Hospital of Bengbu Medical University for Excellent Young Scholars (2025byyfyyq09) and National Engineering Research Center of Science and Technology Information (2025STI135).
Authors and Affiliations
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Wang, Y., Peng, C., Hu, R. et al. A knowledge-enhanced domain-aware large language model agent for atrial fibrillation management.
npj Digit. Med. (2026). https://doi.org/10.1038/s41746-026-03038-x
-
DOI
:https://doi.org/10.1038/s41746-026-03038-x
