Gigantic New 3D Map Traces Every Neuron In a Tiny Mouse Brain (livescience.com) 35
Rick Schumann shares a report from Live Science: Researchers at the Allen Institute for Brain Science, a Seattle nonprofit dedicated to neuroscience, have been painstakingly recording every brain cell and every connection between those neurons in mice for the past several years. The result represents major progress since an earlier, simpler map they released in 2016. The now-complete map encompasses about 100 million cells, the institute reported in a paper published today (May 7) in the journal Cell.
Typically, researchers trace connections between brain cells using thin slices of tissue that can be imaged and explored layer by layer. To build a comprehensive, three-dimensional map, the Allen Institute team instead broke the mouse brain into "voxels" -- 3D pixels -- and then mapped the cells and connections within each voxel. The result comprises an "average" of the brains of 1,675 laboratory mice, to make sure the map was as standard as possible. [...] Mice are common "model organisms" in neuroscience. Their brains have fairly similar structures to humans', they can be trained, they breed easily, and researchers have already developed robust understandings of how their brains work. The hope is that the map will bring that understanding to a new level, the Allen Institute said. In doing so, neuroscientists will have a tool with which to develop new research programs and accelerate research already underway.
Typically, researchers trace connections between brain cells using thin slices of tissue that can be imaged and explored layer by layer. To build a comprehensive, three-dimensional map, the Allen Institute team instead broke the mouse brain into "voxels" -- 3D pixels -- and then mapped the cells and connections within each voxel. The result comprises an "average" of the brains of 1,675 laboratory mice, to make sure the map was as standard as possible. [...] Mice are common "model organisms" in neuroscience. Their brains have fairly similar structures to humans', they can be trained, they breed easily, and researchers have already developed robust understandings of how their brains work. The hope is that the map will bring that understanding to a new level, the Allen Institute said. In doing so, neuroscientists will have a tool with which to develop new research programs and accelerate research already underway.
Deep Thought vs the White Mice (Score:2)
An artificial mouse, with an artificial brain. Might be handy for autonomous pipe inspection drones.
My second thought was that it combined the multi-dimensional beings in Hitchhikers' Guide with the computer they built (Deep Thought)
Re: (Score:3)
"An artificial mouse, with an artificial brain. Might be handy for autonomous pipe inspection drones."
Sure but first we must wait for the artificial cheese to get invented.
Re: Deep Thought vs the White Mice (Score:2)
Re: (Score:2)
Sadly that was accomplished long ago.
Re: Deep Thought vs the White Mice (Score:2)
An artificial mouse, with an artificial brain.
The size of a planet. Literally.
Atomic-powered cyborg mice* - a la Snowcrash - make a lot more sense. Literally.
a*That dog was fucking badass.
You are Here. (Score:3)
!
Re: (Score:2)
Half the brain missing? One of their two brains is missing. The other can carry on quite nicely.
100 million cells... (Score:2)
It seems that thats the important bit, but its left out.
Re: 100 million cells... (Score:2)
That's the part that remembers well- known recipes
Why not just accurately describe reality instead? (Score:5, Insightful)
researchers have already developed robust understandings of how their brains work.
No, no they haven't.
If they did, we could emulate it. We can't, remotely. We "understand" no level of simplicity of consciousness.
Re: (Score:3)
Re: (Score:3, Informative)
The processing power required for such an effort would be staggering. Our most powerful supercomputers would be about as effective as a baby with an abacus. ASICs aren't going to be anywhere near enough.
Re: (Score:3)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
"regrow an organism from DNA"
DNA on its own doesn't contain the information needed for that.
-mrc
Re: (Score:3)
researchers have already developed robust understandings of how their brains work.
No, no they haven't.
Quite so. It is useful to compare this neuron map to one that has been available for a small marine work, C. elegans, for years now. It has 303 neurons exactly (and 959 cells total). Every connection between every neuron has been mapped, and the *type* of connection determined too.
So we must have a good computer model of this super simple brain right? No. We don't even have a good model of one of its neurons, which is itself a complex computational system.
And we are still trying to develop a good working de
Re: Why not just accurately describe reality inste (Score:2)
We "understand" no level of simplicity of consciousness.
Nor will we, at least not while we're metaphysically in this form; I dare say it'd be logically impossible,
Re: (Score:2)
If I understand the mechanics of a computer I'd have a robust understanding how they work. But that would be on transistor - interconnect level, there are more levels including the software layers. Consciousness would be software if there isn't quantum woo magic involved.
And of course (Score:3)
Neurons are all there is to thought and intelligence.
Re: (Score:3)
Well, mostly neurons. Glial cells do have a role to play in influencing them, and there are chemical signals other than just neurotransmitters. But mostly neurons.
Re: (Score:3)
Like everything else in Life, complexity is layered.
Neurons.
Dendrites.
Microtubules.
Functionality is overloaded in each case to conserve energy. The largest-scale phenomena happen at the smallest level; tuning is distributed.
Roger Penrose's work on this is probably the most advanced you'll find from a renowned academic.
Re: (Score:3)
Humanity is still in the process of understanding how a single neuron works let alone how a huge network of them operate. A good example is Lithium whose isotopes are identical chemically, but have a profound difference in effect on the brain. This is because it's not just the chemical reactions, but the speed at which reactions occur in its many interconnections to other neurons. Even a single neuron and its inter
Missing Disclaimer (Score:1)
I'd like to see them do this and then say, "None of the 1675 mice were harmed in making the 3D map."
so how in the world... (Score:2)
how in the world did they convince rsilvergun to sit still and let them perform this test? ;-)
Why I felt this was an important story to submit: (Score:2)
People who read my comments on Slashdot know very well that I'm against so-called 'self driving cars', and ridicule so-called 'AI' as well, for one simple reason: the 'AI' behind SDCs is inherently, fatally flawed: It has no ability to reason, no cognitive ability whatsoever. Therefore the output you get from it is unpredictable in that if it's nonsensical (in a way that a human bra
Re: (Score:2)
AIs are driving well enough and RIs are driving poorly enough that it's not clear that your hypothesis is actually true. RIs are fatally flawed in that they have ego, and frequently prioritize this ego over basic rules of safety. RIs are also frequently multitasking, and prone to mis-prioritize rules for one task due to problems with unrelated tasks.
While AIs are not self-training the way we like, they are built to work without ego, and thus never get angry or vengeful. They are also physically built with s
Re: (Score:2)
Journalism (Score:1)
So far so good.
> Traces Every Neuron In a Tiny Mouse Brain
Nope, not even close. This is an atlas of mouse brain _regions_. The atlas does not contain information about individual neurons or their connections. The Allen Institute has done other neuron tracing experiments, but those have only sampled a small fraction (on the order of 1/1000) of the neurons in the brain, and in any case they are not included in this atlas.
"Tiny Mouse Brain" (Score:1)
What about non-tiny mice, like those dog-sized rats seen in the cafeteria.
Re: (Score:2)
I take it you don't like the other patrons?
Seems like only yesterday (Score:2)
They were so proud to have simulated a worms nervous system, then a fruit fly, and now a mouse brain.... what a great age to be witness to!
Re: (Score:1)
You may want to hold your horses. I believe this is only a visual map and not a complete simulation of a brain.
The purpose of the map may possibly lead up to something of it. How much this will help us understand the human brain, if it's worth killing 1675 mice for some high-resolution map, or if we could have waited for better tech first (to only use a couple of mice for making this map) ... is a different matter.
Trippy Video Tour of Mouse Brain (Score:2)