by Jan Provoost, IMEC
Executive overview
A prototype headset, combining wireless and low-power electronics, was recently validated for sleep staging. The device is an R&D product of IMEC (Belgium) and Holst Centre (Netherlands). It shows the future direction for body monitoring for medical and recreational use: low cost, extremely low energy, very long autonomy, and comfortable to wear.
February 10, 2010 – Sleep monitoring is a medical technique to assess a person’s sleep quality and to uncover sleep disorders. Such disorders, if left untreated, are a major health risk. It is estimated that 10% of the U.S. population have sleep apnea, resulting in up to 38,000 deaths per year. Worldwide, 1 billion people suffer from chronic nasal congestion during their sleep, which lowers their quality of sleep and puts their health at risk.
The economic challenge
Sleep monitoring, as it is currently done, is expensive and cumbersome. The test — called polysomnography (PSG) — measures various body parameters, each requiring their own instrumentation. As a subset of the complete PSG test, sleep staging, for example, requires running an EEG (electroencephalogram) to monitor the brain activity, an EOG (electro-oculogram) to monitor the eye activity, and an EMG (electromyogram) to monitor the chin muscle activity. All these parameters must be obtained at the same time, and continuously during a night’s sleep. The entire PSG test requires measuring even more parameters.
Julien Penders, R&D program manager at IMEC sums up the complications as follows. "Typically, a polysomnography test is performed in sleep laboratories at hospitals, and the number of people that can be tested is limited by the availability of the equipment and qualified personnel." He further notes that, the current sleep studies are not representative of people’s normal sleep. Rather, they are done in a hospital, which is unnatural at best, and for some, even intimidating. Several companies have already introduced portable systems, allowing monitoring in the home environment. But these systems still have important issues, Penders observes, those being size, weight, and limited connectivity. "Moreover, the patients often misplace the electrodes. And for these portable solutions, as with standard PSG equipment, people have to try and sleep packed in sensors and wires, which is really uncomfortable."
Go miniature and wireless
The solution, says Penders, is to go micro-sized and wireless: "What you need is a small, low power, well-integrated system that can be set up by everyone, and that allows home monitoring. That would make sleep diagnosis accessible for a lot more people. And it would give better, more natural results."
To prove this point, IMEC has integrated the state-of-the-art of its sensor R&D into a comfortable sleep stage monitoring prototype. The monitor is a lightweight, wearable, miniaturized headset. On the headset are three sensor nodes measuring two EEG-channels, two EOG-channels, and one EMG-channel. These five signals provide the minimum required information for sleep staging according to the Rechtschaffen and Kales standard.
Figure 1. Sleep monitoring headset. |
Each sensor node measures only 20 × 60 × 8mm3. Each of them includes two ultra-low power biopotential read-out ASICs to amplify and filter the captured signals and a wireless radio to send the data to a recording computer. Unlike traditional systems, this monitor requires no additional wires from the head to the body or from the head to the recording device, making it comfortable to wear. The nodes consume only 5mA, allowing the 12 hours autonomy that are needed for sleep monitoring.
"These sensor nodes have several exceptional characteristics, but there is one that really stands out," explains Rudy Lauwereins, professor of electrical engineering at the University of Leuven (Belgium). "It’s the exceptional signal conditioning and amplification integrated into the biopotential read-out ASIC." He notes that, for the EEG, the monitor needs to pick up a brain signal of only 60μV. "But you somehow have to filter that 60μV signal from a 1V noise, caused by muscle movements." According to Lauwereins, this can be accomplished by ensuring that two electrodes see the muscle movement as a common signal and the brain signal as a differential signal. He says that, if the amplifier has a common mode rejection ratio (CMRR) of ~120dB, you can extract the much smaller brain signals. However, to obtain this high CMRR, instrumentation amplifiers are needed, and these are traditionally very bulky. "This ASIC integrates this functionality on one IC. A chip that is low-power, moreover, and can handle many channels in parallel."
The portable sleep monitoring system was recently validated in the sleep laboratory at the University Hospital Center in Charleroi, (Belgium), against a commercially available reference system. This validation proves, in general, that wireless headsets could replace current monitoring systems, for the purpose of monitoring sleep stages. It also shows that this particular prototype is mature and ready for product development.
The future of body monitoring
The portable and comfortable home-monitor for sleep is only one example of what will be possible with the next wave of electronics. Electronics combining powerful ICs and sensors in ultra-small packages, flexible or even stretchable, with wireless radios included, and using only ultralow amounts of power will be achievable.
Figure 2. Flexible wireless monitoring node. |
In healthcare, for example, it is thought that this kind of sensor will become widespread in the years to come. They will make healthcare cheaper, replacing often expensive and unwieldy instruments. Additionally, they will enable continuous home monitoring, a major demand for an ageing population, and they will add monitoring possibilities that are now near impossible.
Stephan Claes, professor of psychiatry at the Leuven University (Belgium) explains that researchers have a good understanding of the physiological reactions to stress, and they can measure them. "But we have to do this in the hospital, which is unnatural, and which allows us only to do one measurement, or a few at most," he observes. "That is simply not enough to get a good picture of how someone feels during the day." He explains that, with the sensors we’ll be able to follow patients for a longer time, say a few days. "And we will do that in their natural environment, with their family or colleagues, giving us access to their complete day and night stress rhythm, something we’re not able to measure today."
IMEC’s team at Holst Centre has long been working on the base technology to enable body area networks (BANs). Such BANs are networks of sensors that span the body and that communicate with each other and with the environment. "Our ambition is to create autonomous sensors that tap their energy from the sun or from the temperature gradients between the body and the sensors," says Penders. "So these electronics need to be extremely low-power, running on tens to hundreds of microwatts." IMEC is also working on scavengers that deliver that amount of energy. To demonstrate what is possible and what the future holds in store, IMEC has integrated sensors and scavengers in prototype products. Examples include a T-shirt with integrated ECG monitor that runs on solar cells integrated in the T-shirt’s fabric, and a two-channel EEG headband with integrated thermoelectric converter for human body heat and Si photovoltaic cells.
Recreational monitoring
Next to the obvious medical applications, body sensors will also be applied recreationally. Think of e-learning, for example, where the learning program is adapted to the attention span of the students. Or gaming, where you could use brain waves measured by a portable EEG monitor to steer the game.
Recently, the recreational possibilities were visualized in a work of art. With ‘Steel Sky’ (Staalhemel), Christophe De Boeck, a Belgian artist, created an environment where you can listen to what is going on in your brain. The setting is a large room with steel plates attached to the ceiling. As people walk below the steel plates, their brainwaves are captured by a miniaturized, wireless 8-channel EEG monitor (Fig. 3). These signals activate hammers that tick on the steel plates above their head. It’s an artful rendering of an affective environment, with objects that react to the moods and thoughts of people. All this made possible with advanced, unobtrusive electronics.
Figure 3. Participant at the Staalhemel exhibit. |
Conclusion
Today’s developments in electronics enable fabricating comfortable monitor packages for medical and recreational use. ICs and sensors can be combined in small packages, flexible or even stretchable, with wireless radios included, and using only ultralow amounts of power. At IMEC, researchers are working to develop the necessary components and integrate them with matching embedded software to create invisible, comfortable body monitors. As a technology demonstrator, they have built a sleep monitor; and they have proven that it has an equivalent functionality of more bulky commercial systems used in hospitals.
Biography
Jan Provoost received his Master’s degree in Languages in 1989 and his Master’s degree in Information Sciences in 1993, both at the U. of Leuven Belgium. He is a science writer at IMEC, Kapeldreef 75, B-3001 Heverlee, Belgium; e-mail [email protected].