Scientists have come a step closer to creating a "mind-reading machine" which can show mental images.
A team from the University of Glasgow succeeded in decoding brain signals related to vision.
Six volunteers were shown images of people's faces displaying different emotions such as happiness, fear and surprise.
In a series of trials, parts of the images were randomly covered so that, for example, only the eyes or mouth were visible. Participants were then asked to identify the emotion being displayed while electrodes attached to the scalp measured the volunteers' brainwaves.
The scientists were able to show that brainwaves varied greatly according to which part of the face was being looked at. "Beta" waves, with a frequency of 12 hertz, carried information about the eyes, while four hertz "theta" waves were linked to the mouth.
Information was also encoded by the phase, or timing, of the brainwave, and less so by its amplitude, or strength.
Professor Philippe Schyns, who led the study, said: "It's a bit like unlocking a scrambled television channel. Before, we could detect the signal but couldn't watch the content; now we can.
"How the brain encodes the visual information that enables us to recognise faces and scenes has long been a mystery. While we are able to detect EEG activity in certain areas of the brain when particular tasks are performed, we've not known what information is being carried in those brainwaves. What we have done is to find a way of decoding brainwaves to identify the messages within."
Prof Schyns said the study, published in the online journal Public Library of Science Biology, revealed how the brain tuned into different brainwave patterns to code different visual features.
"It is a bit like radiowaves coding different radio stations at different frequency bands," he added. "This work has huge potential in the development of brain-computer interfaces."