Last Updated on January 25, 2024 by pg@petergamma.org
There is a new paper out which is partly funded by Oura Health Oy:
«Selecting a sleep tracker from EEG-based, iteratively improved, low-cost multisensor, and actigraphy-only devices»
https://www.sciencedirect.com/science/article/pii/S235272182300267X
The author of the papers have a conflict of interest since it was partly funded by Oura Health Oy:
Funding
This work was supported by funds from Oura Health Oy, the Yong Loo Lin School of Medicine, Lee Foundation, and the National Medical Research Council Singapore (STaR May2019-001) awarded to Dr. Michael W. L. Chee.
Declaration of conflicts of interest
Oura Health Oy funded the data collection for the evaluation of its new sleep staging algorithm (OSSA 2.0), but the company did not influence the design of the study, analyses, its interpretation or data presentation. All other equipment was contributed by the Sleep and Cognition Laboratory.
And we suppose also Rob ter Horst has a conflict of interest
He prefers devices like the Oura ring, Dreem 2 and Apple watches for sleep tracking for reasons we cannot follow. Did he receive these devices for free in exchange for a review? The Oura ring is $299.00 Is there another scientist than Rob ter Horst who uses the Oura ring for sleep tracking?
And is ther another medical doctor than Dr. Cody Rall who uses the Muse S for sleep tracking?
According to our own litte experiences with smart watches with optical heart rate monitors:
we suppose that these consumer grade devices will hardy be used for clinical and research application. Did also Dr. Cody Rall receive the devices for free and is this also a paid promotional videos?
Which sleep tracker is the best for us personally?
The Muse S EEG device seems still the best of all conumer grade devices. Is this not logical? Are EEG data not generally superior to other methods, especially if combined with other sensors? Accuracies of around 80 percent are very low and remind us of the accuracy of the good old Garmin Forerunner 235 which had an accuracy of about 80 percent of heart rate measurements. Why do we not have better sleep trackers?
A warning to all
Google now owns Fitbit and after all the negative exeriences with Google products we made we would like to warn everybody from Fitbit products who care about their privacy and want to have their own control over their devices.
We suggest to do some experiments with OpenBCI, more EEG channels and the Python sleep software, and adding features such as eye tracking and other sensors to get closer to the accuracy of the gold standard polysomnography:
And we suggest also to test if a combination of multipe device can increase the accuracy of sleep tracking. We suppose also that scientists previously tested all of these options. But what they did not do was to find devices which are optimized for low-cost high-quality. We further miss papers which have no conflict of interest to answer many of the open questions.
Dreem went bankrupt. This is a warning that Raspberry Pi for private users could have the same fate. We mentioned that previously. Dreem has been taken over by another company, which does not offer consumer grade products anymore. And if scientists want to buy these products we suppose they are expensive. Companies who do not publish the costs of devices often do not do this since to publish those is negative for their sales rates, since it allows price comparisons, and it indicates that these companies haver prices which are not very competitive.This makes those also less interesting for the home lab of scientists and medical doctors.
As with consumer grade optical heart rate monitors, we do not know if Rob ter Horsts result and the above paper will ever be confirmed by other scientists. Do these devices have an accuracy which is interesting for scientists and medical doctors? If they are would there not be more papers and studies which use these devices and we do not know of such studies and papers.
Accoding to our own little exerience there is much potential to increase the accuracy of sleep trackers to a level where they also become interesting for the home lab of scientists and medical doctors. But we are not soft- and hardware developers. And until scientists and medical doctors develop such devices which are eventually open source devices we suppose that it will take time.
Meditation has a strong effect on sleep duration. If we meditate more, sleep duration decreases. Sleep which is something which is not under our control is replaced by something which is under our control and that is meditation. But we do not know if these devices shown here are accurate enough to be used for scientific studies about meditation.