Obstructive sleep apnea (OSA) is a prevalent disorder causing airway collapse during sleep, mainly affecting obese men but also impacting various demographics. A 2019 study estimated 936 million adults aged 30–69 worldwide have mild to severe OSA, with China having the highest prevalence. Most cases remain undiagnosed.
Our study of 66 PSG and voice data sets found that OSA patients exhibit more high-frequency sound features, enhancing detection accuracy from 80.3% to 84.85%. The results suggest these features improve screening potential, though a larger sample size is needed for validation. The results were published in [1] and [2].
The application of the project, title of “Acoustic-based automated obstructive sleep apnea detection method”, for RGC Faculty Development Scheme (FDS) 2025/26 proposed by a research team consists of staff in the Department of Computer Science of our college, United Christian Hospital (UCH) and The university of Hong Kong (HKU) is successful. The total grant awarded is HK$ 1,022,981.
The proposed project is aimed to look for optimal acoustic feature set for OSA detection with the following objectives:
- Construct a dataset of 600 PSG tests paired with high-fidelity sound recordings.
- Identify the best feature set using digital signal processing and artificial intelligence for automated OSA detection.
- Develop user-friendly mobile applications for acoustics-based OSA detection.
Effective OSA detection systems rely on high-quality training materials. Our target sample size is 600 from the United Christian Hospital (UCH) and University of Hong Kong (HKU) along with collected 66 speech and snoring recordings paired with PSG from patients at HKU Dentistry and Queen Mary Hospital (QMH). HKU Dentistry ranked among the world’s top dental schools starting from 2018, ensures high-quality patient records. United Christian Hospital, established in 1973, serves East Kowloon and sees over 600 OSA patients annually. These ensures the quality of data collection and ground truth development.
The Department of Computer Science at Hong Kong Chu Hai College will do its best to conduct impactful research with team members to detect OSA in the community!
The research team members
Hong Kong Chu Hai College
Dr Richard Tai-Chiu Hsung, Associate Professor in the Department of Computer Science
Dr Harris Sik-Ho Tsang, Assistant Professor in the Department of Computer Science
Prof. Wai-Lun Lo, Head and Professor in the Department of Computer Science
United Christian Hospital
Dr Winnie Wing Shan Choi, Dental Consultant in Dentistry & Maxillofacial Surgery
Dr Bonnie Fong Yee Lam, Resident specialist in ENT
The University of Hong Kong
Dr Teddy Weifa Yang, Clinical Assistant Professor in Faculty of Dentistry
Reference
[1]Kang-Gao Pang, Tai-Chiu Hsung, Guozhao Liao,Wing-Kuen Ling, Alex Ka-Wing Law and Winnie W.S. Choi, “Obstructive Sleep apnea Detection using Speech signals with High Frequency Components,” Journal of Communications, vol. 17, no. 1, pp. 49-55, January 2022. Doi: 10.12720/jcm.17.1.49-55.
[2]G. Pang, T. -C. Hsung, A. K. -W. Law and W. W. S. Choi, “Optimal vowels measurements for Obstructive Sleep Apnea Detection Using Speech Signals,” 2020 IEEE 3rd International Conference on Information Communication and Signal Processing (ICICSP), Shanghai, China, 2020, pp. 143-147, doi: 10.1109/ICICSP50920.2020.9231972.
