zlacker

Cubic millimetre of brain mapped at nanoscale resolution

submitted by geox+(OP) on 2024-05-09 21:36:06 | 520 points 199 comments
[view article] [source] [go to bottom]

NOTE: showing posts with links only show all posts
1. teuobk+c4[view] [source] 2024-05-09 22:12:13
>>geox+(OP)
The interactive visualization is pretty great. Try zooming in on the slices and then scrolling up or down through the layers. Also try zooming in on the 3D model. Notice how hovering over any part of a neuron highlights all parts of that neuron:

http://h01-dot-neuroglancer-demo.appspot.com/#!gs://h01-rele...

◧◩
5. talsit+S5[view] [source] [discussion] 2024-05-09 22:25:27
>>eminen+k5
Using a Microtome (https://en.m.wikipedia.org/wiki/Microtome).
◧◩
7. dekhn+q6[view] [source] [discussion] 2024-05-09 22:29:50
>>eminen+k5
The sample is stained (to make thigns visible), then embedded in a resin, then cut with a very sharp diamond knife and the slices are captured by the tape reel.

Paper: https://www.biorxiv.org/content/10.1101/2021.05.29.446289v4 See Figure 1.

The ATUM is described in more detail here https://www.eden-instruments.com/en/ex-situ-equipments/rmc-e...

and there's a bunch of nice photos and explanations here https://www.wormatlas.org/EMmethods/ATUM.htm

TL;DR this project is reaping all the benefits of the 21st century.

10. throwu+J7[view] [source] 2024-05-09 22:41:26
>>geox+(OP)
> The 3D map covers a volume of about one cubic millimetre, one-millionth of a whole brain, and contains roughly 57,000 cells and 150 million synapses — the connections between neurons.

This is great and provides a hard data point for some napkin math on how big a neural network model would have to be to emulate the human brain. 150 million synapses / 57,000 neurons is an average of 2,632 synapses per neuron. The adult human brain has 100 (+- 20) billion or 1e11 neurons so assuming the average rate of synapse/neuron holds, that's 2.6e14 total synapses.

Assuming 1 parameter per synapse, that'd make the minimum viable model several hundred times larger than state of the art GPT4 (according to the rumored 1.8e12 parameters). I don't think that's granular enough and we'd need to assume 10-100 ion channels per synapse and I think at least 10 parameters per ion channel, putting the number closer to 2.6e16+ parameters, or 4+ orders of magnitude bigger than GPT4.

There are other problems of course like implementing neuroplasticity, but it's a fun ball park calculation. Computing power should get there around 2048: >>38919548

◧◩◪
12. nickle+q8[view] [source] [discussion] 2024-05-09 22:46:06
>>jamiek+R6
Crow/parrot brains are tiny but in terms of neuron count they are twice as dense as primate brains (including ours): https://www.sciencedirect.com/science/article/pii/S096098221...

If someone did this experiment with a crow brain I imagine it would look “twice as complex” (whatever that might mean). 250 million years of evolution separates mammals from birds.

◧◩◪◨
15. nickle+E8[view] [source] [discussion] 2024-05-09 22:48:48
>>ignora+U7
I think you would also need the epigenetic side, which is very poorly understood: https://www.universityofcalifornia.edu/news/biologists-trans...

We have more detail than this about the C. elegans nematode brain, yet we still no clue how nematode intelligence actually works.

20. dekhn+ma[view] [source] 2024-05-09 23:07:56
>>geox+(OP)
Annual reminder to re-read "There's plenty of room at the bottom" by Feynman. https://web.pa.msu.edu/people/yang/RFeynman_plentySpace.pdf

Note the part where the biologists tell him to make an electron microscope that's 1000X more powerful. Then note what technology was used to scan these images.

◧◩◪
24. bglaze+5b[view] [source] [discussion] 2024-05-09 23:17:36
>>gibson+v9
Computation is really integrated through every scale of cellular systems. Individual proteins are capable of basic computation which are then integrated into regulatory circuits, epigenetics, and cellular behavior.

Pdf: “Protein molecules as computational elements in living cells - Dennis Bray” https://www.cs.jhu.edu/~basu/Papers/Bray-Protein%20Computing...

◧◩◪
25. falcor+9b[view] [source] [discussion] 2024-05-09 23:18:06
>>gary17+29
I appreciate you're running the numbers to extrapolate this approach, but just wanted to note that this particular figure isn't an upper bound nor a longer bound for actually storing the "state of a single human brain". Assuming the intent would be to store the amount of information needed to essentially "upload" the mind onto a computer emulation, we might not yet have all the details we need in this kind of scanning, but once we do, we may likely discover that a huge portion of it is redundant.

In any case, it seems likely that we're on track to have both the computational ability and the actual neurological data needed to create an "uploaded intelligences" sometime over the next decade. Lena [0] tells of the first successfully uploaded scan taking place in 2031, and I'm concerned that reality won't be far off.

[0] https://qntm.org/mmacevedo

◧◩◪◨
28. falcor+ac[view] [source] [discussion] 2024-05-09 23:25:40
>>dekhn+wa
> without it curing a senator's heart disease

Obviously I'm not advocating for this, but I'll just link to the Mad TV skit about how the drunk president cured cancer.

https://www.youtube.com/watch?v=va71a7pLvy8

◧◩
31. adamwo+Qc[view] [source] [discussion] 2024-05-09 23:30:44
>>g4zj+a7
https://hitchhikers.fandom.com/wiki/Total_Perspective_Vortex
◧◩
41. momojo+ag[view] [source] [discussion] 2024-05-10 00:01:40
>>blinco+Kd
Although the article mentions Artificial Intelligence, their paper[1] never actually mentions that term, and instead talks about their machine learning techniques. AFAIK, ML for things like cell-segmentation are a solved problem [2].

[1] https://www.biorxiv.org/content/10.1101/2021.05.29.446289v4.... [2] https://www.ilastik.org/

◧◩
54. skulk+On[view] [source] [discussion] 2024-05-10 01:24:34
>>brando+Lm
I agree, mostly because it's already being done!

https://www.youtube.com/watch?v=V2YDApNRK3g

https://www.youtube.com/watch?v=bEXefdbQDjw

◧◩◪◨
73. gary17+pR[view] [source] [discussion] 2024-05-10 08:10:07
>>falcor+9b
> we may likely discover that a huge portion of [a human brain] is redundant

Unless one's understanding of algorithmic inner workings of a particular black box system is actually very good, it is likely not possible not only to discard any of its state, but even implement any kind of meaningful error detection if you do discard.

Given the sheer size and complexity of a human brain, I feel it is actually very unlikely that we will be able to understand its inner workings to such a significant degree anytime soon. I'm not optimistic, because so far we have no idea how even laughingly simple, in comparison, AI models work[0].

[0] "God Help Us, Let's Try To Understand AI Monosemanticity", https://www.astralcodexten.com/p/god-help-us-lets-try-to-und...

◧◩◪◨⬒
80. ToValu+Hy1[view] [source] [discussion] 2024-05-10 14:08:33
>>Intral+4J
Do I need different prompts? These results seem sane to me. It interprets laser eye removal surgery as referring to LASIK, which I would do as well. When I clarified that I did mean removal, it said that the procedure didn't exist. It interprets Mid-Atlantic Mountain Range as referring to the Mid-Atlantic Ridge and notes that it is underwater and hard to access. Not that I'm arguing GPT-4 has a deeper understanding than you're suggesting, but these examples aren't making your point.

https://chat.openai.com/share/2234f40f-ccc3-4103-8f8f-8c3e68...

https://chat.openai.com/share/1642594c-6198-46b5-bbcb-984f1f...

◧◩
83. tim333+OB1[view] [source] [discussion] 2024-05-10 14:22:24
>>dekhn+ma
I think it's actually "What you should do in order for us to make more rapid progress is to make the electron microscope 100 times better" and the state of art at the time was "it can only resolve about 10 angstroms" or I guess 1nm. So 100x better would be 0.1 angstrom / 0.01 nm.

We have made some progress it seems. Googling I see "up to 0.05 nm" for transmission electron microscopes and "less than 0.1 nanometers" for scanning. https://www.kentfaith.co.uk/blog/article_which-electron-micr...

For comparison the distance between hydrogen nuclei in H2 is 0.074 nm I think.

You can see the shape of molecules but it's still a bit fuzzy to see individual atoms https://cosmosmagazine.com/science/chemistry/molecular-model...

◧◩◪◨⬒⬓
88. nickle+jS1[view] [source] [discussion] 2024-05-10 15:50:37
>>Animat+bb
Badly: https://www.lesswrong.com/posts/mHqQxwKuzZS69CXX5/whole-brai... (the comments have some updates as of 2023)

Almost every other cell in the worm can be simulated with known biophysics. But we don't have a clue how any individual nematode neuron actually works. I don't have the link but there are a few teams in China working on visualizing brain activity in living C. elegans, but it's difficult to get good measurements without affecting the behavior of the worm (e.g. reacting to the dye).

◧◩◪◨⬒⬓
89. Intral+w82[view] [source] [discussion] 2024-05-10 17:14:54
>>ToValu+Hy1
Tested with GPT-3.5 instead of GPT-4.

> When I clarified that I did mean removal, it said that the procedure didn't exist.

My point in my first two sentences is that by clarifying with emphasis that you do mean "removal", you are actually adding information into the system to indicate to it that laser eye removal is (1) distinct from LASIK and (2) maybe not a thing.

If you do not do that, but instead reply as if laser eye removal is completely normal, it will switch to using the term "laser eye removal" itself, while happily outputting advice on "choosing a glass eye manufacturer for after laser eye removal surgery" and telling you which drugs work best for "sedating an agitated patient during a laser eye removal operation":

https://chat.openai.com/share/2b5a5d79-5ab8-4985-bdd1-925f6a...

So the sanity of the response is a reflection of your own intelligence, and a result of you as the prompter affirmatively steering the interaction back into contact with reality.

◧◩
99. ninju+aC2[view] [source] [discussion] 2024-05-10 20:01:18
>>g4zj+a7
https://scaleofuniverse.com/en
127. blueno+1N3[view] [source] 2024-05-11 12:59:55
>>geox+(OP)
Neal Stephenson has a good novel that deals with this (https://en.wikipedia.org/wiki/Fall;_or,_Dodge_in_Hell)
152. NexusM+Tu4[view] [source] 2024-05-11 20:35:39
>>geox+(OP)
After reading through all comments as of 2024/05/11 I (as a professor at some major university) am quite surprised that not one single comment has asked the obvious question (instead of dishing out loads of (partial) "textbook knowledge" about brain functions, the difference between mammals and birds, AI and LLM etc.), which would be: what do all those strange structures and objects do which we know nothing about whatsoever? Have a look:

https://h01-release.storage.googleapis.com/gallery.html

I count seven.

◧◩
159. hotiwu+AC4[view] [source] [discussion] 2024-05-11 22:11:47
>>ein0p+Rl4
The reverse-engineering issue was popularized by this article, https://www.cell.com/cancer-cell/fulltext/S1535-6108%2802%29...

On the second point, the failure of Openworm to model the very well-mapped-out C. elegans (~0.3k neurons) says a lot.

179. ninash+cG5[view] [source] 2024-05-12 15:00:12
>>geox+(OP)
I developed algorithms for a Neuroglancer fork which annotates unlabeled volumetric data via self-supervised learning. I am looking forward to more developments in large scale brain mapping projects :)

More project details: https://www.ll.mit.edu/sites/default/files/other/doc/2023-02...

◧◩◪
185. dredmo+mV6[view] [source] [discussion] 2024-05-13 04:33:14
>>talsit+S5
NB: tome in Microtome has the same root as the T in CAT scan: computer aided tomography. Which is to say, slimly-sliced cabbage^W X-ray scans.

It's also the tome as in book, more properly one volume of a multi-volume (or multi-part) set, though it now generally simply means any large book.

<https://www.etymonline.com/search?q=tome>

◧◩◪◨
192. seesaw+mJ7[view] [source] [discussion] 2024-05-13 13:08:15
>>creer+oU2
A similar dataset already exists in mouse cortex. More are underway in the field.

https://www.biorxiv.org/content/10.1101/2024.03.22.586254v1

◧◩◪◨
196. habi+nX7[view] [source] [discussion] 2024-05-13 14:28:02
>>dredmo+mV6
https://en.wikipedia.org/wiki/Tomography: `The word tomography is derived from Ancient Greek τόμος tomos, "slice, section" and γράφω graphō, "to write" or, in this context as well, "to describe."`
◧◩◪◨⬒⬓⬔
199. RaftPe+e1k[view] [source] [discussion] 2024-05-17 14:12:32
>>nickle+jS1
Here's a timely bit of new research: https://www.science.org/doi/10.1126/sciadv.adk0002

Summary (my paraphrasing):

They partially figured out how two neurons (AVA, AVB) control forward and backward movement, previous theories assumed one neuron controlled forward and one controlled backward, but that didn't correctly model actual movement.

They found that AVA+AVB combine in a complex mechanism with two different signaling/control methods acting at different timescales to produce a graded shifting between forward+backward when switching directions, as opposed to an on/off type switch (that previous models used but didn't match actual movements).

Interesting learnings from this paper (at least for me):

1-Most neurons in worm are non-spiking (I had no idea, I've read about this stuff a lot and wasn't aware)

2-Non-spiking neurons can have multiple resting states at different voltages

3-Neurons AVA and AVB are different, they each have different resting state characteristics and respond differently to inputs

[go to top]