Consumer devices are not up to date
With so many consumer devices out there that use heart rate variability to assess sleep quality, why did researchers need to create and train a new algorithm for sleep monitoring? It would seem For example, think of Apple devices or Garmin sports watches. Now van Gilst highlights a major drawback of these devices in the clinical setting.
“Algorithms in consumer devices have been tested using a small group of healthy subjects with no sleep disturbances.In addition, these devices evaluate sleep over long timeframes to conserve battery life. As a result, the data details add to that, the performance and accuracy of the different stages of sleep are very poor,” says Van Girst.
Hundreds of PSG-measured sleep records in combination with data from wearable devices are required to develop reliable algorithms for sleep monitoring devices in clinical settings. This data should come from people of different ages and with all types of sleep disorders. On top of that, a sleep expert would have to manually mark the recordings, and only then could he create an AI-based algorithm. “Omitting any of these steps will produce an inferior algorithm,” says Overeem. “When it comes to monitoring sleep disorders, the effectiveness of hardware is only as good as the algorithms designed to analyze the data it generates.”
next step
The researchers’ new algorithms could potentially be used in wearable devices from a variety of manufacturers, provided the devices’ sensors are of high enough quality.
“Such techniques and algorithms could be a very important addition to the toolkit for assessing sleep disorders, and could be a strong complement to the gold standard approach of PSG,” said van Gilst. said.
With the new algorithm introduced, Overeem is excited about the future possibilities. “We can now start research to really see what these technologies bring to the clinical setting. For example, we can better assess the severity and variability of insomnia. We are assessing long-term sleep patterns in patients with insomnia for the purpose of further tailoring treatment.”
Additionally, Overeem would like to use that data to measure data over a 24-hour period. “In this case, we will have a better understanding of the negative impact of sleep disruption on the patient’s daily life.”
