Mary Lou Jepsen
730,585 views • 16:50

People don't realize that red light and benign near-infrared light go right through your hand, just like this. This fact could enable better, faster and cheaper health care. Our translucence is key here. I'm going to show you how we use this key and a couple of other keys to see deep inside our bodies and brains.

OK, so first up ... You see this laser pointer and the spot it makes on my hand? The light goes right through my hand — if we could bring the lights down, please — as I've already shown. But you can no longer see that laser spot. You see my hand glow. That's because the light spreads out, it scatters. I need you to understand what scattering is, so I can show you how we get rid of it and see deep inside our bodies and brains. So, I've got a piece of chicken back here.

(Laughter)

It's raw. Putting on some gloves. It's got the same optical properties as human flesh. So, here's the chicken ... putting it on the light. Can you see, the light goes right through? I also implanted a tumor in that chicken. Can you see it?

Audience: Yes.

Mary Lou Jepsen: So this means, using red light and infrared light, we can see tumors in human flesh. But there's a catch. When I throw another piece of chicken on it, the light still goes through, but you can no longer see the tumor. That's because the light scatters. So we have to do something about the scatter so we can see the tumor. We have to de-scatter the light. So ...

A technology I spent the early part of my career on enables de-scattering. It's called holography. And it won the Nobel Prize in physics in the 70s, because of the fantastic things it enables you to do with light. This is a hologram. It captures all of the light, all of the rays, all of the photons at all of the positions and all of the angles, simultaneously. It's amazing. To see what we can do with holography ... You see these marbles? Look at these marbles bouncing off of the barriers, as an analogy to light being scattered by our bodies. As the marbles get to the bottom of the scattering maze, they're chaotic, they're scattering and bouncing everywhere. If we record a hologram at the bottom inside of the screen, we can record the position and angle of each marble exiting the maze. And then we can bring in marbles from below and have the hologram direct each marble to exactly the right position and angle, such as they emerge in a line at the top of the scatter matrix.

We're going to do that with this. This is optically similar to human brain. I'm going to switch to green light now, because green light is brighter to your eyes than red or infrared, and I really need you to see this. So we're going to put a hologram in front of this brain and make a stream of light come out of it. Seems impossible but it isn't. This is the setup you're going to see. Green light. Hologram here, green light going in, that's our brain. And a stream of light comes out of it. We just made a brain lase of densely scattering tissue. Seems impossible, no one's done this before, you're the first public audience to ever see this.

(Applause)

What this means is that we can focus deep into tissue. Our translucency is the first key. Holography enabling de-scattering is the second key to enable us to see deep inside of our bodies and brains.

You're probably thinking, "Sounds good, but what about skull and bones? How are you going to see through the brain without seeing through bone?" Well, this is real human skull. We ordered it at skullsunlimited.com.

(Laughter)

No kidding. But we treat this skull with great respect at our lab and here at TED. And as you can see, the red light goes right through it. Goes through our bones. So we can go through skull and bones and flesh with just red light. Gamma rays and X-rays do that, too, but they cause tumors. Red light is all around us.

So, using that, I'm going to come back here and show you something more useful than making a brain lase. We challenged ourselves to see how fine we could focus through brain tissue. Focusing through this brain, it was such a fine focus, we put a bare camera die in front of it. And the bare camera die ... Could you turn down the spotlight? OK, there it is. Do you see that? Each pixel is two-thousandths of a millimeter wide. Two microns. That means that spot focus — full width half max — is six to eight microns. To give you an idea of what that means: that's the diameter of the smallest neuron in the human brain. So that means we can focus through skull and brain to a neuron. No one has seen this before, we're doing this for the first time here. It's not impossible.

(Applause)

We made it work with our system, so we've made a breakthrough.

(Laughter)

Just to give an idea — like, that's not just 50 marbles. That's billions, trillion of photons, all falling in line as directed by the hologram, to ricochet through densely scattering brain, and emerge as a focus. It's pretty cool. We're excited about it.

This is an MRI machine. It's a few million dollars, it fills a room, many people have probably been in one. I've spent a lot of time in one. It has a focus of about a millimeter — kind of chunky, compared to what I just showed you. A system based on our technology could enable dramatically lower cost, higher resolution and smaller medical imaging. So that's what we've started to do. My team and I have built a rig, a lab rig to scan out tissue. And here it is in action. We wanted to see how good we could do. We've built this over the last year. And the result is, we're able to find tumors in this sample — 70 millimeters deep, the light going in here, half a millimeter resolution, and that's the tumor it found.

You're probably looking at this, like, "Sounds good, but that's kind of a big system. It's smaller than a honking-big MRI machine, monster MRI machine, but can you do something to shrink it down?" And the answer is: of course. We can replace each big element in that system with a smaller component — a little integrated circuit, a display chip the size of a child's fingernail.

A bit about my background: I've spent the last two decades inventing, prototype-developing and then shipping billions of dollars of consumer electronics — with full custom chips — on the hairy edge of optical physics. So my team and I built the big lab rig to perfect our architecture and test the corner cases and really fine-tune our chip designs, before spending the millions of dollars to fabricate each chip. Our new chip inventions slim down the system, speed it up and enable rapid scanning and de-scattering of light to see deep into our bodies. This is the third key to enable better, faster and cheaper health care. This is a mock-up of something that can replace the functionality of a multimillion-dollar MRI machine into a consumer electronics price point, that you could wear as a bandage, line a ski hat, put inside a pillow. That's what we're building.

(Applause)

Oh, thanks!

(Applause)

So you're probably thinking, "I get the light going through our bodies. I even get the holography de-scattering the light. But how do we use these new chip inventions, exactly, to do the scanning?"

Well, we have a sound approach. No, literally — we use sound. Here, these three discs represent the integrated circuits that we've designed, that massively reduce the size of our current bulky system. One of the spots, one of the chips, emits a sonic ping, and it focuses down, and then we turn red light on. And the red light that goes through that sonic spot changes color slightly, much like the pitch of the police car siren changes as it speeds past you.

So. There's this other thing about holography I haven't told you yet, that you need to know. Only two beams of exactly the same color can make a hologram. So, that's the orange light that's coming off of the sonic spot, that's changed color slightly, and we create a glowing disc of orange light underneath a neighboring chip and then record a hologram on the camera chip. Like so.

From that hologram, we can extract information just about that sonic spot, because we filter out all of the red light. Then, we can optionally focus the light back down into the brain to stimulate a neuron or part of the brain. And then we move on to shift the sonic focus to another spot. And that way, spot by spot, we scan out the brain.

Our chips decode holograms a bit like Rosalind Franklin decoded this iconic image of X-ray diffraction to reveal the structure of DNA for the first time. We're doing that electronically with our chips, recording the image and decoding the information, in a millionth of a second. We scan fast.

Our system may be extraordinary at finding blood. And that's because blood absorbs red light and infrared light. Blood is red. Here's a beaker of blood. I'm going to show you. And here's our laser, going right through it. It really is a laser, you can see it on the — there it is. In comparison to my pound of flesh, where you can see the light goes everywhere. So let's see that again, blood. This is really key: blood absorbs light, flesh scatters light. This is significant, because every tumor bigger than a cubic millimeter or two has five times the amount of blood as normal flesh.

So with our system, you can imagine detecting cancers early, when intervention is easy, or tracking the size of your tumor as it grows or shrinks. Our system also should be extraordinary at finding out where blood isn't, like a clogged artery, or the color change in blood as it carries oxygen versus not carrying oxygen, which is a way to measure neural activity. There's a saying that "sunlight" is the best disinfectant. It's literally true. Researchers are killing pneumonia in lungs by shining light deep inside of lungs. Our system could enable this noninvasively.

Let me give you three more examples of what this technology can do. One: stroke. There's two major kinds of stroke: the one caused by clogs and another caused by rupture. If you can determine the type of stroke within an hour or two, you can give medication to massively reduce the damage to the brain. Get the drug wrong, and the patient dies. Today, that means access to an MRI scanner within an hour or two of a stroke. Tomorrow, with compact, portable, inexpensive imaging, every ambulance and every clinic can decode the type of stroke and get the right therapy on time.

(Applause)

Thanks.

Two: two-thirds of humanity lacks access to medical imaging. Compact, portable, inexpensive medical imaging can save countless lives.

And three: brain-computer communication. I've shown here onstage our system focusing through skull and brain to the diameter of the smallest neuron. Using light and sound, you can activate or inhibit neurons, and simultaneously, we can match spec by spec the resolution of an fMRI scanner, which measures oxygen use in the brain. We do that by looking at the color change in the blood, rather than using a two-ton magnet. So you can imagine that with fMRI scanners today, we can decode the imagined words, images and dreams of those being scanned. We're working on a system that puts all three of these capabilities into the same system — neural read and write with light and sound, while simultaneously mapping oxygen use in the brain — all together in a noninvasive portable that can enable brain-computer communication, no implants, no surgery, no optional brain surgery required. This can do enormous good for the two billion people that suffer globally with brain disease.

(Applause)

People ask me how deep we can go. And the answer is: the whole body's in reach. But here's another way to look at it.

(Laughter)

My whole head just lit up, you want to see it again?

Audience: Yes!

(Laughter)

MLJ: This looks scary, but it's not. What's truly scary is not knowing about our bodies, our brains and our diseases so we can effectively treat them. This technology can help.

Thank you.

(Applause)

Thank you.

(Applause)