Japanese Scientists Claim To Reconstruct Images From Brain Data

In a world first, a research group in Kyoto Japan has succeeded in processing and displaying optically received images directly from the human brain. Here's the Japanese press release for good measure. One step closer to broadcasting your dreams? The research is due to be published today in the US scientific journal Neuron

No pictures?

[pwnies]

Honestly? Come on now. Saying you can retrieve images from the mind, then not showing said pictures is the same as claiming you've achieved cold fusion without showing any energy for it.

I think this is the first time I can scientifically say, "Pics or it didn't happen."

Re:Kinda neat, not that exciting though

[oroborous]

That's true, but advances in Optical Imaging might overcome the current imaging limitations. But sci-fi aside, at least getting the decoding algorithms perfected will answer a ton of basic science questions about network dynamics in primary sensory and motor areas.

Re:Kinda neat, not that exciting though

[Futile Rhetoric]

Can we "keep images in our heads" at all? When I try to, it is more of a feeling than an image, and it's a fragmentary one at that. Wouldn't it make sense if our imagination worked a lot like our vision, i.e. we can only focus on small bits of the visual field at once, and so would only be able to imagine those pieces and attributes of an image pertinent to our needs or wants?

I'm free-balling here, mind. I can't seem to put coherent, complete images in my head, but others very well might.

Re:Dreaming Is A Private Thing

[geekoid]

No he did not.
He wrote a story about something like this. People ahve thought about doing this for years.
There is a difference in predicting something, and writing a story.

He also wrote about a bunch of stuff that never happens, and won't likely happen.
I like the mans work, but come on if he gets put any higher on a pedestal he'll be able to touch the moon.

TLJ

[CannonballHead]

Reminds me of Dreamfall: The Longest Journey.

Does it get lonely up in your crystal tower?

[ovu]

"The current accomplishment is low hanging fruit and therefore uninteresting. Surprising, really, that they found funding for such an unnecessary demonstration at all! By commercializing this technology, it would become sufficiently interesting to deserve my royal approval."

Belittling humanity's incremental advancement as if you're a third party, how's that working out for you?

I think it's tremendously exciting. Thanks for the buzzkill though, it reminds me to get off the computer and interact with people of my choosing.

Emacs Macro!

[140Mandak262Jamuna]

I think code must faster than I can type. Soon I will be able to just wear a sensor filled helmet and think code and this machine would convert it to an emacs macro and fill in the source. Yay!

Some great potential here

[Drakkenmensch]

If scientists can understand perfectly how the brain translates images into neural signals, it opens up the possibility to build full cybernetic eye replacements, even if the nerve tissue is damaged and non-functional. The medical applications are mind-boggling.

Also, I'm reminded of the interrogation device from the movie Barb Wire, the one that pulls out images from your brain whether you want it to or not...

Law Enforcement Sketch Drawing

[dlevitan]

I think this would be amazing for law enforcement sketches. Instead of having to ask a witness what the person looked like, they could just copy it out of their visual cortex. No, it wouldn't be perfect, and it wouldn't be acceptable in court as proof someone was there (since you can just imagine your worst enemy in the place of the actual person), but it would help with sketches for wanted posters and the like. Especially if it was cheap and easy.

Thanks for the link

[tobiah]

I just skimmed both papers, looks like the Japanese group goes well beyond what they did at Berkeley, capturing true images, whereas the Berkeley group only found some evidence that this would be possible.

Re:This is so cool!

[Omestes]

This , if true , will have HUGE implications - we'll be able to see what people THINK.

Data in V1/V2 does not constitute cognition, those areas constitute pretty much a visual map of data gathered by the eye (roughly). Its doubtful that imagined visuals are even represented in these areas. This, in other words, doesn't provide any insight into thoughts, just what people see.

I admit, though, that this is awesome. If we can read it, we theoretically could write to it, which would allow for direct neural interfaces.

Re:Emacs Macro!

[astrodoom]

Only if you actually SEE perfect code in your brain. Personally when I'm coding, I'm thinking much more of a process than a picture. We're still no-where near that with just mental picture projection. What I really look forward to is being able to project memories that can't be fully remembered. I want to see what a partial memory looks like. Generally when I have difficulty remembering, it's difficult to construct a mental image, so I wonder what results when thought processes like that are going on.

They can only detect pre-loaded images

[Anonymous Coward]

If you read the article, you'll note that they can only identify which image from a known set of very simple images the subject is looking at.

Yes, yes, early days, but if you think of the brain patterns being monitored as being like a hash of the actual image, it illustrates how limited this is.

Analogy (no cars):
Take a set of known words, hash each of them. Select a random word from that list. Now hash it. Compare that hash to the list of hashed words to find out what word was selected.

This doesn't mean that downloading entire dreams will every be possible.

Re:Kinda neat, not that exciting though

[jhoger]

Having a device like this, you might be able to learn how to hold an image in your head. If it could give you instant feedback, perhaps you can learn it like any other skill.

So the existence of the device, over time, could change the answer to your question.