Towards Privacy-Compliant Healthcare Data: Developing an Efficient Multilingual Filter for PII and PHI under Safe Harbor Principles
Journal
Proceedings of the 18th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management
ISSN
2184-3228
ISSN-Digital
2184-3228
ISBN
978-989-758-868-6
Type
conference paper
Date Issued
2026-10-28
Author(s)
Abstract
Medical text anonymization is essential for data-driven healthcare but constrained by strict privacy regulations, particularly in multilingual Swiss settings. A lightweight, privacy-compliant filter for removing personally identifiable and protected health information from unstructured medical text is presented. The system combines regular expressions, rule-based logic, and locally-run guided LLMs, where local open-source models are constrained to assign a label to each word. This approach is evaluated on 50 manually annotated SwissGerman medical documents and 520 English i2b2 notes. A regex-only baseline shows substantially lower F1 score and higher leakage on both datasets. Overall, the proposed hybrid approach reduces residual identifiers while maintaining deterministic runtime and modest computational cost, enabling practical medical text anonymization in multilingual contexts.
Language
English (United States)
Keywords
Medical Text Anonymization
PII/PHI Detection
Privacy-Compliant Healthcare Data
HSG Classification
contribution to scientific community
Refereed
Yes
Publisher
SCITEPRESS - Science and Technology Publications
Publisher place
Portugal
Volume
2
Start page
608
End page
615
Pages
8
Event Title
18th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management
Event Location
Angers, France
Event Date
28-30 October 2026
Subject(s)
Division(s)
Contact Email Address
bella.fernandez@unisg.ch
File(s)![Thumbnail Image]()
![Thumbnail Image]()
restricted
Name
152971.pdf
Type
Main Article
Description
Official publication
Size
201.31 KB
Format
Adobe PDF
Checksum (MD5)
e5de93753e76b63f2d6ce71984b31500
open.access
Name
KMIS_2026_63.pdf
Description
Poster shown during the conference
Size
182.06 KB
Format
Adobe PDF
Checksum (MD5)
d04f83f82a5ab242d7fe6a7dcc8ab77a