SnoRelief is sound classification algorithm to detect and classify snoring sounds and sleep apnoea
SnoRelief is a sound classification algorithm to detect and classify snoring sounds and sleep apnoea by utilising a group of deep learning algorithms. The algorithms will classify sleep apnoea completely remotely without the patient needing to attach/wear any sensors, as it uses only audio data to make decisions via a patients Smartphone.
Find your nearest AHSN
Sleep apnoea is when your breathing stops and starts while you sleep. The most common type is called Obstructive Sleep Apnoea (OSA). Around two million people in the UK have moderate or severe symptoms related to OSA This can happen hundreds of times a night, leaving them exhausted, with poor memory and concentration and a significantly higher risk of accidents.
Currently people with OSA are having to wait up to two years for appropriate treatment. Evidence suggests this is a direct result of referral process times, limited availability of sleep clinics, reading the output measurements, the time needed for analysis, appropriate diagnosis and receivership of the appropriate treatment.
The proposed solution, SnoRelief, contains a set of sound classification algorithms that can detect and analyse the snoring sounds of the sleeper via their mobile phone placed beside their bed.
Impact & Outcomes
SnoRelief can provide more details about the sleep apnoea condition type and levels (Mild, Moderate, Severe) enabling a more complex analysis that can just be reviewed by doctors to make a decision, with more accurate data and results in turn shortening the patient pathway and reducing associated costs. Quicker treatment will save time and money for the health industry, with accurate results being provided by one machine rather than doctors needing to analyse and read a lengthy report. Sleep clinics will aslo save time by the patient not having to sleep over in a clinic for 2-3 days to get the best results.
There is potential to deploy the smart sound algorithm on existing solutions within the market rather than seeking approval for a new product, integrating the SnoRelief system with existing technology to enhance performance and accuracy for recommending treatments.