Doing the EEG? Perhaps in the future you will have to remove the biometric data

Electroencephalography is a long-established method for the study of the brain. It would seem, than it can surprise? As it turned out, very, very much. For example, a group of researchers from the School of engineering and applied physics state University of new York at Buffalo-based developed a system for identifying a person by characteristics of his EEG.

The fact that the basis of EEG is the recording by means of special electrodes in the frequency and amplitude of electrical impulses of the brain. U.S. scientists decided to find in the EEG not only common, but also unique to each person signs in order that they could be used as biometric data. The experts took an EEG helmet and modified it so that the number of electrodes for registering doubled (64 vs 32 standard). Each area of the brain register 3 groups of sensors: the first captures the data associated with conscious memories, the second is the recognition of persons and objects, and the third reading.

The participants of the experiment (which, by the way, was 179) in this helmet was showing 3 images a picture of an animal, a portrait of a famous person and text. All 3 pictures show a total of not more than 2 seconds (or 1.2 seconds). The entire series of images was shown 4 times in a row at regular intervals of time. After that all recorded pulses has created a common “brain password”, which, as it turned out, was unique for each participant of the experiment.

Then everything is simple: only need to show once again man the number of images to maintain a “brain password” and compare it with existing passwords in the database. The identification accuracy is 95%, but 5 months after reading it decreases by 1% and this change continues to grow. Thus, at the moment, the use of EEG as a biometric measure has a major drawback in that you need to constantly update the database. However, according to scientists, the problem is solved.

Leave a Reply

Your email address will not be published. Required fields are marked *