Sleep Pose Estimation to Improve Sleep Apnea Condition

Challenge Completed!

Globally, more than 1 billion people suffer from sleep apnea which makes the disease a global challenge. In this high-impact 8-week challenge, 50 AI engineers developed an AI algorithm for sleep pose estimation.

The problem

Sleep apnea is a disorder characterized by repeated interruptions in breathing during sleep. The most common cause of these pauses in respiration is the relaxation of throat muscles, provoking a blockage of the airways (Obstructive Sleep Apnea; OSA). Sleep position has been shown to influence the frequency and severity of OSA [1], and treatment strategies exist specializing in patients with positional sleep apnea.

Sleepiz AG has developed a contactless medical device to monitor a person’s sleep throughout the night. The sensing technology, based on continuous wave radar, measures the person’s movements with sub-millimeter resolution. 

The project outcomes

The team built several machine learning algorithms to detect Periodic Limb Movements from complex Electromyography (EMG) and I/Q radar-based time series data.