Scientists extract images directly from brain (2008) | Hacker News Reader

Scientists extract images directly from brain (2008)

pinktentacle.com

77 pointstheforay41 comments

What they're doing is a high-resolution fMRI scan of V1 (primary visual cortex), a region of the brain that shows a "retinotopic map" of the visual field. It's been known for a while that it's possible to reconstruct an image from V1 activation. What this research shows is that fMRI imaging is now detailed enough to do this non-invasively.

Also, something cool about this research, and also a bit scary, is that it's also known that this region is activated by visual recall (a quick search revealed this: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=48...). I'm not sure if anyone has ever tried to reproduce an image from V1 activation during visual recall, so I don't know if it would be a coherent image. My guess is it would though, in which case an fMRI scan could detect an image of an "N" if you asked the person to picture the letter.

Pretty cool.

I really think if a coherent image had been made from visual recall, that news would travel fast.

My big question would be whether the patterns that they're detecting in the brain are similar for different people? For example, if I imagine an apple, do my "brain patterns" resemble someone else's when they also think of an apple? If not, seems like training this device would be a pain in the ass, because you'd have to do it for each person. But if so...imagine a huge resource of patterns for different thoughts, feelings, mental pictures, etc. Hopefully this comment makes sense...

In general the patterns are the same and the brain areas involved are the same but there is definitely individual variability, too. There is probably much less individual variability in visual cortex than, say, language representation areas (consider English speakers vs Spanish speakers) but there is likely enough that they have to retrain their algorithms for each individual to achieve optimum discrimination.

There is indeed enough variation. I unfortunately can't find the link, but they went further and had people imagine something like the letter 'n,' and of the whopping eight people they tested, one came out completely garbled, even after training.

Also, I suspect that variation increases with abstractness, so it will almost certainly be impossible to extract high level thought without extensive cooperation.

For example, the brain state that the word 'conciousness' invokes in me is going to be entirely different from you because I have different associations, examples, knowledge, and general representation.

Now try to extract an entire sentence without training first. Have fun!

On one hand, this is great. It will lead to new technology which will inevitably lead to better video games.

On the other hand, this provides the opportunity for total loss of privacy. Forget freedom of speech, people might start to attack the freedom to think certain thoughts. In other news, China and Iran are thrilled at this new science.

Then we'll have to encrypt our thoughts.

Our thoughts aren't standardized, so you get encryption for free.

That's not at all clear.

"In other news, China and Iran are thrilled at this new science."

Why? Are they having problems with people imagining politically undesirable images?

I would think the West, with its insane war to protect the cartoon children, would have more use for such technology.

I wouldn't claim the West lacks political repression but it seems obvious that any repressive states would love have such technology. Your argument seems to discount this.

In the long term technology like this will end privacy for those who want to participate in the modern economy. But you shouldn't assume it will happen in a 'top down' way that will benefit totalitarian governments. Totalitarian governments require secrecy to do their evil deeds.

Look at the spread of mobile cameras (mostly in phones). These have allowed many an injustice to be exposed. The police no longer feel like they have immunity to do whatever they want.

Consider the benefits of being able to scan the brains of those accused of crimes against humanity. If people understood that in the long term there are no secrets they would hesitate to do evil deeds.

Not only that, but it should be perfectly clear now that if this technology is perfected and inexpensive in the future, then doing things now that you wouldn't want people to know about in the future is usually a bad idea as well. Act and think as though everything about your thoughts and actions will eventually be known to everyone who might be interested.

I guess how you think and act, then, will be a measure of whether you want to be thought of as moral or interesting. ;)

I think that our exposed thoughts would fall under freedom of speech in America. That, and the 5th amendment ("...nor shall be compelled in any criminal case to be a witness against himself...").

I'm not sure how relevant the constitution will be in a future where you can read peoples minds among other things.

But don't forget the part about "without due process." I think that in the future, we will definitely see the use of technology like this to read the minds of accused felons. I also predict the court will declare this perfectly consitutional, as your right to privacy is effectively waived given the right court order (and the motivating "probably cause").

TRULY innocent people will probably like this tech a lot.

It'll go both ways.

The first thing I thought when I saw this was:

"Oh shit- now nothing is private."

Don't get me wrong- it's amazing stuff, and I'm all for the advancements it can bring. But the mind really is the last bastion of privacy in our increasingly policed world.

Think about it:

1) When you make a call, it can be logged and recorded.

2) When you walk into a bank (or a million other locations), your actions are on camera.

3) When you visit a site, your IP address, as well as other information, is logged.

Almost everything is monitored.

Granted the advancement of technology has brought FAR more benefits than costs, but sometimes I wonder what it would be like to live in the 1800s, when you could do something and not have to worry about your actions being "logged" to some eternal database.

Of course, it's fairly easy to control actions. Don't send really personal stuff via email, don't scratch your crotch in front of a security camera. But when it comes to thoughts, it's a different story. Try not thinking about something, and you'll find you can't do it.

I'm sure every single person in this world has had thoughts he/she would not want to reveal.

But I'm getting ahead of myself. It will probably be a while before we have to seriously worry about the implications of this technology.

Try not thinking about something, and you'll find you can't do it.

People say this, but I find it's not true. Just think of something else. Works most of the time.

I just lost the game.

It works until they can read the subconscious and can detect the thought you're trying not to think.

Granted the advancement of technology has brought FAR more benefits than costs

So far.

I don't think we've given it enough of a trial run. Global warming and overpopulation are two results of technological advancement that could potentially outweigh all the benefits of progress if they get bad enough.

We're all doomed to carry around thousands of pounds of 3G connected MRI.

Not to impugn the author's trustworthiness, but this is so astounding to me that I find it somewhat incredible. Does anyone have an English-language news source for this?

The "extracted" pictures are an order of magnitude clearer than I would have thought possible.

Your skepticism is commendable, but the pictures are real. Here's the paper:

http://www.cell.com/neuron/abstract/S0896-6273%2808%2900958-...

That's fantastic, and the full text is available there too. Thanks.

I believe it, if only because I've read of them doing this with cats several years ago.

Seems like it's a neural network trying to decipher the inner workings of another neural network. How amazingly cool is that ?

Recursively cool.

There were interesting comments about this when it was posted here before:

http://news.ycombinator.com/item?id=394826

I think if I could see any invention in my lifetime, I'd want to see that knowledge-uploader thing from 'The Matrix'. Imagine how much you could learn! Especially before you were too old to use it...

The source for this article is a 404 page. (Translated via Google) http://translate.google.com/translate?u=http%3A%2F%2Fwww.chu...

Questions:

- Will the fMRI patterns for the same image be the same for two different people?

- When I think of images, the vision I have in my mind are not as vivid as when I see them directly. Given this, will the fMRI patterns be the same in both cases?

I can see this being amazingly handy in the (far) future in terms of interface navigation and search. Just picture the icon of the program you want to launch, or the stock photo you are trying to find and boom.

That's already available, to some degree:

http://en.wikipedia.org/wiki/Emotiv_Systems

(There are others as well, but I understand this is one of the best commercial products as of now).

Ah, yes I am familiar with this. I've used a similar device during some biofeedback treatment. The leap with the imaging tech is that it is a 1:1 correlation. As far as I am aware the other tech still relies on mapping cues from the brain that aren't directly 1:1 For example, boosting your high beta waves moves your game character forwards, etc. With the imaging technology, you don't have to perform mental gymnastics, you just "picture" the image you are after and the like.

This should be like front page news. This is something many would have not thought possible. It also poses many problems non-materialists that do not subscribe to a weird hand-wavy coincidentalism.

I don't see how this should pose problems to non-materialists. There's no reason the image can't be formed in the brain before or during transmission to the immaterial mind, for instance.

I'm not a dualist by any stretch but to claim this as evidence for materialism is mistaken.

No, you may not understand materialism. Sure, as I already said, if you subscribe to some weird coincidentalist view maybe not but for those not grasping at straws this removes a lot from a dualist conception and it even further shows it as an unneccessary hypothesis.

Not at all: proving that the brain forms images from sensory inputs only proves that the brain forms images from sensory inputs. It's a tenable dualist position that the material brain handles sense perception while the immaterial mind handles higher level consciousness and thought. In fact, the body was always classically assumed to handle most of sense perception even before this research.

I understand materialism just fine. But this doesn't prove materialism.

Hooray! So now I can download the porn images I saw elsewhere back to my laptop directly from my brain! This way you get all the porn mags for free!

Reply on news.ycombinator.com