Skip to playerSkip to main content
  • 1 day ago
Mapping_the_Mind

Category

📚
Learning
Transcript
00:00Today, we're going on one of the most incredible journeys imaginable.
00:04Not to some distant star, but into the space between our own ears.
00:09A project between Harvard and Google has just given us a map of the human brain like we've never seen
00:15before.
00:15And it really makes you wonder.
00:17You know, it's a wild thought, right?
00:19We've got these unbelievable telescopes that can see to the edge of the universe.
00:23But the thing sitting right inside our own skull is still, in many ways, a complete mystery.
00:28And the reason for that all comes down to scale.
00:31The real action in the brain, the stuff that makes us us, happens at a level so tiny it's almost
00:37impossible to see.
00:38I mean, sure, we've all seen brain scans like MRIs, and they are amazing.
00:44They show us which big neighborhoods of the brain light up when we're happy or solving a puzzle.
00:49But what they can't do is show us the actual streets and houses.
00:53They can't see the individual wires, the synapses, where all the information is actually passed along.
00:59And that's where the code of consciousness is written.
01:01To actually read that code, you have to zoom in.
01:04I mean, way in.
01:06We're talking nanoscale.
01:08We need a map that shows every last wire and every single connection.
01:12And for that, you need a totally different kind of tool.
01:15An electron microscope.
01:17Which lets us see things that are smaller than a wavelength of light.
01:19And that brings us to the HO1 project.
01:23This was this massive effort to finally create that map.
01:27A team from Harvard and Google got their hands on a tiny piece of human brain tissue from the temporal
01:32cortex.
01:33And they set out to map every single thing inside of it.
01:36And when I say tiny, you won't believe how small it was.
01:40The whole sample was just one cubic millimeter.
01:43That is about the size of a single grain of sand.
01:46Now, it came from a patient who was having surgery for epilepsy, which is an important detail.
01:51But pathologists confirmed the tissue itself was structurally normal.
01:55Which means this is the clearest, most detailed look we have ever had at our own wiring.
02:01Okay, so what's inside that grain of sand?
02:04Well, for starters, the team found over 57,000 individual cells.
02:09You've got your neurons, your glial support cells, blood vessels, the whole shebang.
02:13It's like a whole city packed into this microscopic space.
02:17Just let that sink in for a second.
02:19But here's the number that just blows my mind.
02:21The synapses.
02:23The connections.
02:24Inside that one cubic millimeter, they counted nearly 150 million of them.
02:30These are the little junctions that make the whole circuit work.
02:33150 million connections.
02:35In a piece of brain, you could barely even see.
02:37So, what kind of hard drive do you need to store a map of a single grain of sand's worth
02:43of brain?
02:43Well, it turns out you need 1.4 petabytes.
02:47To give you some perspective, one petabyte is 1,000 terabytes.
02:51This data set 150 years.
02:52All for that one tiny speck of tissue.
02:55So the big question is, how on earth did they do it?
02:58I mean, creating this digital universe from a grain of sand was an unbelievable challenge.
03:03It took a mix of super precise biology, powerful microscopes, and just a massive amount of computing power.
03:10It was like trying to solve a puzzle with trillions of microscopic pieces.
03:15Here's the basic idea of how they pulled off this incredible feat.
03:19First, they took that tiny sample and sliced it into more than 5,000 slivers, each one thinner than a
03:26wavelength of light.
03:27Then, a custom microscope took pictures of every single slice, creating millions of little image tiles.
03:32After that, some seriously powerful software had to stitch all those millions of images together into one solid 3D block.
03:40And then came the really amazing part.
03:42They unleashed AI algorithms to trace every single cell through that entire digital maze.
03:48And this slide just shows you how tricky that alignment step was.
03:51See the images on the left?
03:53That's the raw data.
03:54It's all shaky and misaligned, like a messy stack of photos.
03:57But on the right, after the software works its magic, you get this perfectly smooth, continuous block of tissue.
04:04Now just imagine doing that for 196 million pictures to build that final 1.4 petabyte model.
04:10It's just incredible.
04:12Now this is where the AI becomes the hero of the story.
04:16You see, trying to trace 57,000 cells by hand through thousands of images would be, well, it would be
04:22impossible.
04:23It would take centuries.
04:24So they used this clever AI called a flood-filling network.
04:27You can see it working here.
04:29The black and white image is the raw microscope data.
04:32And the colored parts are where the AI has automatically identified and filled in each cell, creating a perfect 3D
04:38model of everything inside.
04:39Okay, so building the map was a technological marvel.
04:43But the real excitement, the reason they did all this, is what they found inside.
04:47This map is so new and so detailed, it's already rewriting textbooks and showing us things we've never, ever seen
04:54before.
04:54Let's check out some of the highlights.
04:56So first off, just taking a basic counthead of what's in there lets us some big surprises.
05:01For one thing, the brain support cells, the glia, actually outnumber the neurons 2 to 1.
05:06And get this, the most common type of cell wasn't a neuron at all, but something called an oligodendroglade, which
05:12makes the insulation for the brain's wiring.
05:14Oh, and they also found a bunch of weird blobs that they couldn't identify at all.
05:17But then things got even weirder.
05:20Deep down in a part of the cortex called layer 6, they found these bizarre neurons.
05:25Take a look at the image here.
05:26They call them compass neurons.
05:28See how they each have one really big branch that points in one of two opposite directions?
05:33They form these perfectly mirrored groups.
05:36What are they for?
05:37Right now, nobody has a clue.
05:39This quote is pulled straight from the scientific paper.
05:42And it hints at what might be the most profound discovery of all.
05:47The scientists started noticing that some of the connections in the brain didn't look random.
05:52At all.
05:53They looked, well, they looked like they were there on purpose.
05:59And this is exactly what they meant.
06:01So most neurons connect to each other with maybe one or two synapses.
06:05It's a pretty weak link.
06:06But the team found these rare, incredibly powerful connections.
06:10Check out the image here.
06:11You can see a single axon reaching out and making 53 separate connections to one single partner neuron.
06:17The odds of that happening by chance are basically zero.
06:20It suggests the brain isn't just a random web.
06:23It's deliberately forging these super connections between specific partners.
06:27So after years and years of work, this first incredible exploration is done.
06:32But here's the best part.
06:34This isn't the end of the story.
06:36It's really just the beginning.
06:38Because the team behind this map didn't just publish their findings and call it a day.
06:42They did something much, much bigger.
06:44They've made the entire thing, all 1.4 petabytes of it, totally free and available for anyone in the world
06:51to explore online.
06:52They even built tools so you can fly through it yourself.
06:55This isn't just a research paper.
06:57It's a new atlas for neuroscience.
06:59It's a gift to science that's probably going to lead to discoveries for decades.
07:03And that just leaves us with this final kind of humbling thought.
07:06But this entire journey, all of these wild discoveries, came from a piece of our brain no bigger than a
07:13pinhead.
07:14So it just makes you wonder, what other beautiful, bizarre, and unimaginable secrets are hiding in the billions of other
07:21millimeters that make up the rest of the human brain?
Comments

Recommended