EEG stands for ElectroEncephaloGraphy. Rolls right off the tongue.
The name, however, already explains what it is - a figure ( gram) of the electrical signals coming from the brain (in greek, enképhalos). Let’s take a few moments to figure out how that works, what an EEG looks like, and what we can learn from it. Perhaps this will make it easier to appreciate just how cool the neuro interface, one of BotX’s latest projects, is.
In today’s short guide, we’ll look at:
Now that we know what an EEG is, how do we get one?
Much in the same way we would attempt to measure any other electrical signal - we use electrodes. A common practice in the medical world, in the context of well-known EKGs - images of the electrical activity of the heart.
To acquire an EEG, we’ll place a good number of electrodes on the head. Often 20, individually attached to the scalp with conductive gel, will do. In research or when we aim for higher resolution, we may use a cap with as many as 500 tiny electrodes.
While it may look a bit daunting, the recording itself is completely non-invasive - the electrodes sit on top of the head.
*Note: An invasive form of this recording exists. Electrocorticography, or intracranial electroencephalography, involves implanting electrodes surgically into specific areas of the brain. This is reserved for very particular medical instances and does not make our subject today.
One electrode will capture electrical signal coming from millions of neurons sitting underneath it. By firing all at the same time, they create a strong enough signal for us to detect. Information from each electrode is then passed on to a computer which processes it and prints out the final recording. The end-product may look something like this:
As confusing and squiggly as it may seem, this is a normal EEG of a person resting.
Each line corresponds to electrodes placed in a particular area of the skull. The first ones, Fp1 and Fp2, for example, show activity in the frontal region of the brain, whereas the lower ones, O1 and O2, show activity in the back of the head - the occipital region.
In time and with practice, doctors and scientists learn how to ‘read’ the squiggly lines and gain information from this monitoring of activity in different brain areas.
Some patterns of brain activity are so consistent that they’ve gained their own names. You are, no doubt, already familiar with them.
Keeping on with the Greek nomenclature, normal brain waves are called alpha, beta, delta, and theta. What sets them apart is their frequency and amplitude- that is, how ‘fast’ and how ‘big’ the waves appear.
A theta rhythm, for example, which is mostly experienced during deep sleep, has a low frequency, of under 3.5 Hz - with big and nicely spread apart waves. On the contrary, the beta rhythm of an awake, focused adult has a high frequency of up to 30 Hz and a much lower amplitude. In between sit the other 2 brain waves: alpha waves, the lazier relative of beta, which depict an awake, but resting brain; and theta, normally observed during shallow sleep states or meditation.
Note: A 5th, gamma wave, has been described too. It has an even higher frequency than beta, making it a bit difficult to measure with current setups. It is sometimes referred to as high beta and has been associated with states of peak concentration and cognitive functioning.
You may start to see a pattern here - the more engaged and focused the brain, the tinier and faster the waves get.
We experience all brainwaves naturally as we go to sleep. When you’re in bed, focusing on the last Tweets of the day, you’re likely in beta rhythm. As you decide to put your phone away and do a couple minutes of meditation, you’ll slowly descend into alpha rhythm, and as meditation becomes light sleep into theta. As your sleep deepens, you’ll eventually reach delta rhythm, then cycle through the sleep stages by periodically switching between theta and delta.
These predictable waves let us know, above all other interpretations, that there are rules and patterns beyond the squiggly lines. It means there is a normal EEG profile, and deviations from that normal which we may be able to detect. It also means that within that normal profile, we can distinguish between broad mind states like asleep, awake or focused.
So when does an EEG recording come in handy?
In contrast to other ways of looking at the brain, like CT scans or MRIs, an EEG does not show the shape and structure of the brain, but instead provides real-time insights into its activity. Sure, the resolution is not astounding, we’re looking at the general activity of a loosely defined portion of the cortex, struggling to get a clean read between artifacts like muscle movement or breathing. However, this ability to witness brain activity at any given time in a non-invasive manner remains very much relevant in both neurology clinics and neuroscience labs.
EEGs are instrumental to diagnosing and monitoring patients with epilepsy. A disease where, by definition, the electrical activity of the brain is impaired for brief periods of time, makes for the perfect candidate to be recorded by electrodes in real-time. Seizures can originate in almost any part of the brain and can have a wide variety of causes. As a consequence, particular seizures leave particular traces on an EEG - helping guide doctors in their diagnoses and even treatments.
As we’ve seen, an EEG can monitor brain activity during sleep, proving useful in patients with sleep disorders. In more severe circumstances, EEGs can also be used to monitor comatose patients, keep track of brain activity in accidents and head injuries, or diagnose brain death in persistently unconscious patients.
In basic research, EEGs are used, often in combination with other methods, to track cortical activity in the hope of figuring out how the human brain works. In simpler experiments, scientists may start out by designing specific tasks, and then monitoring the brain activity of people performing those tasks. Conclusions can be drawn about patterns of brain activity related to specific functions, as well as regions of the brain and even potential networks of neurons that underlie the squiggly lines coming out. Approaches of this kind have been used to study attention, learning, visual processing, memory, motor processing and many other elementary as well as complex functions.
It is astounding how a modern concept may revive a technique as old as the EEG (invented in 1929 by psychiatrist Hans Berger). AI may do just that.
We’ve seen how the EEG gives us real-time insight into how the brain works - a sometimes messy, hard to interpret readout, but nevertheless accessible and effective. We’ve also seen how data coming out of the EEG follows at least some rules, some patterns - even if we can’t always see them.
Machine learning, a technology that feeds on patterns to solve problems and answer questions, may just elevate EEG interpretation to a completely new level.
By feeding sufficient EEG data into an AI, we could try and “teach” it how to distinguish brain states. In time, it may even pick up the pace and leave us behind, detecting subtler changes and patterns that are invisible to the naked eye. By increasing the sensitivity of data interpretation, the computer may go beyond detecting just normal vs abnormal activity, or awake vs asleep.
Ultimately, EEG, this almost 100-year-old technique, could provide the input for a primordial brain-computer interface. Pilot interfaces of this kind are already in use - for example, in prosthetic limbs. Here, the electrical activity of the brain is fed into the computer and deconvoluted until the command to move the animatronic limb is registered and understood. Refining such techniques may forever change the lives of individuals suffering from limb loss or paralysis.
But what else can we detect from an EEG-based interface? Commercial possibilities for the wider public are getting more and more intriguing, with prototype EEG sensors now even included in VR gaming systems. What else is achievable needs to be explored, and this is what BotX’s Neuro-Interface aims to do. Watch a video about it here, and stay tuned for our exciting news and advancements.
R. Douglas Fields - Electric Brain: How the New Science of Brainwaves Reads Minds, Tells Us How We Learn, and Helps Us Change for the Better