Comparing the Accuracy of Deidentification in Japanese Text Using Large Language Models
Studies in Health Technology and Informatics · 2025
DOI: 10.3233/shti251144 ↗Kurz gefasst
Vergleichsstudie in der Zeitschrift Studies in Health Technology and Informatics, 2025. Verfasst von Ayako Yagahara und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Radiological Technology, Hokkaido University of Science. Hauptthemen laut MEDLINE: Natural Language Processing, Data Anonymization, Electronic Health Records, Confidentiality. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).
Abstract (Originalsprache)
The purpose of this study is to evaluate the accuracy of personal information extraction using large language models (LLMs) to develop an deidentification tool for in-house use. We used three LLMs (BERT, GPT3.5 and GPT4o-mini) to target the extraction of personal names, facility names, and place names. As a pilot study, we analyzed 20 Japanese newspaper articles containing these three types of information and assessed the extraction accuracy. The results showed that BERT achieved the highest accuracy, with F1-scores of 0.94 overall, 0.99 for personal names, 0.89 for facility names, and 0.93 for place names. Therefore, BERT is the most useful for the automatic extraction of personal information from clinical data.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Radiological Technology, Hokkaido University of Science.
- Data Science Center, HELIOS, Hokkaido University Hospital.
Zitieren
APA
Yagahara, A., Mori, H., & Nishimoto, N. (2025). Comparing the Accuracy of Deidentification in Japanese Text Using Large Language Models. Studies in Health Technology and Informatics. https://doi.org/10.3233/shti251144
Vancouver
Yagahara A, Mori H, Nishimoto N. Comparing the Accuracy of Deidentification in Japanese Text Using Large Language Models. Studies in Health Technology and Informatics. 2025. doi:10.3233/shti251144
BibTeX
@article{Yagahara2025,
author = {Yagahara, Ayako and Mori, Haluna and Nishimoto, Naoki},
title = {Comparing the Accuracy of Deidentification in Japanese Text Using Large Language Models},
journal = {Studies in Health Technology and Informatics},
year = {2025},
doi = {10.3233/shti251144}
}
EUVIMED