Sleep apnea is a prevalent nocturnal breathing disorder and causes a high burden to the affected individuals, as well as to society and economy. However, current diagnostic methods for sleep apnea are complicated and labour-intensive processes involving various shortcomings related to measuring, analysing and reporting of data.
In this research project we introduce novel diagnostic approaches for sleep apnea utilizing minimally sleep-disturbing wearable sensors and state-of-the art artificial intelligence and machine learning solutions. This responds the growing pressure to move most sleep studies from sleep lab to home and to reduce the workload and costs related to manual analysis of nocturnal sleep recordings.
We aim to develop new metrics which correlate better with the disease severity, related daytime symptoms and risk of severe health consequences than the presently used diagnostic parameters. The project is based on large international patient pools collected and analyzed in collaboration with world top level clinicians, physicists and engineers in Norway, Iceland and Finland.
By adapting modern artificial intelligence techniques, we can identify such features from polysomnographic signals that could be used as novel biomarkers to personalize sleep apnea subtypes and severity. The better understanding of sleep apnea subtypes and underlying pathological mechanisms is crucial while developing personalized approaches to disease management.
The outcomes of this project can lead to paradigm shift in the diagnostics and treatment of sleep apnea and have significant positive impact in Nordic health and socioeconomy. Wearable sleep monitoring systems and sophisticated artificial intelligence based solutions could open global scalable business potential for industry partners operating at medical device sector as well as for health centers and sleep clinics by offering telemedicine sleep diagnostic and treatment follow-up services.
