Chapter 10: Vision: From Eye to Brain

Welcome to Last Minute Lecture.

This free chapter overview is designed to help students review and understand key concepts.

These summaries supplement not replaced the original textbook and may not be redistributed or resold.

For complete coverage, always consult the official text.

Welcome back to the Deep Dive.

Today we're doing something a little different.

Usually we look at the world around us, you know, economics, history, tech, but today we're looking at the

world around us.

We are indeed.

We're turning the camera around and examining the lens itself.

Which is a pretty fitting metaphor, actually, because we're diving into chapter 10 of behavioral neuroscience.

The topic is vision.

Specifically, vision from eye to brain.

And it's one of those topics where what you think you know is probably just the tip of the iceberg.

Oh, absolutely.

I have to say, going into this reading, I thought I had a decent handle on how eyes work.

You know, it's a camera.

Light goes in and image gets taken.

The brain sees it.

Simple, elegant,

and well,

almost completely wrong.

Painfully wrong.

I mean, the complexity here is honestly staggering.

We're talking about quantum physics turning into chemistry, which then turns into electricity, which then turns into, well, our reality.

It's probably the most complex biological task you will ever perform.

And you're doing it right now just by listening to this.

A massive portion of your brain, a huge chunk of cortical real estate, is dedicated solely to making sense of light.

So to kick this off, we're not starting with anatomy diagrams.

We're starting with a mystery.

A story that I think really exposes just how strange, how profoundly weird this whole system is.

The story of patient DF.

It's a classic case study.

It really blew the field of visual neuroscience wide open back in the early 90s.

And it just challenges everything we intuitively believe about what it even means to see.

So set the scene for us.

Where does the story begin?

It begins with a tragedy, unfortunately.

We have a young woman referred to only as DF in the medical literature to protect her privacy.

She's taking a shower one day.

It was a cold day.

The windows were all closed and she had a gas heater running in the bathroom to keep warm.

Oh, I think I know where this is going.

Not good.

Carbon monoxide.

It's colorless, odorless.

The heater was functioning and she had no idea.

The gas was just flooding the room, binding to the hemoglobin in her blood and starving her brain of oxygen.

And someone found her.

Her husband found her unconscious on the floor.

She was rushed to the ER and for a while it was really touch and go.

But she does survive.

She survives.

She wakes up and, you know, initially the doctors are relieved.

She can speak perfectly fine.

She can move her arms and legs.

She could feel a hand on her shoulder.

A memory completely intact.

So on the surface she seems okay.

The basic functions are there.

Exactly.

Until she opens her eyes.

She looks around the hospital room and she can't tell you what anything is.

What do you mean?

Like blurry?

Not blurry.

It's more profound than that.

She looks at her husband's face, a face she's known for years, and it's just a landscape of shapes to her.

She knows it's a face because it has the right parts.

Eyes, a mouth, a nose, but she has no idea who it is.

So she has face blindness, prosopagnosia.

It went way, way beyond faces.

If you held up a pencil, she couldn't name it.

She'd say a long, thin object.

If you held up a flashlight, she might guess it was a kitchen utensil because it was shiny and metallic.

She had lost the ability to visual recognize objects.

The term for that is visual agnosia.

Next one, visual agnosia.

Okay, so in my mind that means she's functionally blind.

If you can't identify a flashlight, for all intents and purposes, you can't see the flashlight.

That's the intuition.

Absolutely.

But this is where the researchers, Mel Goodale and his team,

stepped in and found something that just,

it shouldn't be possible.

They did what's famously called the slot experiment.

This was the part of the reading that made me just stop and stare at the wall for a minute.

It really does break your brain a little.

So here's the setup.

They hold up a disc and it has a slot cut into it, like a mail slot, and they can rotate this disc to any orientation, vertical, horizontal, diagonal, whatever.

Simple enough.

And they ask DF, tell us the orientation of the slot.

Is it vertical?

Is it horizontal?

And she can't do it.

She has no idea.

She's just guessing.

Her performance is at random chance.

She literally cannot consciously perceive the angle of the slot.

If you ask her to draw it, she can't do that either.

Okay, so she's blind to orientation.

That part of her vision is gone.

But then, and this is the incredible twist, they hand her a card, like a letter, and they say,

don't tell us what the angle is, just put the card through the slot.

And she does it.

She does it perfectly, instantly, without any hesitation.

She rotates her wrist to the exact correct angle before her hand even reaches the disc and just slides it right in.

Wait a second, let me process that.

She doesn't know what the angle is, but her hand knows what the angle is.

That's it.

Exactly.

Her conscious mind is blind, but her motor system can see perfectly.

They did it with other objects, too.

She couldn't tell you that the object in front of her was a uniquely shaped rock.

But if you asked her to pick it up, her fingers would pre -shape into the perfect grip size and configuration to grab it.

That creates such a profound disconnect.

It just shatters the idea that seeing is one single process.

It's the smoking gun.

It implies that we have at least two systems operating in parallel, one for knowing what's out there and another completely separate one for doing things with what's out there.

And in D .F.'s case, the phone line between them was cut.

Precisely.

Her body could see, but her conscious mind was in the dark.

And that is our mission for today.

We are going to trace that whole path.

We're going to try and figure out how the brain pulls off that incredible magic trick.

We are going to follow a single photon of light from the moment it hits the surface of your eyeball through the bizarre counterintuitive wiring of your retina and then along the distinct neural highways in the brain that separate the what from the where.

And I think it's important to frame this whole discussion with a waterfall concept that was mentioned right at the start of the chapter.

The visual waterfall.

I like that term.

It's this idea that we are constantly being bombarded.

It's an almost unimaginable amount of data.

If you actually think about the resolution of your eye and the number of photo receptors and the fact that it's refreshing constantly, it's like trying to drink from a fire hose.

So if your brain actually tried to process every single bit of visual data that hit your eye, what would happen?

You'd crash.

Your consciousness would just be overwhelmed.

You'd probably be catatonic, just locked up, trying to process the sheer volume of information.

So the story of vision isn't really a story of seeing everything.

It's a story of filtering.

It's compression.

It's aggressive editing.

It's taking this absolute flood of reality and stripping it down to the bare essentials that matter for survival.

Is there food?

Is there a mate?

Is there a tiger in the bushes?

That's it.

Biology fundamentally doesn't care if you can appreciate the subtle breaststrokes on a Monet.

It cares if you can survive the next five minutes.

Okay.

That's a perfect frame.

So let's keep that in mind.

Every mechanism we're about to discuss is an evolutionary hack designed for survival.

So let's start at the very beginning.

The instrument itself,

the eye.

It's almost a cliche at this point to compare the eye to a camera, but I mean, the structural similarities are pretty undeniable, aren't they?

They really are.

You have a dark chamber, which is what camera actually means in Latin, camera obscura.

You have an aperture to control the amount of light coming in, a lens to focus it, and a light sensitive surface at the back to record the image.

But let's dismantle that biological camera because the biological version is a lot squishier and frankly a lot weirder than a Nikon.

Okay.

Let's start with the very front element, the cornea.

The cornea is that clear dome -shaped tissue at the very, very front of your eye.

And here's a fun pop quiz question that catches a lot of first -year med students.

Which part of the eye does most of the focusing?

Well, my intuition immediately says the lens.

I mean, it's called the lens.

That's its job.

And your intuition would be wrong.

It's the cornea.

The cornea provides the vast majority of the eye's refractive power, something like two -thirds of it.

Really?

Why is that?

It all comes down to a principle in physics.

The change in the refractive index,

refraction, the bending of light,

happens most dramatically when light travels from one medium to another with a very different density.

Like light going from air into water.

Exactly.

The jump from air to the fluid -filled tissue of your cornea is a huge change in density.

That bends the light sharply.

But the jump from the fluid inside your eye to the lens itself, that's a much smaller change in density.

So the lens does less bending.

Okay.

So the cornea is the heavy lifter.

It's like the fixed primary focus lens.

So what is the actual biological lens doing then?

The lens is the fine -tuner.

Think of the cornea as getting the image in the right ballpark.

But if you want to switch your focus from looking at a mountain in the distance to looking at the text on your phone, you need to adjust that focus.

In a camera, you do that by twisting the barrel, and the glass element physically moves back and forth.

Right.

But you can't really do that inside an eyeball.

There's no room to slide a lens back and forth.

So mammals evolved a much cleverer mechanism.

We don't move the lens, we change the shape of the lens.

This is the process of accommodation.

Yes.

Your lens is this flexible elastic little thing.

It's like a tiny clear bag of jelly.

And it's suspended by these tiny ligaments called zonial fibers, which are attached to a ring of muscle called the ciliary muscle.

And this is where the mechanics get really counterintuitive, I thought.

When I flex a muscle, like a bicep, I expect things to get tight and pull on something.

You would think so.

Oh.

But in the eye, it's basically reversed.

When your ciliary muscle is relaxed, the ring of muscle is wide.

That pulls the fibers tight, which in turn stretches the lens out, making it flat.

A flat lens is for distance vision.

A flat lens is for distance vision.

So relaxing your eye muscles equals seeing things that are far away.

Okay.

So relaxing equals seeing far.

What about seeing up close?

To look at something close, like the book in your lap, the ciliary muscle has to contract.

When it contracts, the ring of muscle gets smaller.

That releases the tension on the fibers.

And the lens, being naturally elastic, just sort of pops into a rounder, fatter shape.

And a fatter, more curved lens bends light more sharply.

Precisely.

So that rounder lens allows you to focus on things that are close up.

Which means looking at something close requires active, continuous, muscular effort.

It is a constant workout.

And that's exactly why staring at a screen six inches from your face all day gives you a headache and eye strain.

Your ciliary muscles are basically doing a bicep curl for eight hours straight.

And this of course leads us to the sad reality of aging.

The arm's length menu reading phenomenon.

Presbyopia.

It happens to literally all of us.

So why does that happen?

Does the ciliary muscle just get weak over time?

The muscle itself stays reasonably fine, for the most part.

The problem is the lens itself.

The material fails.

Over decades, new layers of cells are added to the lens, kind of like tree rings.

It gets denser and denser.

It loses that youthful elasticity.

So it turns from a soft jelly into more of a hard rubber?

It's a perfect analogy.

So the ciliary muscle can contract all it wants, releasing the tension.

But the stiff old lens just refuses to bulge.

It stays flat.

So you lose your ability to focus up close.

And that's when we have to go by reading glasses.

It's a mechanical failure of the material, not the muscle.

It is.

Just simple wear and tear.

Okay, so we've got the light focused.

But before it even hits the lens, it has to pass through the aperture.

The pupil.

Which isn't really a thing in itself, is it?

It's just a hole.

The structure that matters is the iris.

The colored part of the eye?

Right.

The iris is essentially a sphincter muscle.

And what's fascinating here is that you have absolutely zero conscious control over it.

It is entirely run by your autonomic nervous system.

This is that constant battle between fight or flight and rest and digest.

Exactly.

The sympathetic nervous system, that's your fight or flight system, it wants as much information as possible.

It screams dilate.

It rips that aperture wide open to let in every single available photon because you need to spot the potential danger lurking in the dark bushes.

And the parasympathetic system does the opposite.

It chills everything out.

That's your rest and digest system.

It constricts the pupil.

It says we're safe.

We're in bright light.

We don't need to be blinded by the sun.

So let's close down the shutters a bit for a sharper image.

There is a fantastic historical nugget in the text about this.

The Belladonna plant.

Oh, yes.

The beautiful lady.

Italian, right.

Belladonna.

That's the one.

So back in the Renaissance and even much later, having widely dilated pupils was considered the height of attractiveness.

It's a biological signal of arousal, of excitement, of interest.

If someone looks at you and their pupils dilate, your brain subconsciously registers that as, oh, they like what they see.

So people were trying to hack this biological signal.

They were.

Women would use extracts from the deadly nightshade plant, whose scientific name is Atropa belladonna, as eye drops.

Which is, as the name implies, highly poisonous.

Very much so.

But it contains a chemical called atropine.

And atropine is an acetylcholine antagonist that blocks acetylcholine receptors.

And acetylcholine is the neurotransmitter that the parasympathetic system uses to tell the pupil to constrict.

Exactly.

So if you block that signal, you essentially paralyze the close command.

The finger muscle of the iris relaxes, the pupil blows wide open, and you get that doe -eyed seductive look.

And you probably also get extremely blurry vision and a nasty headache.

And you risk systemic poisoning if you absorb too much.

But, you know, fashion.

Fashion is pain.

Okay, moving on.

We have the light through the cornea, through the pupil, focused by the lens.

And we have three pairs of extraocular muscles moving the whole eyeball around.

Now, we finally arrive at the destination.

The screen at the back.

The retina.

The retina.

This is the magical place where physics becomes biology.

Up until this point, it's all just been optics bending light beams.

Now, that light has to become a neural signal.

And the structure of the retina, I have to ask, why on earth is it backward?

It is the great, and I mean great, design flaw of the vertebrate eye.

If you look at figure 10 .2 in the text, you can see the different layers very clearly.

Explain this for our listeners, because it really does defy all logic.

So if you were an engineer, and you were designing a digital camera sensor, you would put the light -sensitive pixels at the very front, facing the lens, and then you would run the wires out the back.

Of course.

That's the only sensible way to do it.

The vertebrate retina is the exact opposite.

The actual light -sensitive cells, the photoreceptors, the rods and cones, are buried at the very, very back of the tissue pressed up against the wall of the eye.

The output wires, the ganglion cells whose axons become the optic nerve, are at the front.

So you're telling me the light has to pass through all the circuitry and wiring before it gets to the sensor?

Yes.

The incoming light has to travel through layers of ganglion cells, bipolar cells, amicrine cells, horizontal cells, and a network of blood vessels before it finally hits the rods and cones that are supposed to be detecting it.

Doesn't that degrade the image?

It seems like it would be a blurry mess.

It absolutely does.

It scatters the light, it blurs things.

The brain has to do a tremendous amount of computational work later on to clean up that messy signal.

But why?

Why did evolution do this?

It seems so suboptimal.

Well, it's interesting because not all eyes are like this.

Cephalopods sow octopuses and squid.

They have it the right way around.

Their receptors are at the front.

But somewhere way back in our evolutionary lineage, the way the neural tube folded in on itself during development just, it resulted in the retina being put in backward.

And evolution is a tinkerer, not an engineer.

Once a solution works well enough to keep you alive, it tends to stick.

Good enough for government work, as they say.

All right, let's talk about those actual sensors then, the rods and the cones.

It's really two separate visual systems working in parallel here, isn't it?

It is.

You have the scotopic system and the photopic system.

Scotopic is the rods.

Right.

An easy way to remember it is S for shadow or scoto, which means darkness.

Rods are your night vision specialists.

They're unbelievably sensitive.

So sensitive, in fact, that a single photon, the smallest possible quantum unit of light can trigger a response in a rod cell.

A single photon, that's insane.

It is.

But there's a trade -off.

To get that incredible sensitivity, they have to sacrifice two things, resolution and color.

Rods are completely colorblind.

That's why at night, in very dim light, the world appears to be in shades of gray.

They use a pigment called rhodopsin that is just incredibly volatile to light.

And on the other team, we have the cones.

The photopic system.

These are the divas of the retina.

They need a lot of light to wake up and do their job.

They're completely useless in the dark.

But in return for that highlight requirement, they give you two amazing things.

Full color vision and sharp high QED detail.

And their distribution across the retina is very different, right?

They aren't just mixed in together evenly.

Not at all.

The cones are packed incredibly densely into a tiny, tiny pit right in the center of the retina.

It's called the fovea.

And the text mentions something really cool about the fovea.

It's nature's workaround for the backward retina problem.

It is.

It's a brilliant little hack.

At the fovea, the layers of blood vessels and all those wires, the ganglion bipolar cells, are physically pushed to the side.

It's like the cells part to open a window.

Like opening a curtain.

Exactly.

So at that one tiny critical spot,

the incoming light gets a direct unobstructed path to the cones.

This is why when I want to see something very clearly, like reading small text, I have to look directly at it.

You are pointing your fovea at it.

That's the only high -resolution part of your vision.

Your peripheral vision is dominated by rods.

It's blurry.

It's color -weak.

But it's fantastic at detecting motion.

Which makes perfect evolutionary sense.

If I'm in the jungle focusing on picking berries, which is fovea work, I need my peripheral vision to notice the flicker of motion of a tiger creeping up on my side.

That's rod work.

That's the exact division of labor.

Now, we have to get a little technical for a moment.

Because the way these cells actually fire, it's the most Alice in Wonderland, topsy -turvy part of neuroscience.

This is the chemistry of transduction.

I had to read this section three times.

Because it's backward too, just like the retina's structure.

It's completely backward from every other neuron in your body.

Okay, walk us through it.

Normal neuron logic is a stimulus happens, the cell gets excited, it depolarizes, and it fires an action potential.

Right.

You poke a sensory cell in your skin, sodium channels open, it fires a signal that says ouch.

But photoreceptors, rods and cones.

Imagine a rod cell sitting in absolute total darkness.

You would think it would be silent resting, it's not.

It is working its little tail off, it is constantly depolarized.

Sodium channels are held open, and sodium is just pouring into the cell.

We call it the dark current.

And because it's depolarized, it's constantly releasing neurotransmitter.

Constantly.

In the dark, the cell is continuously dumping glutamate into the synapse.

It's essentially shouting darkness, darkness, darkness to the next cell in the chain.

The on switch is taped down.

Okay, and then a single photon of light comes flying in and hits it.

That photon hits the pigment molecule inside the rod, which is rhodopsin.

Rhodopsin absorbs the photon's energy and instantly changes its shape.

This is the trigger for everything that follows.

It's like a key turning in a lock.

A very fast key.

That shape change in rhodopsin activates a G protein called transducin.

Transducin, okay, I'm falling.

Transducin then finds and activates an enzyme called phosphodisterase, or PDE.

This sounds like a Rube Goldberg machine made of molecules.

It is a chemical Rube Goldberg machine.

Yeah.

And it's designed for amplification.

So this PDE enzyme swings around, and its job is to chew up a molecule called cyclic GMP, or CGMP.

Now, CGMP is the crucial part.

It was acting like a molecular doorstop, propping the sodium channels open.

So the enzyme comes along and destroys the doorstop.

And the sodium channels slam shut.

And when those positive sodium ions start flowing in...

The cell hyperpolarizes.

It becomes more negative inside, and crucially, it stops releasing glutamate.

So the signal for light is silence.

The signal is a reduction in noise.

The brighter the light, the quieter the cell becomes, the less neurotransmitter it releases.

That seems unbelievably convoluted.

Why not just have a system that fires when it's hit by light?

The reason is that massive amplification I mentioned.

That chemical cascade is like a signal booster.

One single photon hits one molecule of rhodopsin.

But that one molecule can activate hundreds of transducent molecules.

And those hundreds of PDE enzymes can break down thousands of CGMP molecules.

So it turns a tiny, single -photon whisper of light into a massive, unmistakable cellular event.

It's a biological high -gain amplifier.

It's what gives you that incredible night vision.

Okay, so we have the signal.

The rods and cones have gone silent, or at least quieter, which tells the next layer of cells, the bipolar cells, hey, we saw something.

And then the bipolar cells talk to the ganglion cells.

But we need to pause here for a second, because there are other cells involved that do something incredibly important.

The amicrine and horizontal cells.

Exactly.

These cells allow for lateral communication.

The cells in the retina aren't just independent pixels in a camera.

They talk to their neighbors.

Specifically, they gossip.

They tell their neighbors to shut up.

This is lateral inhibition.

It is.

The basic concept is that an active neuron tends to inhibit its immediate neighbors.

If cell A is getting blasted with bright light, it sends a strong signal forward to the brain.

But it also sends an inhibitory signal sideways to cell B next to it, saying, I'm very active, so you need to be less active.

Why would it do that?

What's the point?

Contrast enhancement.

Imagine you are looking at a sharp edge, a black bar on a white background.

The cells looking at the middle of the white page are all active, and they're all inhibiting each other a little bit.

OK, they're all kind of holding each other down.

Right.

But now look at the cell right on the border, just inside the white area.

It's getting inhibition from its neighbors in the white area.

But on its other side, its neighbors in the black area, that neighbor is quiet.

It's not sending any inhibition.

Oh, so the cell right on the edge gets less total inhibition than the cells in the middle of the white patch.

Exactly.

So with less inhibition holding it back, it fires harder.

Your brain perceives that edge as being even brighter than the rest of the white area.

And the reverse happens on the dark side of the edge.

Yes.

The cell just inside the black border gets extra inhibition from its very active, bright neighbor.

So your brain perceives that edge as being even darker than the rest of the black bar.

This is what creates those optical illusions called Mach bands, right?

Where you see these faint stripes of extra brightness and darkness at edges that aren't physically there.

If you look at figure 10 .7 in the book, you can see this effect very clearly.

Your retina is literally doing calculus.

It's calculating the second derivative to exaggerate boundaries, to make objects pop out from their background.

So we are already lying about reality before the signal has even left the eyeball.

We are enhancing reality for our own good.

Fair enough.

Okay, so the final process signal leaves the eye via the axons of the ganglion cells.

And this big bundle of wires is the optic nerve.

And because this massive cable has to physically punch a hole through the back of the retina to exit the eye, we're left with the blind spot, the optic disc.

Right.

An area with absolutely no photoreceptors, just cable.

But I don't see a black hole in my vision right now.

No, because your brain is a very helpful and slightly deceitful editor.

It sees the blank spot in the data stream and it just fills it in.

It takes the information from the surrounding area, the color, the texture, and essentially hallucinates the missing patch so that you have a smooth, continuous visual field.

Incredible.

So now we are really leaving the eye.

We are traveling down the optic nerve, heading deep into the brain.

And we hit the first major intersection, the optic chiasm.

The Greek letter chi, which looks like an X, this is the crossover point.

Now, the pop science fact that everyone thinks they know is your left eye goes to your right brain and your right eye goes to your left brain.

And as a neuroscientist, every time I hear that, a little piece of my soul dies.

It is not eye to brain.

It is visual field to brain.

This is a crucial distinction.

Let's break this down very carefully because it's so important.

Okay, do this with me.

Close your left eye.

Just look with your right eye.

You can see your nose over on the left side of your vision, right?

Yes, on the nasal side of my vision.

And you can also see things off to your right, towards your ear, on the temporal side of your vision.

Correct.

So one eye sees both the left and right sides of the world?

Exactly.

But the brain wants to organize everything by field of view.

It wants all the information from the left world, everything to the left of your nose, to go to the right hemisphere of your brain for processing.

And voice versa.

So how does it sort the signals?

It uses the anatomy of the retina.

The retina in each eye is split in half.

You have the nasal hemorrhantina, the half closer to your nose, which sees the outside edge of the world.

And you have the temporal hemorrhantina, near your temples, which sees the center of your vision.

So at the chiasm?

The fibers from the nasal hemorrhantina, the ones looking outward, cross over to the opposite side of the brain.

The fibers from the temporal hemorrhantina, the ones looking inward, stay on the same side.

They do not cross.

The result being that the left hemisphere of your brain gets a complete stitched together map of the entire right visual field composed of signals from both the left eye and the right eye.

Perfect.

After the chiasm, that bundle of nerves is now called the optic tract.

And its first major stop is the thalamus.

Specifically, the LGN.

The lateral geniculate nucleus.

This is the main sensory relay station or switchboard for vision.

Most of the axons from the retina terminate here.

And then from the LGN, a huge fan of fibers called the optic radiation projects back to the cortex to V1, the primary visual cortex.

Also called the striate cortex.

We have finally arrived at the back of your head in the occipital lobe.

This is where seeing, in the sense we usually think of it, really begins.

Why is it called striate?

Does that mean something?

Striate just means striped.

If you take a slice of this part of the brain and look at it under a microscope, you can see a prominent white stripe of myelinated axons in walled out layers, layer IV specifically.

That visible stripe is the massive bundle of incoming axons arriving from the LGN.

It's the highway off -ramp for all that visual data.

And the map here is interesting.

It's a direct map of the retina, right?

A retinotopic map.

Yes.

It's a point -for -point map.

If you stimulate a specific spot on the retina, a specific corresponding spot in V1 will light up.

But the scale of that map is wildly distorted.

This is cortical magnification.

It is.

We're back to the fovea.

The fovea is a tiny, tiny part of the retinalis than 1 % of the total area.

But in the visual cortex,

nearly 50 % of the entire tissue is dedicated to processing the input that comes just from the fovea.

50 % of your visual brain power is focused on that one tiny dot in the center of your vision.

It just shows you what the brain truly prioritizes.

It doesn't care that much about what's happening in the periphery.

It cares deeply about high -resolution detail right in the center of your gaze.

It's like having a gigapixel camera for the center of your view and an old blurry potato camera for the edges.

And this part of the brain is plastic.

The text brought up the example of video games.

Yes.

Finally, some validation for gamers everywhere.

So you're telling me I'm actually improving my brain by playing first -person shooters.

You actually are, to a degree.

Studies have shown that playing fast -paced action video games, which require intense central focus, peripheral awareness, and very fast reaction times,

can actually sharpen the processing in these cortical circuits.

Players get better at contrast sensitivity and can allocate their visual resources more efficiently in V1.

I am absolutely telling my partner that tonight.

I'm not gaming, dear.

I'm engaging in cortical plasticity therapy.

Let me know how that goes for you.

So we are in V1.

The signal has arrived.

What are these individual neurons actually looking for?

They aren't just seeing light dots from the retina anymore, are they?

No, not at all.

This is the Nobel Prize -winning work of David Hubel and Torsten Meisel back in the 1960s.

They discovered that neurons in the cortex are incredibly picky.

They have complex receptive fields.

Tell us about the famous cat experiment.

It's a legendary story in neuroscience.

They had an esatized cat and they were recording from a single neuron in its visual cortex.

They were projecting little spots of light onto a screen in front of the cat's eyes, trying to get the neuron to fire, and just nothing.

They tried spots everywhere.

The neuron was silent.

They were getting frustrated.

So the dot of light just wasn't the right stimulus.

It wasn't.

But then as they were pulling the glass slide out of the projector,

the sharp edge of the slide cast a faint shadow, a bar of light, across the screen for a moment, and the neuron just went crazy.

Pop, pop, pop, pop.

The audio monitor lit up.

It didn't want a dot.

It wanted a line.

It wanted an edge.

And as they discovered, it specifically wanted an edge at a very particular orientation.

And this is the distinction they found between simple cells and complex cells.

Right.

A simple cortical cell monitors a small patch of your vision.

It might respond vigorously to a bar of light, but only if it's perfectly vertical or only if it's at a 45 -degree angle.

If you tilt that bar even slightly, the cell just stops firing.

So I literally have a neuron in my brain that loves vertical lines and a different one right next to it that loves horizontal lines.

You do.

You have columns of these cells for every possible orientation, for every point in your visual field.

They are specialized edge detectors.

And complex cells, how are they different?

They take it a step further.

They also orientation selective, but they don't care exactly where the line is within their larger receptive field, as long as it's there.

And very often, they demand motion.

So a complex cell might have a preference that says, I only fire for a vertical line that is moving from left to right.

Exactly.

Or a diagonal line moving up.

So you can see how the brain is building a representation of the world from the bottom up.

It starts with dots in the retina, combines them into lines in simple cells, and then combines those into moving lines in complex cells.

The next step is shapes and then objects.

That was the old hierarchical theory.

It's very intuitive.

But the book says the modern view is a bit more abstract.

It's called the spatial frequency filter model.

This was the part that sounded more like audio engineering to me than biology.

It's literally the same math.

It's Fourier analysis.

The idea is that the visual system analyzes the entire scene, not as a collection of edges, but as a combination of overlapping waves of light and dark.

Can you break down spatial frequency for us?

Okay.

Imagine looking at a picket fence.

Lots of thin vertical slats very close together.

That's a high spatial frequency.

Lots of rapid changes from light to dark in a small amount of space.

This corresponds to fine detail and sharp edges.

And low spatial frequency would be the opposite.

Imagine a hazy cloudy sky.

Very gradual transitions from light gray to dark gray.

Big soft shadows, broad shapes.

That's low spatial frequency.

And the brain processes these different frequencies separately.

It has different neural channels that are tuned to different frequencies.

And the Mona Lisa is the absolute perfect example of how this plays out in our perception.

Ah yes, her famously elusive smile.

Why does she smile?

Or, more accurately, why does she seem to stop smiling the moment you look directly at our mouth?

It's so spooky when you experience it.

It's Leonardo da Vinci being a brilliant neuroscientist.

He was hacking your spatial frequency channels.

The smile itself, that subtle upturn of the lips, is painted almost entirely in low spatial frequencies.

Soft, blurry shadows.

Okay, so it's in the blurry channel.

Exactly.

When you look at her eyes, your high -detail vision is on her eyes.

But your peripheral vision is looking at her mouth.

And your peripheral vision is naturally blurry.

It's much better at picking up low spatial frequencies.

So your peripheral vision picks up the soft shadows of the smile.

She looks happy.

But then I get curious and I look directly at her mouth to check.

And now your fovea, your high -frequency, high -detail scanner, is pointed at the mouth.

It filters out all those blurry low -frequency shadows and looks for sharp lines and crisp details to define the expression.

But da Vinci deliberately didn't paint any sharp lines for the smile.

So when you look for it with your best vision, it vanishes.

It's a ghost smile.

It only exists when you aren't looking directly at it.

It's absolutely brilliant.

And it's definitive proof that what you see depends entirely on which filter your brain is currently applying.

Let's add another major layer of complexity to this model.

Color.

Ah, the greatest illusion of them all.

Don't tell the artists, but color isn't a real physical property of objects, right?

It's not.

Objects don't have color.

They just have surface properties that cause them to absorb some wavelengths of light and reflect others.

Red is just a label your brain invents and assigns to a specific wavelength of reflected light, somewhere around 600 to 700 nanometers.

Blue is just your brain's interpretation of short wavelength -like, around 400 nanometers.

And the first step in seeing color is the trichromatic hypothesis.

That's at the level of the cones.

Right, from Young and Helmholtz.

It's simple enough.

We have three different types of cones, and each type is most sensitive to a different part of the light spectrum.

We have short -wavelength cones, blue, medium -wavelength cones, green, and long -wavelength cones, red.

So the brain just looks at the relative amount of activity from each of the three cone types and uses that ratio to calculate the color, like mixing red, green, and blue paint.

But that doesn't explain everything.

The text brought up the opponent process hypothesis, which adds another layer.

Right, because if it's just about mixing three primary colors, why can't we imagine or perceive a color like reddish -green?

Yeah, that's a great point.

I can easily imagine blue -green, that's teal.

I can imagine red -blue, that's purple.

But reddish -green is impossible for my brain to construct.

It just becomes brown or mud.

The same with yellowish -blue.

That's because at the next level of processing in the ganglion cells and the LGN, the brain doesn't just look at the raw cone signals, it pits them against each other.

It creates specific opponent axes.

And blue versus yellow.

And then a non -color one, black versus white, for brightness.

So a single neuron can signal more red by increasing its firing rate.

Or it can signal more green by decreasing its firing rate.

But it can't possibly signal both at the same time.

They're physiologically antagonistic.

These are called spectrally opponent cells.

For instance, if you shine red light on a plus LM cell one that's excited by long wavelengths and inhibited by medium wavelengths, it fires like crazy.

But if you hit that same cell with green light, its firing is suppressed below its baseline rate.

And this explains after images perfectly.

It does.

If you stare at a bright red image for 30 seconds, you are fatiguing the red side of that red -green opponent system.

You're adapting those cells.

Then you look away at a blank white wall.

Your red sensors are tired and don't respond as strongly.

So the green sensors, which are normally balanced out by the red, suddenly dominate the signal.

Your brain interprets this imbalance as green and you see a green ghost image.

And all of this leads us inevitably to the dress.

The dress that broke the internet.

I have to ask, what did you see?

Blue and black or white and gold?

To me, it was clearly, unequivocally blue and black.

And I couldn't understand how anyone saw anything else.

I was 100 % on team white and gold.

I thought the blue and black people were crazy.

So spoiler alert for anyone who's been living under a rock.

The physical dress was in fact blue and black.

But you weren't wrong for seeing it as white and gold.

Your brain was just making a very sophisticated, and in this case, incorrect assumption about the lighting in the photo.

This is the concept of color constancy.

Yes.

Your brain is a genius at this.

It knows that the color of the light source changes how an object looks.

Sunlight is yellowish.

A shadow is bluish.

Indoor lighting is tungsten.

Your brain tries to automatically subtract the color of the light source to figure out the true color of the object's surface.

So if my brain looked at that photo and assumed the dress was in a deep shadow?

It computationally subtracts the blue of the shadow from the image, and what's left behind looks gold and white to you.

And if your brain assumed the dress was under bright, washed -out fluorescent light?

It subtracts the bright yellowness of the light, and what's left behind is clearly blue and black.

So we weren't even arguing about the dress.

We were arguing about the invisible room that our brains had independently constructed around the dress.

We each construct our own reality based on unconscious assumptions and context.

Which brings us back full circle to where we started with Patient DF and the grand architecture of the visual system,

the two streams.

We've left V1 now.

The basic image has been processed.

It has edges.

It has color.

It has motion.

Now the brain asks, what do we do with this information?

And the signal splits.

It forks into two major pathways, the ventral stream and the dorsal stream.

The ventral stream heads down, or eventually into the temporal lobe.

We call this the what pathway.

And its job is identification.

Exactly.

It compares the visual input to your vast library of memories.

Is that a cup?

Is that a face?

Is that my grandmother?

Is that a lion?

It's all about recognition and meaning.

And the dorsal stream goes up?

It heads up, or dorsally, into the parietal lobe.

This is the where pathway.

Or, perhaps more accurately, the how pathway.

Why how, instead of where?

Because its job isn't just to know where something is in space.

Its job is to guide your actions in relation to it.

Where's the object, and how do I move my hand to interact with it?

It handles spatial location, movement, and all the calculations needed for motor control.

So patient DF.

DF had a perfectly preserved dorsal stream.

Her how system was working just fine.

It could see the male slot, calculate its orientation, and guide her hand perfectly.

But her ventral stream was destroyed.

Her what system was knocked out by the carbon monoxide poisoning.

It received the visual data, but couldn't make any sense of it.

It couldn't identify the slot, the flashlight, or her husband's face.

It's just amazing that we can have these zombie -like subroutines running that we aren't even consciously aware of.

Her hand was smarter than she was.

And you see the tragic opposite of this condition as well.

A condition called optic ataxia.

Is that where the dorsal stream breaks down?

Yes.

In these patients, the ventral wood stream is perfectly fine.

They can look at a spoon on the table and say, that is a spoon.

It's for eating soup.

They know exactly what it is.

But if you ask them to reach out and pick it up, They miss.

They miss by a mile.

They grope for it clumsily.

They can't translate the visual data of its location into the correct motor coordinates for their hand.

They know what it is, but not where it is for the purpose of action.

We really are just a collection of different machines all stuffed into a trench coat and pretending to be one person.

That is a disturbingly apt description of modern neurology.

And speaking of those machines working together, the text mentions mirror neurons are found in the dorsal stream area, specifically in the premotor cortex.

Yes, this is such a fascinating area of research.

Mirror neurons are cells that fire when you perform an action like reaching for a peanut, but they also fire when you simply watch me reach for a peanut.

So my brain is internally simulating your action as if I were doing it myself.

Exactly.

It maps the visual input of someone else's action directly onto your own motor output programs.

It's thought to be the basis of empathy, and it's definitely the basis for learning by imitation.

It's how monkey see, monkey do, literally works at a neural level.

It helps dissolve the barrier between me doing something and you doing something.

I wanna pivot now to the clinical side of things before we close out.

We've talked about how this intricate system works, but what about when it breaks during development?

And more importantly, can we fix it?

Let's talk about amblyopia, lazy eye.

This is a huge topic because it affects so many kids.

And the common misunderstanding is that people think it's a muscle problem.

Oh, the eye muscle is weak, so the eye wanders.

That's not that.

No, the eye itself and the muscles are usually perfectly fine.

The brain is the problem.

If a child's eyes are misaligned, either from strabismus or a big difference in prescription, they send two different conflicting images to the brain.

The brain hates double vision.

It finds it confusing and intolerable, so it makes a ruthless decision.

It just cuts the feed from one of the channels.

It actively suppresses the input from the bad eye.

It learns to ignore it completely, functionally blinding itself to that eye's input, just to get one clear, coherent image from the good eye.

And if you let that go on for too long during childhood?

The neural connections from the lazy eye to the visual cortex wither and die from disuse.

The cortical real estate in V1 that was supposed to be for that eye gets taken over by the good eye.

You permanently lose binocular vision and depth perception.

Which is why you see kids wearing eye patches.

You patch the good eye, you put a pirate patch on it to force the brain to stop ignoring the bad eye and start listening to its signal again.

But the conventional wisdom was always that this only works on kids.

The text offered some new hope for adults.

It used to be absolute dogma that once you passed the critical period around puberty, that was it.

The wiring was set, you were stuck with amblyopia for life.

But new research is showing that there is more adult plasticity than we ever thought.

Let me guess, video games again?

Sometimes, yes.

But more specifically, using virtual reality or specific exercises that force the two eyes to work together.

If you can present an image where you subtly suppress the input from the good eye just enough and boost the signal to the bad eye, you can encourage the adult brain to start rewiring itself.

It takes a lot longer and a lot more work.

But the plasticity is still there.

And then there's the really sci -fi stuff.

The idea of restoring sight to the completely blind.

Can you tell us about that mouse study?

This is the kind of research that gives you goosebumps.

So imagine you have a disease like retinitis pigmentosa.

It's a genetic condition where your photoreceptors, the rods and cones, gradually die off.

But the wiring behind them, the bipolar cells, the ganglion cells, the optic nerve is all still perfectly intact.

So the camera sensor is broken, but the cable running to the computer is still good.

A perfect analogy.

So what these researchers did is they took precursor cells, basically stem cells that are destined to become rod cells, and they injected them into the retinas of adult mice that had been rendered blind by a similar condition.

And they didn't just die or get rejected.

Now, this is the amazing part.

They survived.

They integrated into the existing retinal circuitry.

They grew and formed new correct synapses with the bipolar cells that were waiting there.

And the ultimate question,

did the mice see again?

They did.

They regained pupillary reflexes to light.

And in behavioral tests, they could track moving bars on a screen.

They regained functional, measurable visual behavior.

That is absolutely incredible.

It suggests that the adult retina and the brain are much more accepting of new connections than we ever thought.

It opens the door to the possibility of one day being able to reseed a damaged human eye using stem cells.

So we have traveled all the way from a single photon and bending through the cornea to the backward wiring of the retina, the chemical silence of the rods, the great sorting at the optic chiasm, the edge detection in V1, and the final split into what and where.

A journey that happens in your brain dozens of times per second in a few hundred milliseconds.

What is the one thing, the final provocative thought, that you want the listener to walk away with after all of this?

For me, it's the profound realization that vision is a user interface.

It's not a window.

Expand on that.

What do you mean, user interface?

We don't see reality as it truly is.

We see a reconstruction, a model that's useful for our survival.

We have a giant blind spot in each eye that our brain just paints over so we don't notice it.

We perceive colors that are adjusted based on what we think the lighting in the room is.

We see sharp edges that are mathematically enhanced by our own retinal cells.

We see smiles that can disappear if we look too close.

So we are essentially living inside a simulation that is being generated in real time by our own cortex.

That's a great way to put it.

It's a dashboard.

It's a set of icons on a desktop.

It's designed for efficiency and survival, not for objective truth.

And if our most trusted sense, our vision, is actually this highly edited, filtered, and sometimes completely fabricated model of the world,

it forces you to ask a really deep question.

What else is just a model?

What other aspects of our reality are just efficient shortcuts?

Is my memory of yesterday a true recording or is it a reconstructed story?

Is my sense of time an accurate measurement or is it just a useful interface to keep my organism running long enough to survive and reproduce?

Well, on that existential crisis, I think we'll wrap up.

I'm going to go stare at a wall and wonder if it's really white or if my brain just wants it to be.

Keep your eyes open.

Or maybe closed.

I'm not sure which is more honest anymore.

This has been The Deep Dive.

Thank you for listening from the Last Minute Lecture Team.