Thanks to unobtrusive sensors and artificial intelligence, sleep studies can soon be conducted more efficiently at the patient’s home, with a reduced workload required for analysis.
A Nordic research consortium develops better diagnostics for sleep apnoea, relying on wearable sensors and neural networks.
The University of Eastern Finland leads a Nordic research consortium that seeks to develop new, increasingly accurate and highly automated methods for the diagnostics of sleep apnoea.
Obstructive sleep apnoea is a prevalent nocturnal breathing disorder causing a high burden to the public health care system. When untreated, it increases the risk of cardiovascular diseases, for example. However, the methods currently used to diagnose and assess the severity of sleep apnoea are laborious, expensive and sometimes inaccurate.
“New methods are based on wearable, unobtrusive sensors and modern computational solutions drawing on artificial intelligence and machine learning,” says Professor Juha Töyräs, director of the research consortium.
New automated methods reduce the workload and costs related to manual analysis of nocturnal sleep recordings. These methods are also expected to respond to the growing pressure to move most sleep studies from sleep laboratories to patients’ homes.
“It is now possible to evaluate the severity of sleep apnoea much more precisely, as the most significant risk factors associated with nocturnal breathing cessations and the related physiological burden can be taken into account individually for each patient,” Senior Researcher Timo Leppänen explains.
The research consortium comprises academic partners such as Reykjavik University and Akershus University Hospital, as well as industrial partners such as Nokia Solutions and Networks, CGI Finland, Bittium Biosignals, Screentec and Nukute.
Photo caption: A Nordic research consortium develops better diagnostics for sleep apnoea, relying on wearable sensors and neural networks.