The Research problem
Insomnia and obstructive sleep apnea (OSA) are among the most prevalent sleep disorders worldwide. Epidemiological studies estimating that insomnia affects approximately 10–30% of the general population, while OSA affects 3–7%. Comorbidity between these two conditions is common: up to 67% of individuals with insomnia also experience of sleep apnea, a condition known as comorbid insomnia and sleep apnea (COMISA). Despite their high prevalence and significant impact on mental health and quality of life, current diagnostic pathways rely heavily on in-laboratory polysomnography (PSG), which is resource-intensive, associated with long waiting times, and poorly suited for large-scale or longitudinal assessment. The limited availability of reliable and user-friendly home-based EEG technologies further contributes to inaccurate or incomplete sleep evaluation. In this context, AI-powered sleep technologies represent a promising and scalable approach to improving global sleep health and mental wellness.
Research Design
This project will establish a long-term Sleep AI Research Network focused on developing self-operable home EEG devices and AI-based sleep scoring models validated across diverse clinical datasets. A triangular international collaboration among National Cheng Kung University (NCKU), the University of Tsukuba (UT), and the University of Technology Sydney (UTS) brings together complementary expertise in AI, biomedical engineering, and clinical sleep medicine across Taiwan, Japan, and Australia, with a particular focus on the Asia-Pacific region. The project will:
- Establish and sustain triangular international partnership to promote long-term knowledge exchange and collaborative innovation in sleep technology and medicine.
- Design self-operable, home-based EEG sleep monitoring devices capable of acquiring high-quality neural signals suitable for both automated AI analysis and expert manual scoring, offering a practical alternative to conventional in-lab PSG and helping to alleviate diagnostic bottlenecks in sleep centers.
- Develop and validate automated AI-based sleep scoring models using diverse, multi-site clinical datasets, reducing the burden of manual scoring while ensuring explainability, reliability, and clinical trustworthiness.
- Organize a joint international workshop that includes research presentations, technical discussions on AI-powered sleep monitoring and enhancement technologies, and professional PSG scoring training to strengthen cross-institutional capacity building.
This collaborative effort is coordinated by the NCKU Digital Biomedical Research Center to maximize scientific integration, translational relevance, and international impact.
Project Objectives
The overarching objective of this project is to establish a long-term Sleep AI Research Network that delivers scalable and clinically trustworthy solutions for early detection and assessment of major sleep disorders, including insomnia, obstructive sleep apnea, and their comorbidity. Supported by the WUN Research Development Fund, the collaboration enables international knowledge exchange and on-site research visits among Taiwan, Japan, and Australia, creating a robust platform for cross-disciplinary innovation. By uniting technology development, clinical validation, and global collaboration, the network seeks to safeguard sleep health and mental wellness while establishing a sustainable foundation for long-term international research and translational impact.