Skip to main content



Networks of the Brain

http://gizmodo.com/5843117/scientists-reconstruct-video-clips-from-brain-activity

On September 22 2011, UC Berkeley scientists published a study in Current Biology sharing their findings on being able to reconstruct visual signals from the brain. The current process utilizes a functional Magnetic Resonance Imaging system to scan and record the subjects brain activity and blood flow through the visual cortex while they watched a video. The recorded data was then translated into voxels (three-dimensional pixel units), which allows scientists to connect shape and motion information into specific brain actions/signals. 18 million seconds of random youtube videos were then analyzed by the computer to create a database of potential brain activity associated with the videos. The computer then combines 100 of the videos it determines to be closely matched to the video the subject viewed. The resulting video is low-resolution and blurry, but does capture much of the defining characteristics of the original video.

The implications for these findings are enormous, where its applications span from dream visualization to psychological therapy, and to even visualizing the thoughts of humans who are otherwise unable to communicate. This subject has deep ties to the study of networks. The basic functioning unit of the brain is the neuron, which receives and transmits information through chemical and electronic signals. Neurons act by receiving electronic signals with its dendrites and transmitting signals with its axon terminals. Neurons form complex networks in the brain, which act to interpret sensory information, instigate an action, and form memories. The network of neurons activated due to a stimulus will be similar for other similar stimuli. Neurons can then be thought of as nodes, with a collection of nodes with edges between them forming a specific experience. All nodes do not have edges with all other nodes, however, differentiating one stimulus from another. This is what the technology developed at UC Berkeley interprets. The more accurately we can map out the neuron network associated with a specific stimulus, the more information we could potentially reconstruct into visual data or otherwise. I think the development of this technology has a hopeful future, and will allow us to understand and interpret just what cognition is.

Comments

Leave a Reply

Blogging Calendar

September 2011
M T W T F S S
 1234
567891011
12131415161718
19202122232425
2627282930  

Archives