Released in summer 2017, Dreem, from Rythm is a game-changing wearable that actively improves your sleep. We talk to CEO Hugo Mercier about neurotechnology, big data and AI to find out more.
In 2013, President Obama announced the launch of the BRAIN Initiative, an ambitious, long-term research program aimed at radically improving our understanding of the human brain.
The press hailed this new era of scientific enquiry as the ‘century of the brain’, with talk of discoveries that would be able to produce a ‘revolutionary new dynamic picture of the brain’.
>> Update: Read our exclusive in-depth Dreem 2 review at SleepGadgets.io
It’s easy to understand the excitement behind the project. Neuroscience represents a rapidly evolving, cross-disciplinary frontier, with numerous, varied areas of research – from finding a cure for diseases like Parkinsons’ and Alzheimer’s, to building brain-computer interfaces which turn science fiction into real life.
These examples won’t likely arrive any time soon. However, in another area of brain research – the science of sleep – years of hard work are about to come to fruition.
Dreem: sleep meets neurotechnology
In an increasingly sleepless society, options for diagnosing and treating sleep disorders are extremely limited, unless that is, you’re willing to pay thousands to undergo a clinical sleep study.
Scheduled for release this summer, Rythm’s first consumer product, the Dreem headband is the culmination of years of research and testing. With an impressive scientific pedigree and millions invested in research and engineering, Dreem looks set to be the most impressive hi-tech sleep aid to hit the market yet.
Dreem’s custom-designed EEG (electroencephalography) sensors monitor your brainwaves in real-time whilst you sleep, which according to Rythm co-founder and CEO Hugo Mercier, provides accurate sleep monitoring that far exceeds the capabilities of other consumer sleep trackers.
The big selling point however is not Dreem’s sleep tracking, but its biofeedback system which, based on your brain activity, generates ‘pink noise’ audio tones to enhance the quality of deep sleep.
With an imminent product launch and the prospect of a potentially game-changing sleep technology product, we spoke to CEO Hugo Mercier via email about the ideas and background to Dreem, big data, artificial intelligence and the future of sleep.
Watch this space for more news on the official release date and prices for Dreem
Jeff Mann (JM): Hi Hugo, thanks for taking the time to speak to us. We’ve been covering developments at the forefront of consumer sleep technology for quite a few years now, and of all the products out there, Dreem seems to me to be one of the most exciting and promising pieces of sleep tech I’ve come across. For those not in the know, could you give our readers the ‘elevator pitch’ for the Dreem – what it is, what are the benefits, and what sets it apart from the competition?
Hugo Mercier (HM): Dreem is the only active and complete sleep solution that merges the latest and greatest in neuroscience research and advanced technology. It’s a wearable device that monitors brain activity (EEG), and stimulates the brain with sound to enhance the quality of deep sleep.
On March 2016, we launched a limited series beta release of the first version of Dreem. We had more than 6500 applicants to the program and we delivered the product to the 500 selected among them.
This program has generated more than 20,000 nights that validate our technology and the efficiency of the product on sleep improvement. It makes it one of the largest ecologic sleep study of the history. It sets the stage for a broader consumer launch coming in summer 2017, with a new version of the Dreem headband.This product is going to be much improved in term of size, comfort, easiness to use, but also from a technology and features standpoint.
As far as competitive products are concerned, existing solutions on the market do not work. Medical solutions, like sleeping pills are addictive, invasive with strong side effects on health.
On the other side of the spectrum, consumer solutions, like sleep trackers and wearables, have shown a strong lack of accuracy, due to the simplistic technologies used, and inefficiency in term of sleep improvement, because of their passive nature. As of today, no similar product has ever been developed.
Dreem is safe, non-invasive, and efficient. It is not only the most accurate sleep monitoring solution (tested against the best consumer sleep trackers available on the market), but it is also actively improving sleep via sound stimulations at night.
JM: There’s an incredible amount of innovation in the sleep technology market right now. But there’s also been a lot of criticism with regards to the accuracy and efficacy of consumer sleep tech. Your background in neuroscience gives you an advantage over many other startups. Can you briefly explain your journey from scientist to CEO and how your vision for Rythm might differ from other firms trying to break through in the the sleep technology market?
HM: I started Rythm with the vision of solving the current sleep epidemic, by using recent neuroscientific breakthroughs on sleep, and developing advanced technologies to bring neurosciences to the consumer.
Before founding the company, I was an engineering student at Ecole Polytechnique in Paris, and was fascinated by neuroscience and learning more about the human brain. I started to collaborate with a team of neuroscientists to stimulate the brain at night and promote deep sleep using sound.
A few months after starting our project, we found that our stimulation process had shown great efficiency in the improvement of sleep quality, and also improved cognitive and physical performance at awakening.
This is why I followed my journey in building Rythm and developing the Dreem headband. It is innovative from a neurosciences standpoint, the stimulation process has been patented and publicated in several leading scientific journals, like Nature and Cell. The team structure is quite innovative.
All the technologies have been developed in-house by experts in machine learning, electronics, mechanics, materials, chemistry, mathematics, and neurosciences. Rythm is daily mixing engineering of new hardware and software technologies, with fundamental neurosciences and clinical research, under a single roof.
That is what allowed us to overcome the challenges and differ from other firms in sleep tech market. The technologies and the neuroscience efforts led by Rythm are more than promising to answer the various sleep challenges, by introducing a new kind of a sleep solution for millions who are dealing with poor quality of sleep.
JM: Like the sadly now defunct Zeo headband, Dreem uses EEG (brainwave activity) sensors to determine the wearer’s sleep architecture. However, the vast majority of consumer sleep monitoring devices use metrics such as movement, heart-rate and breathing to attempt to achieve the same goal. What are your views on the usefulness of non-EEG sleep monitoring? In your opinion can techniques such as ballistocardiography, or measuring heart-rate variability be an effective way to determine sleep physiology?
HM: There are different methods of non-EEG sleep monitoring currently being used, but none are as effective as EEG sleep monitoring. For example, the least effective method of non-EEG sleep monitoring is the use of accelerometer-based sleep trackers.
Accelerometers measure movement, which is ineffective in accurately distinguishing between the different stages of sleep because we move the same amount whether we are in light or deep sleep.
Ballistocardiography or measuring heart-rate variability is a more effective way to determine sleep physiology than accelerometer-based sleep trackers because our heart-rate is known to increase and decrease while transitioning through specific sleep stages.
For example, our heart-rate slows down when transitioning from wakefulness to light-sleep, increases slightly during REM, and then becomes the slowest during deep-sleep, so ballistocardiography a better indicator of sleep physiology than accelerometer-based sleep trackers.
However, neither of these processes are as effective as EEG sleep monitoring because EEG allows for the most precise measurements of activity. With EEG, we can measure the exact frequency, amplitude and type of brain-wave that’s responsible for each sleep cycle, and what’s even better is that you can manipulate these frequencies to deepen sleep — this can’t be done with ballistocardiography.
You might like : The Ultimate Guide to Sleep Tracking
We have one of the most accurate EEG dry acquisition line in the world with more than 90% of correlation with medical setups, and we have the most accurate algorithms for automatic sleep analysis and staging, with a 95% accuracy compared to the gold standard (currently done by sleep doctors).
About Zeo, it is true that they used EEG as a medium to monitor sleep. But for an end-user suffering from sleep issues, they need much more than just monitoring. We’ve seen considerable success with our beta program because the product not only monitors with accuracy, but it also uses sound to improve on the overall quality of sleep.
I can’t delve more into the consumer product at this point, but we’re going to have various new features in upcoming product that will help our users in their quest to improve how they sleep. Stay tuned for the latest news on our site: www.dreem.com
JM: Consumer sleep tracking is right now, largely a passive activity – users can view their sleep data, but it’s up to them to change their behaviour and habits. Dreem is very much an active device – using biofeedback to actually alter the brain state of the wearer. Can you explain in layman’s terms what audio bio-feedback is and how it works?
HM: Audio bio-feedback is the process of using sound to alter the frequency and amplitude of your brainwaves. Think of your brain as an organic orchestra. When you fall asleep, the rhythms of its waves slow down. In deep sleep or slow wave sleep (SWS), the tempo reaches their slowest point, at approximately 1 Hertz.
It means that your orchestra is playing for half a second, quiet for the next half a second, and on and on. Dreem’s electrodes are able to record and analyze brainwaves in real time to stimulate sound at precise moments to sync with slow wave activity. The result is sound working with your brain to impact deep sleep to deliver on higher frequency and amplitude than before to improve overall sleep quality.
JM: Dreem is different to a lot of sleep tech products in that it is designed to enhance the deep, or slow wave sleep stage (SWS). Can you explain the rationale behind choosing to focus on SWS rather than say, on REM or the sleep cycle as a whole?
HM: We’ve chosen to focus on deep sleep or slow-wave sleep (SWS) rather than REM or the sleep cycle as a whole because SWS is the stage of sleep that is responsible for some of brain and body’s most important physiological processes.
For example, a disruption in SWS is linked to impaired memory, high blood pressure, obesity, and decreased mental and physical stamina. This is because SWS is the time at which our bodies repair, regulate and recharge for the next day, whereas REM and light-sleep aren’t linked to these same consequences.
JM: Although it’s not yet hit the wider market, you’ve been evaluating the product on your Dreem First program for some time now. You’ve gathered a serious amount of real-world sleep data in that time. To what extent does ‘big data’ play a role in the development of Dreem and your future products, and what are the distinctions between working with a large dataset, compared to smaller scale clinical data?
HM: Working with a large dataset compared to smaller scale clinical data makes all the difference when it comes to diagnosing and treating sleep disorders. If a physician or sleep-lab was trying to diagnose their patient but they only had a small set of data, their baseline for “normal” may be inaccurate.
Having access to a large dataset like Dreem makes diagnosing disorders much easier because you can tell where this patient fits in on a large spectrum of individuals. Utilizing big-data helped with the development of Dreem because we were able to obtain massive amounts of highly accurate sleep data in a short amount of time, and we will use our data to increase the accuracy of future products even more.
To that point, last week I introduced Morpheo, the first open and secure sleep diagnosis AI platform. We’re building this platform in collaboration with École Polytechnique and the Université Paris Descartes.
Morpheo is an open-source and secure initiative to help develop machine learning models for automatic and predictive diagnosis of sleep disorders. The first platform of its kind, Morpheo will explore and provide sleep clinicians with a free online software that will automatically analyze sleep patterns and help improve and speed diagnosis related to sleep. Morpheo will also enable software developers and machine learning experts as they create models for the diagnosis other illnesses.
JM: Rythm is quite a large company now and has had some serious backing from investors. Without giving away your trade secrets, can you tell us your vision for the company and what see as future developments in the sleep technology sector?
HM: About ⅓ of the U.S. population has reported sleep issues. The numbers are similar in Europe. This is one of the most critical epidemics of our generation that needs solutions right away. Our ultimate goal is to solve the sleep epidemic with neurotechnology by providing non-invasive and efficient solutions to improve sleep quality. Dreem aims to become the prefered sleep solution for the consumer in the coming years.
To be completely honest though, I think we’re just getting started. I had the vision of a consumer product that everyone would be able to use and benefit from easily. In a little more than 2 years, my team has developed several disruptive technologies allowing automatic and embedded analysis of the brain activity, in a miniaturized and comfortable headband. And now with projects like Morpheo and other large scale data collaborations with medical doctors, sleep experts, academic researchers, data scientists, web developers and cryptologists – the opportunities for sleep diagnosis and treatment are endless.