@inproceedings{singh-etal-2025-redactor,
title = "{R}edact{OR}: An {LLM}-Powered Framework for Automatic Clinical Data De-Identification",
author = "Singh, Praphul and
Dzialo, Charlotte and
Kim, Jangwon and
Srivatsa, Sumana and
Bulu, Irfan and
Gadde, Sri and
Kenthapadi, Krishnaram",
editor = "Rehm, Georg and
Li, Yunyao",
booktitle = "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track)",
month = jul,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.acl-industry.36/",
doi = "10.18653/v1/2025.acl-industry.36",
pages = "510--530",
ISBN = "979-8-89176-288-6",
abstract = "Ensuring clinical data privacy while preserving utility is critical for AI-driven healthcare and data analytics. Existing de-identification (De-ID) methods, including rule-based techniques, deep learning models, and large language models (LLMs), often suffer from recall errors, limited generalization, and inefficiencies, limiting their real-world applicability. We propose a fully automated, multi-modal framework, RedactOR for de-identifying structured and unstructured electronic health records, including clinical audio records. Our framework employs cost-efficient De-ID strategies, including intelligent routing, hybrid rule and LLM based approaches, and a two-step audio redaction approach. We present a retrieval-based entity relexicalization approach to ensure consistent substitutions of protected entities, thereby enhancing data coherence for downstream applications. We discuss key design desiderata, de-identification and relexicalization methodology, and modular architecture of RedactOR and its integration with Oracle Health Clinical AI system. Evaluated on the i2b2 2014 De-ID dataset using standard metrics with strict recall, our approach achieves competitive performance while optimizing token usage to reduce LLM costs. Finally, we discuss key lessons and insights from deployment in real-world AI-driven healthcare data pipelines."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="singh-etal-2025-redactor">
<titleInfo>
<title>RedactOR: An LLM-Powered Framework for Automatic Clinical Data De-Identification</title>
</titleInfo>
<name type="personal">
<namePart type="given">Praphul</namePart>
<namePart type="family">Singh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Charlotte</namePart>
<namePart type="family">Dzialo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jangwon</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sumana</namePart>
<namePart type="family">Srivatsa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Irfan</namePart>
<namePart type="family">Bulu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sri</namePart>
<namePart type="family">Gadde</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Krishnaram</namePart>
<namePart type="family">Kenthapadi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Georg</namePart>
<namePart type="family">Rehm</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yunyao</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Vienna, Austria</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-288-6</identifier>
</relatedItem>
<abstract>Ensuring clinical data privacy while preserving utility is critical for AI-driven healthcare and data analytics. Existing de-identification (De-ID) methods, including rule-based techniques, deep learning models, and large language models (LLMs), often suffer from recall errors, limited generalization, and inefficiencies, limiting their real-world applicability. We propose a fully automated, multi-modal framework, RedactOR for de-identifying structured and unstructured electronic health records, including clinical audio records. Our framework employs cost-efficient De-ID strategies, including intelligent routing, hybrid rule and LLM based approaches, and a two-step audio redaction approach. We present a retrieval-based entity relexicalization approach to ensure consistent substitutions of protected entities, thereby enhancing data coherence for downstream applications. We discuss key design desiderata, de-identification and relexicalization methodology, and modular architecture of RedactOR and its integration with Oracle Health Clinical AI system. Evaluated on the i2b2 2014 De-ID dataset using standard metrics with strict recall, our approach achieves competitive performance while optimizing token usage to reduce LLM costs. Finally, we discuss key lessons and insights from deployment in real-world AI-driven healthcare data pipelines.</abstract>
<identifier type="citekey">singh-etal-2025-redactor</identifier>
<identifier type="doi">10.18653/v1/2025.acl-industry.36</identifier>
<location>
<url>https://aclanthology.org/2025.acl-industry.36/</url>
</location>
<part>
<date>2025-07</date>
<extent unit="page">
<start>510</start>
<end>530</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T RedactOR: An LLM-Powered Framework for Automatic Clinical Data De-Identification
%A Singh, Praphul
%A Dzialo, Charlotte
%A Kim, Jangwon
%A Srivatsa, Sumana
%A Bulu, Irfan
%A Gadde, Sri
%A Kenthapadi, Krishnaram
%Y Rehm, Georg
%Y Li, Yunyao
%S Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track)
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-288-6
%F singh-etal-2025-redactor
%X Ensuring clinical data privacy while preserving utility is critical for AI-driven healthcare and data analytics. Existing de-identification (De-ID) methods, including rule-based techniques, deep learning models, and large language models (LLMs), often suffer from recall errors, limited generalization, and inefficiencies, limiting their real-world applicability. We propose a fully automated, multi-modal framework, RedactOR for de-identifying structured and unstructured electronic health records, including clinical audio records. Our framework employs cost-efficient De-ID strategies, including intelligent routing, hybrid rule and LLM based approaches, and a two-step audio redaction approach. We present a retrieval-based entity relexicalization approach to ensure consistent substitutions of protected entities, thereby enhancing data coherence for downstream applications. We discuss key design desiderata, de-identification and relexicalization methodology, and modular architecture of RedactOR and its integration with Oracle Health Clinical AI system. Evaluated on the i2b2 2014 De-ID dataset using standard metrics with strict recall, our approach achieves competitive performance while optimizing token usage to reduce LLM costs. Finally, we discuss key lessons and insights from deployment in real-world AI-driven healthcare data pipelines.
%R 10.18653/v1/2025.acl-industry.36
%U https://aclanthology.org/2025.acl-industry.36/
%U https://doi.org/10.18653/v1/2025.acl-industry.36
%P 510-530
Markdown (Informal)
[RedactOR: An LLM-Powered Framework for Automatic Clinical Data De-Identification](https://aclanthology.org/2025.acl-industry.36/) (Singh et al., ACL 2025)
ACL