Development of a Clinical Speech Corpus with Temporally Aligned Sensitive Health Information Annotations
Authors:
- Hsieh-Chih HsuHHHsieh-Chih Hsu
- Liang-Chun FangLFLiang-Chun Fang
- Tatheer Hussain MirTMTatheer Hussain Mir
- Ching-Tai ChenCCChing-Tai Chen
- Hui-Hsien FengHFHui-Hsien Feng
- Jiunn-Ru LaiJLJiunn-Ru Lai
- Pratham NandyPNPratham Nandy
- Omkar PanchalOPOmkar Panchal
- Wei-Hsiang LiaoWLWei-Hsiang Liao
- You-Zhen TienYTYou-Zhen Tien
- Pin-Zhen ChenPCPin-Zhen Chen
- Yun-Ru LinYLYun-Ru Lin
Preprint
Abstract
Objectives: Publicly available datasets for clinical speech de-identification remain scarce due to privacy constraints and annotation complexity. We address this gap with the 2025 SREDH/AI Cup SHI Speech Corpus, a time-aligned clinical speech dataset for detecting sensitive health information (SHI). Methods: The dataset integrates three sources: the OpenDeID-v2 corpus, the dataset for automated medical transcription and medical dialogues derived from Taiwanese television dramas, combining English medical speech with Mandarin Chinese utterances. All audio was manually annotated with millisecond-level, time-aligned SHI segments using a schema of eight primary categories and 38 subcategories of which 26 were attested. Results: The corpus comprises approximately 20 hours of annotated audio (19.36 hours English, 0.89 hours Chinese) and 7830 SHI entities, split into training (10 hours, 1539 files), validation (5 hours, 775 files) and test (5 hours, 710 files) sets, supporting evaluation of both automatic speech recognition (ASR) and SHI recognition. Interannotator agreement (Fleiss’s kappa) reached 0.907. In the 2025 SREDH/AI Cup competition, top-performing systems achieved a mixed error rate of 0.1147 in ASR and a macro F1 of 0.7103 in SHI recognition. Discussion: SHI category distribution is long-tailed, and English–Chinese data volume is highly imbalanced, underscoring the scarcity of Chinese clinical speech resources and the difficulty of time-aligned annotation in low-resource settings. Conclusion: This corpus approximates dialogue-based clinical scenarios and supports research on medical speech de-identification, offering a unique reference for low-resource clinical speech privacy protection. Key message: What is already known on this topic: - Publicly available clinical speech datasets containing time-aligned sensitive health information (SHI) annotations for clinical speech de-identification research are limited. What this study adds: - The 2025 SREDH AI Cup SHI Speech Corpus introduces a clinically oriented speech dataset with millisecond-level time-aligned SHI annotations covering 26 of the 38 SHI subcategories defined in the annotation schema. - The corpus integrates structured annotation protocols and standardized data processing to support the reproducible benchmarking of speech-based de-identification models. How this study might affect research, practice or policy: - The availability of time-aligned SHI annotations may facilitate research on real-time or streaming de-identification systems beyond conventional, transcription-focused approaches. - This dataset may support the development of privacy-preserving technologies for clinical speech in bilingual settings involving English and low-resource languages.Keywords
Funding Information
- Ministry of Education, ASUSTeK Computer Inc and the National Science and Technology Council
This publication has 17 references indexed in Scilit:
- Benchmarking Recording-Level Clinical Speech De-identification for Privacy-Preserving Clinical AudiomedRxiv, 2026
- Large Language Models for Automatic Deidentification of Sensitive Health Information in Clinical Speech2026
- Large Language Models for Automatic Deidentification of Electronic Health Record Notes2025
- Code-Switching Automatic Speech Recognition for Nursing Record Documentation: System Development and EvaluationJMIR Nursing, 2022
- Open Source MagicData-RAMC: A Rich Annotated Mandarin Conversational(RAMC) Speech Dataset2022
- SLURP: A Spoken Language Understanding Resource Package2020
- AISHELL-1: An open-source Mandarin speech corpus and a speech recognition baseline2017
- Annotating longitudinal clinical narratives for de-identification: The 2014 i2b2/UTHealth corpusJournal of Biomedical Informatics, 2015
- Talking Voices2007
- The Rich Transcription 2006 Spring Meeting Recognition Evaluation2006