Thursday, September 15, 2011

Cloud computing in the brain




I meant to post on this earlier this summer but got distracted... work & life have a way of doing that.

This past July saw one of my most grueling projects finally come to a close. Published in the Journal of Neuroscience and titled “How each movement changes the next: an experimental and theoretical study of fast adaptive priors in reaching" this paper goes well beyond the topic of simple motor control and into fundamental issues about how our brains learn from experience (this is the reason why it took over 2.5 years to get through the review process!).  In particular we show how neurons in the brain might do something akin to cloud computing.

Confused? Let me explain.

First I'll have to talk about a little hard science for some background.  What my co-author and I found was that whenever we move, our brain keeps track of where we go.  Over time, the motor control regions of the brain begin to generate an internal model of the recent movements we've made.  It then uses this memory when we make future actions.  For example, if I keep reaching for this soda can on my desk (and return it to the same position), then over time my brain retains the history of all the reaches I've made to that same spot on my desk.  Over time my reaches to the soda get a little more accurate.  So in a sense, practice makes perfect.

Now, here's the kicker... if I suddenly want to reach for something else ("Hey, that cookie over there looks mighty tasty!") then my brain biases this new movement in the direction of the soda can. We don't really notice it that much, but it's detectible with the fancy machines that we use to monitor your movements.  The more often I reach for that soda can, the more biased my reach for the cookie gets.

It turns out that this type of learning is really sophisticated and follows what appears to be rigid statistical principles.  What I mean is that our brain somehow encodes our recent actions as a prior probability distribution (think of the "bell curve" you've heard about).  It then integrates this prior with all the incoming sensory information you're getting from your eyes, your hand, etc.  This integrated information is then used when you make your next reach.  The stronger the prior, the more biased your future actions will be.  The stronger the sensory input is, the less biased you'll be.

For the math-nerds out there, this is a form of adaptive Bayesian inference.  I wont go in to the awesome details of Bayesian statistics because I don't want to lose 90% of whoever it is actually reads these posts ("Hi Mom!"). This is a branch of mathematics that's used to filter Spam from your inbox, improve images of the stars from telescopes, optimize airline travel, make video games more difficult, and almost everything else that's cool these days.  Needless to say, these are some pretty freaking sophisticated computations that our brains are doing almost effortlessly.  And not just for any high level cognitive process (I mean this process worked in both Shakespeare's brain and in Pauly D's brain)... but for something as simple as reaching for a can of soda.

Okay... hopefully I haven't lost anybody.  Because here's the truly insane part.

Through simulations of neural tissue, my co-author and I found that this really cool mathematical computation likely happens through a form of "cloud computing" in the brain.  For those of you who don't know, cloud computing is the process by which a computation is broken down into a set of little chunks and then distributed to a whole bunch of computers that live in the vast ether of the internet (Note: So technically what I'm talking about is more similar to a computer clusters, but "cloud computing" is the hip new thing these days).  You know those nasty things called "botnets" that take down servers in foreign countries or send you all that spam in your inbox... they're cloud computing gone bad.

We found that a similar principle might work in the brain.  There might not be a single neuron or group of neurons that store this statistical prior.  Instead, we were able to show how this memory can naturally emerge in the dynamics of the information passing between neurons.  Thanks to Hebbian learning ("neurons that fire together wire together") our brain is able to store little bits of information distributed across a mass of connected cells.  Basically populations of neurons remember their collective pattern of activity from the recent past.  Over time, this distributed collective memory shapes the way the network responds to new inputs.  Eventually, this learning exhibits very sophisticated properties that look almost exactly like human behavior, as well as the expectations of statistical theory.  So the whole is, in fact, greater than the sum of its parts.

Let's put this all together shall we?  We've got a group of neurons in your brain that are building a complex statistical model of everything you just recently did. But this model isn't encoded by the activity of any one cell or even in the response properties of a group of cells.  Instead this model simply exists in the abstract dynamics of how these neurons talk to one another.  Mind blown yet?

Now to be completely honest this is hardly the first time that someone has come up with the idea that information is broken down and stored across a network of neurons (in fact there's a formal name for this called "sparse coding").  But what's interesting from our study is that complicated, mathematically principled information can just emerge naturally in the brain thanks to the fact that neurons are recurrently connected (i.e., send information back and forth) and they have associative learning.  This dramatically increases the complexity of information that our brain can store.  Information isn't just encoded in how the cells fire, but also in how they talk to each other as a group. It's a sort of meta-level type of information storage.

Let's end with this... our simulation used about 180 simulated neurons (with 32,400 connections) and we were able to do some pretty fancy mathematical processing.  The human brain has more neurons than there are stars in the Milky Way.  Each of these neurons has, on average, 10,000 connections or so.  A conservative estimate puts it as about 100 billion neurons with about 100 trillion axons.   Think about just how complex of a biological computer that this system could hold!

Yup.... crazy stuff like this is why I do science!

Wednesday, September 7, 2011

If you don't like the message, kill the messenger

As some of you may know, one of my goals in life is to facilitate the role of science in society.   Last week I came across a little gem that reminds me of what we, as scientists, are up against as far as how science is discussed outside the lab.

Now before I start, let me be honest.  There have always been, and always will be, those who will instinctively disavow science.  For some it's a matter of fear ("That appears to threaten my world view and scares me").  For others it's simple ignorance ("That doesn't easily make sense to me, so I don't believe it").  But for a few it is really more a matter of power ("By minimizing science and its scope, I can act how I want regardless of what evidence there is that I shouldn't.")

The first two groups I can understand and even empathize with.  It's our nature to be cautious and skeptical.  Hell, as a scientist I'm trained to be skeptical and to wait for mounting evidence to make the case that I should reject my current beliefs.   However, usually those who dislike science out of fear or ignorance can be reasoned with if approached in the right way.  Remember that even Pope John Paul II, a man who lead an institution that was terrified of the concept of evolution, eventually conceded that it exists.

It is that third group of science denier terrifies me. And that brings me to what started this whole post.  Check out this clip of Presidential candidate Rick Santorum speaking to a law school audience last week.



Let's listen very carefully to what Mr. Santorum says.  First he appeals to authority" "Because I believe what the Catholic church teaches with respect to homosexuality" he therefore should be allowed to hold his beliefs that same-sex couples should not have equal rights in terms of marriage and adoption.  Now if it were just a personal belief that would be fine.  We all have to appeal to authority at some point in order to have baseline assumptions with which to act in the world.  But when faced with new evidence that those assumptions are incorrect, a logical person would change the assumption.  Mr. Santorum inherently acknowledges this fact when he goes on to say that there is no evidence to refute his "bigoted" beliefs.

But now listen to what Mr. Santorum says when confronted with two critical pieces of information that there is, in fact, evidence to refute his assumptions.  First the fact that the American Psychiatric Association (not the American Psychological Association as stated in the video, although they take the same stance) delisted homosexuality as a mental illness in the DSM-III in1974.  This, by the way, was after numerous empirical studies showing that, in all other aspects, homosexuals had no other presenting symptoms or negative life outcomes than heterosexual individuals.  Second, the fact that there is overwhelming peer-reviewed research by behavioral scientists that children growing up in same-sex households are no different, in terms of mental health and emotional well-being, than those who grow up with heterosexual parents.

Mr. Santorum replies "The American Psychological Association is made up of people who agree with the American Psychological Association." He then follows up with this gem, "A lot of psychologists don't belong to the American Psychological Association. A lot of doctors don't belong to the American Medical Association."

Now keep in mind this is a serious (albeit unlikely) candidate for the GOP nomination.  He's a trained lawyer and served in Congress both as a Representative and a Senator.  So he should, by all accounts, be a very bright and logical man.  Yet his response to just the existence of scientific evidence refuting his beliefs is the same as that taken by the lunatic anti-science fringe: if you don't like the message, attack the messenger.

As a scientist, I'd be fine if Mr. Santorum argued with the science itself.  Every study and every field has its weaknesses. Had he argued that showing a lack of a difference is not the same thing as there not being a difference, he'd actually be correct.  In science, we call it "arguing the null."  Had he said that the research was still preliminary because the children adopted by homosexual couples are just now reaching adulthood, that would also be a valid and reasoned response.

But that wasn't Mr. Santorum's reaction.  This initial reaction reveals something far deeper than a simple lack of understanding of the issue.  It shows a complete and utter lack of respect for the science being discussed.  In fact, he turned the appeal to authority on it's head.  It's okay for Mr. Santorum to hate homosexuality because the Catholic church says so, but it's not okay for scientists in the APA to feel that there's nothing wrong with homosexuality because they're just preaching to the choir.  Do you see that Catch-22?

If Mr. Santorum were the only major politician to act this way toward science, I could let it slide.  He'd be annoying but not threatening.  Unfortunately this is the rule these days and not the exception.  For many politicians, as well as media personalities and a growing minority of the U.S. population, refuting a scientific fact outright because scientists say it is a badge of honor.  It's something to brag about.  In fact, we live in a world where most of the contenders for a major political party's presidential nomination don't believe in evolution or global climate change.  Both of which have thousands (that's right.. thousands!) of empirical publications supporting them.

And that brings us back to the motive. As I mentioned before, Mr. Santorum has to be a smart and logical man in order to get as far as he did.  The same is true for the other leaders who instinctively attack science as a whole (with the exception of Mrs. Bachmann who's a whole other kind of crazy).  I'm afraid, their attack comes from a need for power.  "I can ignore the research on homosexuality and evolution because I want to court the religious vote."  "I can ignore the evidence of climate change because I want to court the industries who it will affect."

Unfortunately, this is an endgame move in the debate.  There's nothing you as a scientist can say or do that will change their beliefs because their ignorance is a source of power.  Arguing with those who take this stance is like arguing with the three year old holding his fingers in his ears... you just have to wait until they grow up so you can talk to them like an adult.

Update: As if on cue, last night's GOP primary debate featured Rick Perry demonstrating my point perfectly.  "Just because you have a group of scientists standing up and saying 'Here are the facts'... Galileo got outvoted for a spell."  The important point being distorted by Mr. Perry is that Galileo was not outvoted by his scientific peers... he was outvoted by the Catholic Church.  His "natural philosophy" peers respected and built off of his work.  In fact, by taking such a stance against the empirical findings, Mr. Perry is putting himself in the same boat as those who "outvoted" Galileo.