Large Language Model Empowered Privacy-Protected Framework for PHI Annotation in Clinical Notes
Studies in Health Technology and Informatics · 2025
DOI: 10.3233/shti250965 ↗Kurz gefasst
Fachartikel in der Zeitschrift Studies in Health Technology and Informatics, 2025. Verfasst von Guanchen Wu und 9 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Science, Emory University. Hauptthemen laut MEDLINE: Electronic Health Records, Confidentiality, Computer Security, Personally Identifiable Information. Die Arbeit wurde bisher 1-mal zitiert (Crossref). Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).
Abstract (Originalsprache)
De-identifying private information in medical records is crucial to prevent confidentiality breaches. Rule-based and learning-based methods struggle with generalizability and require large annotated datasets, while LLMs offer better language comprehension but face privacy risks and high computational costs. We propose LPPA, an LLM-empowered Privacy-protected PHI Annotation framework, which uses few-shot learning with pre-trained LLMs to generate synthetic clinical notes, reducing the need for extensive datasets. By fine-tuning LLMs locally with synthetic notes, LPPA ensures strong privacy protection and high PHI annotation accuracy. Experiments confirm its effectiveness, efficiency, and scalability.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Computer Science, Emory University.
- Department of Computer Science, University of Chicago.
- School of Computing, University of Georgia.
- Nell Hodgson Woodruff School of Nursing, Emory University.
Zitieren
APA
Wu, G., Zheng, L., Xie, H., Xiang, Z., Lu, J., Liu, D., Bold, D., Li, B., Hu, X., & Yang, C. (2025). Large Language Model Empowered Privacy-Protected Framework for PHI Annotation in Clinical Notes. Studies in Health Technology and Informatics. https://doi.org/10.3233/shti250965
Vancouver
Wu G, Zheng L, Xie H, Xiang Z, Lu J, Liu D, et al. Large Language Model Empowered Privacy-Protected Framework for PHI Annotation in Clinical Notes. Studies in Health Technology and Informatics. 2025. doi:10.3233/shti250965
BibTeX
@article{Wu2025,
author = {Wu, Guanchen and Zheng, Linzhi and Xie, Han and Xiang, Zhen and Lu, Jiaying and Liu, Darren and Bold, Delgersuren and Li, Bo and Hu, Xiao and Yang, Carl},
title = {Large Language Model Empowered Privacy-Protected Framework for PHI Annotation in Clinical Notes},
journal = {Studies in Health Technology and Informatics},
year = {2025},
doi = {10.3233/shti250965}
}
EUVIMED