Post Your Entry!
MONDAY, JULY 04, 2011
 
What I’m going to term as “dream seeing,” since there is no term for this scientific study yet, should not be confused with reading dreams. Dream reading is just listening to a person explain what they remember from their dream and then interpreting them. However, what I’ll be covering, dream seeing, involves a third party being allowed to actually view the images that are going through a person’s brain as they dream. And if the dreamer forgets what they dreamt about, the person who was watching will have every dream by the dreamer stored on a computer; So that both the dreamer and observer can go through the images that appeared in the dream again.

This may sound like science fiction right now but ground breaking technology has created the potential for such machines. Scientists from Japan’s ATR Computational Neuroscience Laboratories have released the new study in the Neuron journal.

The computers which can view the images are similar to MRI scan machines, except they use: Functional Magnetic Resonance Imaging (fMRI). The coolest thing about the fMRI scans is that by observing multiple scans of different scales the researchers can actually translate that brain activity into the image that is being viewed by someone. So say someone is looking at the shape of a square, then the fMRI can create the image of square simply by examining the brain’s activity. The brain's activity is displayed in an fMRI scan by using voxels. A voxel, is like a pixel but in 3 dimensions instead of 2. As the study explains: "We reconstructed visual images by combining local image bases of multiple scales, whose contrasts were independently decoded from fMRI activity by automatically selecting relevant voxels and exploiting their correlated patterns." 1

As the image is formed in the brain, the fMRI scans are examined and able to produce a pixelated copy of the original image.

The small patch images were also recreated without the use of prior images, the study explains: "Binary-contrast, 10 × 10-patch images (2100 possible states) were accurately reconstructed without any image prior on a single trial or volume basis by measuring brain activity only for several hundred random images." The images created by the fMRI have yet to be perfect. They look very closely to pictures taken by cheap low-resolution digital cameras but the results are accurate. That is, if someone is looking at a square you won’t see a triangle appear on the fMRI scan, what you get is what you saw.

The scientists have also concluded that such examinations could be applied to someone as they're sleeping to view any images that appear in a person's dream. Although the experiment hasn't been performed, the scientists don't seem to see many major problems in accomplishing such a task. Right now what waits to be worked on is refining the precision of the image produced by a person's thoughts. In doing so, a dream's images will be easier to understand.
Comments

Hogan

Hogan

So let me get this straight. These fMRIs can produce low-resolution images of what people are looking at, but not, at least at the moment, what they might dream of. Right?

The question this begs, for me, although it may be easily answered, is this: When people dream, are the visual centers in the brain active in the same way in which looking at something makes them active?

If so, the fMRIs could produce the same kind of images for dreams as for looking, presumably. But when we dream are our visual centers activated in the same way as looking? In a dream we're not actually "seeing" anything. We're asleep with our eyes closes. So, is looking at x the same as dreaming of looking at x, neurophysically? Is it the same in the brain?

This seems to me to be a central question. What do you know about this, Alamir?

The article and the study behind it is pretty friggin' cool, especially if it mght apply to dreams as well. I've dreamt, so to speak, of such a possibility.

Dreams are a strange brew of emotions and virtual sense experience. They seem as real as real life, and the emotions they evoke are just as, and often more, intense than in reality. But is it the same as actually experiencing the sensations, neurologically? That, to quote Hamlet, is the question.
REPLIES: Alamir

Alamir

ALAMIR

Replying to Hogan:
That part is confusing as I've read conflicting reports from both the actual study and reports on the study which are really just translations of Japanese reports on the study. But if I go what I've found in the study itself it seems the images we think in our brains are what produce the specific brain patterns for the fMRI to pick up and later be translated into images for third party observers. So we don't actually need to see the image, we only need to think it. This obviously makes it much more easier to see a person's dreams if we only need to see what they think they see.
This is not the first time I've written about this companies research team (the Japanese scientists are so much cooler than ours) and in their previous research I know that humans only needed think certain thoughts in order for the fMRI to interpret the brain patterns accurately. With that said, I'll have to do a bit more research to be 100% sure about it (I'm only about 90% sure) and/or find out how the eye-sight is related. I'll be updating this article and probably writing a response to it on some more advancements that have been made in this field.



USERNAME:
PASSWORD:
REMEMBER ME
Forget your password?
Don't have an account? Sign Up, it's free!
Most Discussed Articles Top Articles Top Writers