By Diagnostic World News Staff
February 20, 2025 | Researchers at the University of Cambridge have developed washable and comfortable ‘smart pajamas’ to help monitor sleep disorders at home. The pajamas negate the need for sticky patches, monitoring equipment, or visiting sleep clinics for specialty tests. The pajamas have printed fabric sensors that can monitor breathing by detecting small movements in the skin.
A common issue of sleep tests is discomfort. Sticky patches attached to the skin are often uncomfortable, which can hinder sleep for patients. And while Home Sleep Apnea Testing (HSAT) systems were developed for less invasive methods, they only target specific conditions (e.g. sleep apnea) and may not address other sleep disorders. Wearable devices, such as smartwatches, are less obstructive, they do not possess the level of accuracy of lab-grade equipment. More sophisticated designs that incorporate physical designs, such as humidity, mechanical, and acoustic sensors, have been proposed to help with the accuracy levels of wearable devices, but these typically increase bulkiness, energy consumption, and—not to mention—discomfort (PNAS, DOI: 10.1073/pnas.2420498122 ).
To combat these obstacles, research lead Luigi Occhipinti, professor from the Cambridge Graphene Centre, and his team built on an earlier work of a smart choker for people with speech impediments. They redesigned the sensors for breath analysis and increased the sensitivity. The sensors were then trained using a “lightweight” AI algorithm and can identify six different sleep states with 98.6% accuracy. It can ignore sleep movements, such as tossing and turning, and can even still detect breathing if fitted loosely around the neck and chest.
These smart pajamas will be useful in helping millions who struggle with disordered or irregular sleep. Sixty percent of adults suffer from poor sleep quality, leading to a loss of 44 to 54 working days on an annual basis (PNAS, DOI: 10.1073/pnas.2420498122 ). Sleep behavior like sleep apnea, mouth breathing, and snoring are major contributors to poor sleep quality, and poor sleep can lead to illnesses including diabetes, cardiovascular disease, and emotional disorders.
The research team also developed a machine learning model called SleepNet, which uses the signals captured by the sensors to identify sleep states such as nasal breathing, mouth breathing, snoring, teeth grinding, central sleep apnea, and obstructive sleep apnea. SleepNet reduces computational complexity and can be run on portable devices without needing to be connected to computers or servers.
The team plans to adapt the sensors for other health conditions or home uses, such as baby monitoring, and are currently working with different patient groups. They are also going to improve the durability of the sensors for long-term use.