Scientists Extract Images Directly From the Brain

[INRM]

Scientists Extract Images Directly From the Brain
http://www.slipperybrick.com/2008/12/scientists-extract-images-directly-from-the-brain/

In what could be the first step toward recording your dreams, researchers from Japan’s ATR Computational Neuroscience Laboratories have developed new brain analysis technology that can reconstruct the images inside a person’s mind and display them on a computer monitor. That means it won’t be long before you can share your thoughts and dreams with others the way you share your flickr pics. They’ve successfully displayed simple images produced in the human brain on a computer screen.

Considering this devices by converting electrical activity to an image... does this use some kind of EEG for it to work? Or does it require some other device?



BTW: I need not state that this technology will almost inevitably end up being used by law enforcement and by the government and truthfully will probably end any and every last form of privacy in existance

[Twiler]

Why couldn't laws be passed to restrict their use?

Such laws could be broken, but so could the laws against torture and threats.

[rhtufts]

Yes! One step closer to having Google connected directly to my brain.

[technoextreme]

The same technology dates back to 1999 where they did the same exact thing with cats. INRM do you always get intensely paranoid over anything you obviously have no clue whats going on? Here is a helpful hint. Don't start extrapolating science fiction scenarios from news reporters that are probably complete and utter idiots when it comes to science.
Considering this devices by converting electrical activity to an image... does this use some kind of EEG for it to work?
FMRI which measures the blood flow in the brain. Not need to be really paranoid over that type of technology. It's impossible for law enforcement to broadside people with a one ton humming machine that can rip metal out of your body.

[X]

I think this is cool.

What does this mean for understanding how the brain processes images?
For example, could the technology be used to improve camera systems granting sight to the blind?

And it seems from the article that this just works for images being viewed at the time. Does anyone know if it can work for "remembered" images?

[thatguywhojuggles]

In the comments below the article, someone pointed out that the two "n"s are identical down to the static. Is the eye and brain that accurate?

[Kaylee]

Researchers (http://news.yahoo.com/s/afp/20081211/sc_afp/sciencejapanbrainoffbeat_081211052641) at the ATR Computational Neuroscience Laboratories succeeded in processing and displaying images directly from the human brain, they said in a study unveiled ahead of publication in the US magazine Neuron.

<snip>

When people look at an object, the eye's retina recognises an image that is converted into electrical signals which go into the brain's visual cortex.

The team, led by chief researcher Yukiyasu Kamitani, succeeded in catching the signals and then reconstructing what people see.

In their experiment, the researchers showed people the six letters in the word "neuron" and then succeeded in reconstructing the letters on a computer screen by measuring their brain activity.

The team said that it first figured out people's individual brain patterns by showing them some 400 different still images.

It seems that each person’s brain needs to be mapped individually. I wonder how long that process takes.

I would think that mapping thoughts (vs. mapping how each person's brain processes a printed letter) would take much longer and be prone to many misinterpretations – especially in real world applications where I imagine most people would not be interested in cooperating.

[Lennart Hyland]

Isnt there already a thread about this?

[Dancing David]

Whoa everybody, at least the excited ones.


In their experiment, the researchers showed people the six letters in the word "neuron" and then succeeded in reconstructing the letters on a computer screen by measuring their brain activity.


Now all the pop science articles say about this is that.

So images are focused on the retina, photoreceptors then trigger and there are further neurological transmission by neural transmission.

Some of these images pass up the optic nerve into the brain where eventual they are turned into perceptions.

So this begs the question:

Is this:
-in the optic nerve bundle?
-in the pathways to the visual cortex?
-in the visual cortex?

Each is a little closer to the actual areas where the perceptions are created.

So NO they can NOT look at your dreams!

So far they can look at something, most likely you optic nerve bundle, and process something from there. Now can they distinguish between the word 'neuron' and 'moron' or 'nervous', they don’t say. Can they tell any visual image from any other visual image? They don't say.

I doubt it highly.

As Kaylee points out, you have to map the individual pattern of signals for this to work. each person has to be sat down and the device calibrated.

Considering each of us has similar brain morphology but unique and individual neural fuzzy logic programming, to map a thought would require a whole lot of effort. It would be unique for each person. And it would take a huge amount of calibration, like sitting you down and making you watch movies to see what you respond to, and what areas, but how do you control for the side chatter in thoughts. what if the person deliberately thinks of poetry and nonsense phrases during that time.

Your calibration would be garbage and full of spurious signals and information.

Maybe five hundred years from now this will be an issue.

[arthwollipot]

Colour me skeptical. I'd like to see more information.

[Skeptic Ginger]

If you can convert brain waves into images, then there is a mechanism for ESP to work. Not saying it does work or that the brain waves are strong enough to be transmitted over any distance, nor that there is any mechanism to receive the brain waves so don't bother going there.

My point is, do our brainwaves themselves contain information that is universal, or do brainwaves only represent activity in neurons that contain the actual information.

When you drive a car, the fuel runs the motor. The fuel contains no car parts, no car identity, no car 'information'. If you collect the electricity which ran through a light bulb, you cannot distinguish it from electricity which ran through a radio.

I saw this article but haven't seen anything which discusses what is meant by reading one's thoughts. Knowing the area of the brain which is active gives you information about one's thoughts. Learning how to control a device with nerve impulses (thoughts) is only possible because one learns the particular 'thought' which triggers the device. Biofeedback in other words.

I don't believe our brainwaves contain information about the thought which resulted in the brain wave. It would be incredible if that hypothesis was incorrect.

[Skeptic Ginger]

OK, just read DDavid's comments. I see that one must first interpret X's specific brain wave pattern in order to read X's thoughts, and only the thoughts one has correlated with the interpretation.

X's brainwaves do not contain some universal information about Y's thoughts.

Thanks, DD.

[technoextreme]

If you can convert brain waves into images, then there is a mechanism for ESP to work. Not saying it does work or that the brain waves are strong enough to be transmitted over any distance, nor that there is any mechanism to receive the brain waves so don't bother going there.

FMRI!!! FMRI!!!!! FMRI!!!! Your not measuring brain waves with FMRI!!!!

[Skeptic Ginger]

FMRI!!! FMRI!!!!! FMRI!!!! Your not measuring brain waves with FMRI!!!!FMRI, brain waves, who cares?

[Skeptic Ginger]

OK, I've rethought the brain imaging vs electrical activity issue. But I'm stuck on the electromagnetic spectrum.

With an EEG, brain waves, aka electrical activity, can be detected by measuring the transmission onto the surface of the brain. The fMRI, OTOH, takes a picture of heat/bloodflow within the brain. But in order to take this image, an electromagnetic wave still must reach the monitor which is measuring the activity.

Neither one is likely to carry information which any other individual's brain could interpret. So again, I wasn't making that claim. But any energy which can be detected outside the brain which originates from thought is something that is detectable.

This particular research merely confirms what I previously concluded, that something which can be detected, does not contain the information, only a reflection of the information which the brain contains. It can be measured and correlated with the brain activity which it originates from. But it doesn't contain inherent images or vocabulary.

So, it remains as not being a possible mechanism for ESP.

[X]

OK, just read DDavid's comments. I see that one must first interpret X's specific brain wave pattern in order to read X's thoughts, and only the thoughts one has correlated with the interpretation.


Disclaimer: My mind is too be read solely at your own risk. What happens to your sanity is not something I can be held accountable for.
:covereyes



X's brainwaves do not contain some universal information about Y's thoughts.


Yes they do.



--- --- --- ---

I still think, if this is an accurate report, that it might have promise for restoring sight to the blind. But I'm not a biologist, a doctor, or a neurologist, so what do I know (aside from Y's every thought)?

[Skeptic Ginger]

...

Yes they do. ...I was trying to avoid a useless semantic argument about what constitutes "information".

What 'universal information' pray tell, is contained in the brain's electrical activity that does not require the individual's brain structure with it? I'm not talking about sleep/wake patterns yadda yadda. I am talking about images and words which constitute thoughts.

[Dancing David]

Um, electrical signals they are not.

Biochemical signals involving a phase shift in a bilipid membrane they are.

There is an 'electrical' effect. But think of a biochemical impulse traveling on the neuron, not an eletrical signal in wires.

[Belz...]

If true, neat.

Could be used to gather evidence... though I suppose, like hypnosis, won't be much help in sorting out fantasy from reality.

[Belz...]

If you can convert brain waves into images, then there is a mechanism for ESP to work. Not saying it does work or that the brain waves are strong enough to be transmitted over any distance

Ooh... My sci-fi nerd neurons just fired: I imagine cybernetic investigators armed with thought-reading devices in their brains who hack into unsuspecting victims and record their thoughts during interrogation.

Wait! I've had this one before. Stupid reality, making my fantasies true.

[Kuko 4000]

Colour me skeptical. I'd like to see more information.


Download the full paper from:

http://uploading.com/files/DYB9I37X/sdarticle.pdf.html

Previous thread:

http://forums.randi.org/showthread.php?t=130619

[INRM]

It uses fMRI? I thought it used some kind of electrical measuring device.

What's a "multivoxel" pattern decoder? Also, what's a "voxel"?


Twiler,

Why couldn't laws be passed to restrict their use?

There could. Unfortunately I don't see anyone doing anything to that effect.


INRM

[JPK]

One of my favorite movies has this as the theme. Until the End of The World.
http://www.imdb.com/title/tt0101458/
A scientist played by Max von Sydow invents a camera that can record someones experiences. Not just what they see, but how they feel emotionaly as well. He sent his son, William Hurt, all over the world to record images to bring back. Sydows wife was blind. One of the other things this camera could do was record a persons dreams as well.
Makes for good SciFi, I'm not sure how this could work in the real world.
JPK

[arthwollipot]

There's certainly a long way to go before this technology can record someone's dreams. Yes, Kuko 4000, I've now seen more information on the subject. :) It's a fascinating technology, but I think it's a gateway for more useful applications to be developed, rather than an end result in and of itself. The potential is great.

[uruk]

Yes! One step closer to having Google connected directly to my brain.

To hell with Google, what about delivering porn directly into the brain!!!!!!

[uruk]

~snip~
BTW: I need not state that this technology will almost inevitably end up being used by law enforcement and by the government and truthfully will probably end any and every last form of privacy in existance
How would they tell the difference between an image of something you dreamed or imagined from an image of something you actually experianced? I think you would still have the problem of false or distorted memory.

What would really interesting is if they could implant an image in the Brain.

Watch the movie "Strange Days" with Ralph Finenes and Angela Basset. Itt deals with type of technology.

[uruk]

What's a "multivoxel" pattern decoder? Also, what's a "voxel"?
INRM

The 3D equivilent of a pixel. "Volumetric picture element"

A voxel contains three dimensional position information as well as 2D and color information.

[X]

I was trying to avoid a useless semantic argument about what constitutes "information".

What 'universal information' pray tell, is contained in the brain's electrical activity that does not require the individual's brain structure with it? I'm not talking about sleep/wake patterns yadda yadda. I am talking about images and words which constitute thoughts.


It was a joke.

[CFLarsen]

Memory isn't a tape recorder.

[BenBurch]

Memory isn't a tape recorder.

No, but it is a physical medium. Some day, we will be able to read it out.

I'll likely be worm food for a few centuries by then.

[INRM]

arthwhollipot,

It's a fascinating technology, but I think it's a gateway for more useful applications to be developed, rather than an end result in and of itself. The potential is great.

Yeah, great for violating people's privacy...


BenBurch,

No, but it is a physical medium. Some day, we will be able to read it out.

That's probably correct.

I'll likely be worm food for a few centuries by then.

That I doubt. I think we'll be able to do it in five to ten years minimum, and about 10-20 years tops. I do not simply mean the ability to read the images in a person's mind but the ability to identify even subtle and abstract thoughts.

Eventually this ability may even become remotely achievable. I assume within the 20 year period this will probably happen. When this does, privacy as we know it will cease to exist.


INRM

[uruk]

Eventually this ability may even become remotely achievable. I assume within the 20 year period this will probably happen. When this does, privacy as we know it will cease to exist.


INRM That would depend if they can do it from a distance or if a device has be in contact with the person.


Anyhoos, people willingly give up thier privacy everytime they post on Myspace or Facebook.

[CFLarsen]

You can't read memory. Memory isn't stored. It is (re)constructed.

[JPK]

You can't read memory. Memory isn't stored. It is (re)constructed.

Getting back to that movie I mentioned earlier, this was a point. The person who originaly recorded the images had to rewatch them while they were being tranmitted to someone else in order to reconstruct the memory and how they feel about the images.

John Kardel

[arthwollipot]

Yeah, great for violating people's privacy...Nuclear technology can destroy cities, or it can provide electricity for them. Genetic engineering can create megaviruses, or it can feed millions of people.

The fact that a technology may be used badly at some time in the future is no reason not to pursue it.

[Skeptic Ginger]

It was a joke.Crum! That's twice in a week and just after the news report people who miss sarcasm have a particular type of dementia. :eek:

[Skeptic Ginger]

... Memory isn't stored. It is (re)constructed.That's absurd. Some part of memory has to be stored, regardless of the process of replying it and revising it.

[Dancing David]

No, but it is a physical medium. Some day, we will be able to read it out.

I'll likely be worm food for a few centuries by then.


Memory is partly stored but mostly reconstructed, under some current models. It again would be prone to people thinking contrary thoughts and the like, as far as recovery.

[Dancing David]

arthwhollipot,



Yeah, great for violating people's privacy...


BenBurch,



That's probably correct.



That I doubt. I think we'll be able to do it in five to ten years minimum, and about 10-20 years tops. I do not simply mean the ability to read the images in a person's mind but the ability to identify even subtle and abstract thoughts.

Eventually this ability may even become remotely achievable. I assume within the 20 year period this will probably happen. When this does, privacy as we know it will cease to exist.


INRM

Wow, that is a nice assumption, how will they overcome the fact that our brains are all unique and have unique patterns of firing?

Each person would have to be calibrated individualy. Think of a snowflake made up of some trillion neurons. Then think of the fact that each persons set of patterns is idiomatic. the neural association that I have to the thought 'dog' are not going to be related to the neural patterns that you have for the thought 'dog'. In that the cells may be in totally different positions and relations to each other.

Brains are not hardwired like computers, they grow in place, gross morphology is similar but actual neural connections are not. Each brain then develops the set of patterns it uses to to form association is not. So the patterns of firing are idiomatic and contingent upon history, then they are plastic even after that.

So each person would have to hooked up to the machine, each person then run through a very long process to calibrate the machine. And then due to plasticity each person would have to be recalibrated.

[Dancing David]

That's absurd. Some part of memory has to be stored, regardless of the process of replying it and revising it.


Again this is very speculative, theories are tending towards reconstruction right now. Neural associations are a huge part of the process. (And they are very plastic) So patterns of a tendency to be part of a reconstructed memory are going to be there, but given the nature of confabulation, hard wired memory, like bits on a disk, seems unlikely. Rehearsal is also important and may happen at a preconscious level, I have been able to recover some pieces of memorized verse many years after the fact. But it takes time and practice. Even then i had to look up some parts.

Please do not assume that brains are hardwired, which you may not have, everything is fuzzy and reverberating.
:)

[INRM]

DancingDavid,

According to most of these fMRI tests, they were able to conclude that most human brains were close enough to each other that they could determine pretty much via algorithms what was going on in any brain.

[CFLarsen]

If some parts of memory has to be stored, which memories cannot be misremembered?

It should be an absurd no-brainer.

[Dancing David]

DancingDavid,

According to most of these fMRI tests, they were able to conclude that most human brains were close enough to each other that they could determine pretty much via algorithms what was going on in any brain.


Really, where did it say that?

:)

I still haven't found out where the pattern they interpreted was located. My guess is still the optic nerve.