Scientists Can Now Identify Individuals Based on Brain Waves—And It’s 100% Accurate

Brainprints

For the longest time, the world has relied on fingerprints as a way to easily identify individuals. This is because a fingerprint is unique (and permanent) for each person—so what happens when someone’s fingerprint is stolen? This may sound a little farfetched, but it is easy enough to take a fingerprint from another, you just need to swipe a glass that they held.

And of course, you can’t grow a new finger to replace the print. This means that, when it gets compromised, it’s compromised forever. But now, researchers from Binghamton University have found a more efficient way to identify people—using brain waves.

Because people react to various stimuli—such as foods, or words, or celebrities—differently, it means people’s brain waves will also exhibit varying patterns when these things are presented to them. No two people will show the same brain pattern when presented an image of, say, a dog.

This led researchers to look into how this can be applied to security and identification. They tested the brain activity of 50 individuals who wore electroencephalogram devices as they were shown 500 images (which included images of pizza, Anne Hathaway, a boat, and words).

The result? Significant discrepancies in brain activity that was enough to define a ‘brainprint’ with 100 percent accuracy.

Applications

Future applications of this technology will probably not be used to unlock smartphones. However, it could be used for locking much loftier technologies. The team behind the work sees a lot of potential for it in terms of high-security settings, where a limited number of authorized users need ultrahigh-level access.

Brainprints are more efficient for such purposes, given that the brain biometrics of an individual is very hard to steal, and in the unlikely event that it does get compromised, the specific identifier can easily be cancelled and reset. At which point, the authorized user can simply create a new brainprint.

The study was published recently in The IEEE Transactions on Information Forensics and Security.