[Reading Time - 2 minutes 50 seconds]
Your retina, fingerprint, voice, iris, and face are all common forms of biometrics, which are used for computer authentication based on the features and characteristics of the individual. Whereas password logins are based on what you know, biometric logins are based on what you are. The advantage of biometrics is that you always carry your retina or voice with you, and you don't have to memorize anything new or risk losing it.
But now there's research for adding a new biometric feature to that list. It's your brainwaves. And although it holds some promise, there are also some downsides to it, too.
Your brain cells communicate through electrical impulses. These impulses are always active, even when we sleep. These electrical impulses can be read using an electroencephalogram (EEG) machine. This machine can detect electrical activity in your brain using electrodes (small metal discs) that are attached to your scalp; the electrical impulse activity shows up as wavy lines on the EEG device and are recorded. An EEG is one of the main diagnostic tests for epilepsy or other seizure disorders and may also be used for diagnosing brain damage from head injury or inflammation of the brain (encephalitis), and is also used to detect sleep disorders.
Everyone's brain activity reflects the unique pattern of the neural pathways in our brain. And because these are all different, they result in different EEGs for everyone. Thus, unique EEGs could be used as one form of a person's unique identification and could be the basis of computer authentication.
It all begins with a user connected to an EEG machine through electrodes attached to the scalp. The person then closes their eyes as a list of 75 acronyms (like "FBI" or "DVD") is read to them. Why listen to these acronyms instead of just recording the brain's electrical signals? It's because just measuring the brain's signals produces too much extraneous "noise" that makes it difficult to find a unique pattern.
By listening to acronyms the EEG machine can see brainwaves that move from just one region of the brain to another, specifically to the region associated with the task of reading and recognizing words. Known as "semantic memories" ("episodic memories" record our past experiences) these are collection of meanings that we associate with words and subtly differ from person to person, providing an individual pattern. And unlike episodic memories, semantic memories do not change too much over time. If you are bitten by a dog, the episodic memory neurons that fire when you next read the word “dog” are likely to be different, but the semantic memory neurons are thought to behave roughly the same as before. The end result is a much clearer brainwave signal that can be measured more accurately and quickly.
These EEGs are recorded during the readings, forming a baseline EEG. When a user wants to log in to their computer, they put on a cap with electrodes that reads their EEG and then compares it to the baseline EEG. If the EEGs match up the users is logged in.
How accurate is using brainwaves for authentication? In one study there was a 94 percent accuracy rate.
Sounds great, right? Well, there are a few drawbacks.
First, a 94 percent accuracy rate is not quite good enough for computer authentication. Also, a user would have to don a helmet with electrodes whenever she wanted to get into her computer. That's not much of a fashion statement; in fact, there's nothing that says nerd quite like wearing in the office a wired helmet with sensors glued to your scalp. And there's also the problem of identifying the brainwaves when the user does not have the acronyms being read to them (conceivably the user could read through a list of acronyms in an attempt to recreate those same brainwaves).
But an even more significant drawback is that the EEGs can change during or immediately following certain situations. Stress can change an EEG, as can overall alertness. Physical exercise can likewise alter an EEG (although it bounces back to normal after about one minute). And EEGs can also be impacted by opioids, caffeine, and even alcohol.
So, are we likely anytime soon to log in to our computers using EEG? Probably not. But they do hold promise if we want to start preventing someone who has had too much to drink to start typing on their computer!
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.