SMARTSLEEP: Standardizing Measures and Advancing Reliable Technologies for SLEEP

The project's objective is to develop standardized, transparent algorithms for wearable devices to measure key sleep metrics—total sleep time (TST), sleep onset latency (SOL), and sleep efficiency (SE). It leverages AI/ML techniques to ensure accurate and clinically validated predictions. Wearable technologies address the limitations of traditional tools like polysomnography, which, while effective, are intrusive, costly, and not suitable for real-world contexts. 

This project is critical as sleep disturbances significantly affect societal productivity and health costs. By enabling validated, continuous monitoring of sleep, SMARTSLEEP facilitates proactive healthcare interventions and democratizes access to sleep health technologies. Its outcomes will support clinical integration, allowing for effective treatment of sleep-related conditions and the development of innovative digital health solutions for a healthier society. 

The SMARTSLEEP project has contributed to the scientific field by designing and validating a standardized pipeline for sleep-wake classification using wearable sensor data. The core scientific value lies in the development of a device-agnostic algorithm based on triaxial accelerometer and gyroscope data, with a focus on generalizability and transparency. This pipeline was tested against PSG-derived labels in a challenging population (suspected sleep apnea patients), achieving reasonable performance using only inertial sensor data.