Obama Seeking to Boost Study of Human Brain • Hacker News Board

Re: Obama Seeking to Boost Study of Human Brain

#41

Post by etrautmann » Mon, Feb 18, 2013, 5:29 PM UTC

I'm currently working to develop neural prosthetics (computer cursor, robotic arm, etc.) and it's clear to me that the president's proposed research will have real clinical impact over the timescale of 5-10 years. Our lab and others already have paralyzed human patients controlling computers in clinical trials via a brain-machine interface, but it is hard to overstate the potential possibilities that remain hidden due to our lack of low-level understanding of low level circuits.

To clear up some misconceptions in this thread, this new work is NOT focused on top-down cognition using hand-wavy tools. Instead, it's focused on understanding neural circuits starting with simple organisms and working up to primates/humans by developing new technologies that allow us to radically scale up the scope of question we can ask.

Re: Obama Seeking to Boost Study of Human Brain

#42

Post by etrautmann » Mon, Feb 18, 2013, 5:39 PM UTC

gph wrote: post #9

I may be way out of my depth on this, but isn't this project assuming that there is a 'normative' human brain, and that we all have the same neuron structure? I know we all have the same general brain regions, and in most of us they function approximately the same. But to go down to the neuronal-level will any single brain be able to give a map that would be useful to the population as a whole? Or are they planning o…

I'm a PhD candidate in neuroscience, I'll take a stab at that.

You highlighted a critically important question in the field. We do all have the same brain structures, and neurons within those structures have similar shape and function, but the exact wiring of the circuits will not be the same for two individuals. Understanding how the structure of circuits leads to behavior, memory, cognition, etc. are the high level goals that the whole field is striving for.

This project seeks to answer quite a bit more than just mapping areas in the brain. I think the name is more of a way to connote that the project seeks to be comprehensive.

Re: Obama Seeking to Boost Study of Human Brain

#43

Post by etrautmann » Mon, Feb 18, 2013, 5:42 PM UTC

JoeAltmaier wrote: post #12

Sounds like snake oil. Too early to build a complete human brain model. Simpler brain models (cat, lizard) haven't been done yet (?) but that's not headline-worthy. How about "President announces program to solve all human suffering"? About as meaningless, but would actually be more useful if by some astronomical fluke they succeeded.

This research program is not about human brain modeling like the EU bluebrain. Instead, it's try to do exactly what you suggest, understanding simpler brain models and building up from that to more complex organisms.

Re: Obama Seeking to Boost Study of Human Brain

#44

Post by etrautmann » Mon, Feb 18, 2013, 5:48 PM UTC

ibuildthings wrote: post #6

http://bluebrain.epfl.ch/cms/lang/en/pid/56882 Here are more details of the European project (labelled Blue Brain Project) other posters had mentioned about in this thread. Also, there is this TED talk http://www.ted.com/talks/henry_markram_supercomputing_the_br... by Henry Markram, the project director of the blue brain project.

To be clear, there are not many similarities between Markram's project and this program. Critics of bluebrain should be very interested in the proposed approaches and technologies outlined here. Rather than starting by simulating a whole brain before understanding a functional circuit-level description of different brain areas, this program proposes to start by studying simple and accessible circuits and build up to larger systems.

Re: Obama Seeking to Boost Study of Human Brain

#45

Post by simonster » Mon, Feb 18, 2013, 5:57 PM UTC

keypusher wrote: post #15

While there has been a huge growth in brain research over the last decade or so, the methods and analysis are still immature, and results are often quite fuzzy (even if researchers would like their reviewers to believe otherwise). There also doesn't seem to be a clear goal here, other than to throw a bunch of money at some interesting questions. The Genome Project had the very clear goal of mapping the human genome.…

I've thought a lot about why fMRI data analysis tools are so terrible, and I've come up with a few reasons:

- It's impossible to implement the algorithms for fMRI data analysis efficiently in most "dynamic" programming languages due to the performance hit you take from using a dynamic language. (It might actually be possible in Julia, NumPyPy, Python with Numba, but these languages are not yet well-established.) On the other hand, dynamic programming languages are much better suited to exploratory data analysis than C is, so essentially all fMRI data analysis ends up being a mixture of C code and glue code in some other language. In this regard, I don't think SPM (MATLAB with C MEX files) is really that bad. It's fast and it avoids having to read the data from disk multiple times.

- People use what the tools they know, not the tools that are best for the job. FreeSurfer is a mess of C, C shell, and Tcl/Tk, but there's nothing else that can visualize fMRI data with comparable ease and accuracy. Most people in neuroimaging only know MATLAB, which is pretty terrible for analyzing large data sets because it can't mmap files (and it doesn't have the language features necessary to make this possible, and it's closed source).

- Related to the above, it's easier to get funding to develop a novel algorithm than to implement an existing algorithm in a way that makes it more useful/accessible to researchers. I believe this is slowly changing.

- There are a lot of different algorithms used for analyzing fMRI data, and no single package implements all of them. The necessity of each algorithm differs by lab and researcher, according to scientific necessity, personal preference, or the conventions of their subfield. People end up writing their own code to glue together methods from different analysis packages, which is, again, often written using the wrong tools.

- Us graduate students who know how to code well need to publish papers. There is comparatively little incentive to publish code.

Re: Obama Seeking to Boost Study of Human Brain

#46

Post by adventured » Mon, Feb 18, 2013, 6:02 PM UTC

jivatmanx wrote: post #38

I'm skeptical, but if it's reasonable to spend over a Trillion on a manned fighter project (F-35) considering the pathetic state of that project, then the few billion for this is worthwhile a million times over

How about neither.

The Paul Allen brain Atlas, among many other private initiatives, is capable of pursuing this goal better than the federal government, in the same way and for the exact same reasons that Craig Venter & Co. embarrassed the human genome project.

Re: Obama Seeking to Boost Study of Human Brain

#47

Post by abecedarius » Mon, Feb 18, 2013, 6:03 PM UTC

bnewbold wrote: post #31

abecedarius wrote: post #23

On the claimed $800 billion dollar return on the human genome project: http://www.nature.com/news/2011/110511/full/news.2011.281.ht... (It's skeptical but vague; it looked like the rest of the top hits just parroted the press release.) Anyone have access to the brain-mapping proposal? http://www.cell.com/neuron/abstract/S0896-6273%2812%2900518-... "The function of neural circuits is an emergent property that arises f…

At columbia.edu via scholar.google.com: http://academiccommons.columbia.edu/catalog/ac%3A147969 http://academiccommons.columbia.edu/download/fedora_content/...

Thanks!

On skimming it sounds well thought out as research on its own terms, though the connection to curing diseases is just as handwavey.

Re: Obama Seeking to Boost Study of Human Brain

#48

Post by adventured » Mon, Feb 18, 2013, 6:10 PM UTC

AutoCorrect wrote: post #17

How about a 10 year project to pay off the National Debt? Quit wasting money on science projects that aren't needed (this is already being done).

3% interest times $20 trillion = $600 billion per year in interest. (interest rates will not stay this low indefinitely, and even 3% would be crazy low given the actual default risk via inflation that the US poses)

That interest basis alone would effectively wipe out the ability to continue Social Security (or almost the entire US military, take your pick).

The national debt can never be repaid under any circumstances. We can't afford the real interest cost right now, which is why the Fed is paying for that via debt monetization (aka QE). Throw in just a trillion per year in principle, and it becomes a sad joke.

We will never, and have no plans to ever pay for the national debt. There's no scenario under which the math works out, unless our government suddenly becomes hyper disciplined and initiates an uncompromising 50 year payback plan that takes a hatchet to the entire welfare state (corporate, social, military).

Re: Obama Seeking to Boost Study of Human Brain

#49

Post by JoeAltmaier » Mon, Feb 18, 2013, 6:27 PM UTC

lars wrote: post #25

JoeAltmaier wrote: post #12

Sounds like snake oil. Too early to build a complete human brain model. Simpler brain models (cat, lizard) haven't been done yet (?) but that's not headline-worthy. How about "President announces program to solve all human suffering"? About as meaningless, but would actually be more useful if by some astronomical fluke they succeeded.

It's on the same timescale as the European Blue Brain Project (and also Kurzweil's predictions, if you put any faith in that). If you believe that this an information technology that has an exponential growth, it's not a sufficient objection that we're far off now. The same objections were raised for the Human Genome Project, which was far from completion most of its duration, but followed an exponential growth and c…

Genome is a couple megabytes. Brain is terabytes. A billion times more complex. And its not all about the chemistry, its about the physical wiring. So the same techniques won't apply.

Re: Obama Seeking to Boost Study of Human Brain

#50

Post by etrautmann » Mon, Feb 18, 2013, 6:31 PM UTC

jahewson wrote: post #14

I went to a fantastic talk by Microsoft Research's Stephen Emmott recently in which he referred to fMRI as "phrenology for the 21st century" and advocated the building of computational models of the brain. Mixed feelings about this announcement - some of the newer methods do look promising.

The proposed research is not focused on fMRI...