Neuroscientists at the University of Toronto in Scarborough, Canada, have developed a 'brain-reading' technique that allows volunteers wearing EEG to view several face images before reproducing the avatars. This technique is expected to help inability to speak Communicate with others.
In this technology experiment, the EEG records the activity of the brain. A set of programs called 'machine learning algorithms' generate face images on the screen similar to those seen by volunteers based on recorded data.
Adrian Nestor of the University of Toronto in Scarborough previously used functional magnetic resonance imaging to record human brain activity and 'reproduce' human face images in the brain. "The technique of using EEG to do this has been described by Ness Special postdoctoral researcher Dan Nemrudorf led the development.
The researchers said using functional magnetic resonance imaging can capture more detailed activity in specific areas of the brain, but the EEG is more practical because of its relative portability and cheapness.
Nestor's lab will study how to reproduce facial images in memory and images other than face with the technology of reproducing images based on EEG data.
Nestor said that the technology of EEG-based image reproduction "may provide a way for people who can not communicate through verbal communication," and may not only be able to reproduce the image being thought of in the brain but also the memory and imagination in the future It may also help law enforcement officers to generate images of suspects on the eyewitness's head without having their hands drawn on the witness's verbal portrayal of the suspect's head.
The research report will be published electronically by the Journal of Neuroscience.
