Learning to See Again: A New Era of Sight Restoration, Pt. I

Just another WordPress site

Learning to See Again: A New Era of Sight Restoration, Pt. I

[MUSIC PLAYING] CHERYL KAISER: I have the distinct honor of welcoming you to our Allen Edwards Public Lecture Series that focuses on improving behavior through brain science Tonight’s lectures focus on neuroscience of sight restoration with a talk by Professor Geoff Boynton and a conversation in talk with sight recovery patient, Michael May Before introducing this evening speakers, let me just make a few comments about this lecture series and how it came about This annual series is the result of the generous support of Professor Allen L. Edwards, who made a substantial gift, an endowment that ensures that this series can take place free of charge to all of us So Professor Edwards was part of our department for half a century until his death in 1994 He was an outstanding teacher, researcher, and writer whose scholarship address personality measurement, statistics, and research design And his statistics books were gold standards for psychological research So the Edwards family contribution to the psychology department and his example of what can be accomplished with the support from members of our community– as outlined in the lecture show that I hope you enjoy this event this evening Our improving behavior through brain science series is one of our department’s new visions that help raise our impact on society to a whole new level Here, I’m going to ask that you join our efforts by becoming a friend of the psychology department by visiting our web site, or you could speak with me or our department chair, Sheri Mizumori, after tonight’s lecture And I thank many of you in the audience who have already made contributions to our department Without further delay, I want to introduce our first speaker I’m delighted to introduce Dr. Geoff Boynton as our speaker this evening Dr. Boynton earned his PhD in psychology from the University of California at Santa Barbara He’s currently a professor of psychology in our department, and he joined us in 2007 after a decade at the Salk Institute in San Diego His primary research interests are in neuroscience, and more specifically, in sensory and perceptual processes His research explores how attention shapes the way the brain processes visual information in our environment He’s pioneered the use of functional magnetic resonance imaging technology as a tool for studying vision perception, and his work has been foundational in shaping the development of vision science His work has relevance in understanding applications as diverse as sight restoration, which we’ll learn about today And one of my personal favorites is how our attention operates, so we can prevent ourselves by being hit by looming objects in our visual field that might be going straight at our head He’s authored many papers on these topics He’s been awarded grants from the National Institute of Health as well as private foundations He serves on the editorial board for the Journal of Vision and Vision Research Here at the University of Washington, Geoff is a member of the Center for Brain Imaging Director’s Committee He’s also a highly regarded teacher of statistics and neuroscience in our department And it’s such an honor to introduce my colleague, Geoff Boynton, to talk with you this evening Thank you, Geoff [APPLAUSE] GEOFFREY BOYNTON: Thank you, Cheryl I’m not going to talk about objects flying at people’s heads today Instead, I’ll be talking about some of the research that I’ve been doing with my close colleagues, Ione Fine, Michael Baylor, Ariel Rokem, on work we’ve been doing trying to understand and the visual experience might be like for someone who has had surgery for a retinal prosthetic And after my talk, which is going to be about 45 minutes, there will be some time for question and answer, and then we can do a transition over to our second speaker You can think of me as being the warm up band for Mike May, who we’re lucky to have here tonight, who, in addition to having a very interesting life, is a sight recovery patient And so he’s going to tell you a lot about his experience of having his vision restored as an adult So the work we’re going to be talking about is work related to something called retinal prosthetics And if you want to learn more about that and you want to do a web search, you might want to search under the term “bionic eye.” Now, that’s the colloquial term for this sort of technology And when I think– I’m a guy of a particular generation When I think of bionic eye, I immediately think of Steve Austin, The Six Million Dollar Man, the TV show from the 70s Do you remember this? So back then, this was, I think, 1974 when the first season came out The main character had, after a horrible test plane accident, had his legs replaced, his right arm replaced, and his left eye replaced with a bionic eye

And the whole idea was to make him– what was it? Better than he was, faster than he was, and stronger than he was Well, what we’re doing, now that it’s more than 40 years later, we are actually– we, I would say companies– are actually developing such devices to try to restore vision in the blind And we’re really trying to assess how well these devices work to help with help the engineers develop these devices in the future Full disclosure here, I am a PhD, not an MD So I think of this as the less useful kind of doctor So my background is in math and computer science before moving and seeing the light, so to speak, moving over to psychology and neuroscience So if there’s any kind of medical emergency tonight, I’m not much use But if anyone has a serious urgent math or statistics problem, I can be the first responder So why all this interest in retinal prosthetics? And why is it so much in the news? Well, blindness is a big problem And not just being blind is a problem, but the prevalence of blindness is a problem too The most common causes of blindness in adults are disease related And they’re related The two main causes are macular degeneration and retinitis pigmentosa Both of these are diseases of the retina, which I’ll talk about this in a moment in a crash course on the physiology of the retina But these are diseases where the photo receptors in the back of the eye, the light gathering cells, degenerate There are a variety of causes Generally, the causes are genetic, and there are many, many genetic gene mutations that can lead to this sort of vision loss And they’re age related And as our population ages, the prevalence of age-related diseases is increasing So this is a real problem The two main classes of retinal degeneration– macular degeneration typically leads to the loss of photoreceptors in the cones, related to central vision, your foveal vision And that tends to spread from the fovea to the periphery Retinitis pigmentosa, sometimes called just RP, starts with the generation of photoreceptors called the rods, which we use for night vision So the first symptoms tend to be a loss of vision at low light levels, and that tends to lead to photoreceptor death in the cones, the day vision photoreceptors And that tends to lead to loss of vision from the periphery to the fovea over time I’m going to talk a little bit more about what happens in the retina for these diseases But first, let’s do a crash course on retinal physiology So this is a cartoon version of the eye and the retina And as many of you know, the eye is much like a camera, at least the optics of the eye are, where the cornea and the lens help to form an image of the world on the back of the eye here And the image is upside down and backwards This seemed to cause a lot of trouble for the early philosophers, but if you think about it, the brain has always experienced the world this way It doesn’t know anything different So there’s really not a problem for the brain It’s not like we have to do this in order to see the world properly In the back of the eye, there’s a thin tissue, neural tissue, called the retina Now, tissue is the right name for it If you ever see surgery for a retinal detachment, and you get to see what the retina actually looks like when it’s removed from the back of the eye, it looks like wet tissue paper It’s thin It’s easily tearable It’s very delicate If you zoom in on the retina, you can see that the retina actually consists of various cell layers Light comes in from this end, and eventually, if a photon is lucky, it will hit one of these cells down here called photoreceptors These photo receptors are amazing devices They act like light waveguides that can collect the light, and if a photoreceptor does come in here, if it’s lucky, it can do something called isomerize, and lead a really interesting cascade of events that effectively amplifies the energy in the photon, leading to an electrical change that can cause these intermediate neurons to start firing on its way to sending signals to the brain These intermediate layers, amicrine cells, bipolar cells, horizontal cells, do some really interesting image processing For the most part, they tend to add and subtract signals coming from the photoreceptors up from the bottom And so, for example, this cell might respond very well to a bright light shining on that photoreceptor, but subtract light coming from these photoreceptors So the early stages of visual processing happens in the retina The signal that’s leaving the eye is not like a pixel version of the world out there And the image information does leave the eye through this last layer called the ganglion cell layer These ganglion cells are very specialized neurons that send spiking signals out the eye through basically a hole in the back of the eye, a long cable all the way back into

the middle of the brain, in a part of the brain in the central brain called the lateral ventrical nucleus It’s interesting You might have noticed here that light comes down through here It passes through all this junk that are not light receptive cells And eventually, you hit the photoreceptors down here This seems backwards, right? We are now, when you’re staring at me, you’re looking through a bunch of nontransparent cells in order to be able to see me We can’t see that We don’t experience that, because anything that doesn’t change over time in the visual system tends to fade away perceptually And so why is that the case? Well, you read textbooks on this There’s a lot of post hoc explanations Well, these photoreceptors down here are metabolically very active They shed their so-called disks over time, and the vascular system’s down here Maybe there’s something about keeping things clean and housekeeping But actually, I guess last week there was some discussion about the octopus Many of you were here last week The octopus is back It turns out that the octopus and other mollusks, their retinas flip around the other way, in a sensible way The photoreceptors are right here Light comes down, hits the photoreceptors Then the signals pass through a layer of cells and out the eye What’s the deal? Why are we this way? Perhaps we’re not as evolved as the octopus I pointed this out to my father-in-law, who happens to be a philosophy professor at NYU And he thought for a moment, and said, maybe God is an octopus [LAUGHTER] So here’s a painstakingly long drawn diagram from a well-known paper showing the so-called axon pathways of these ganglion cells and how they leave from a particular location in the retina, lead out to a particular part in your back of your retina, and then out the back of your eye to your brain Now, this is an amazingly distinct pattern It’s like a fingerprint And these ganglion cell pathways, well, they don’t cross streams, because that would be bad And if you think about it, you have to somehow get signals from every point on your retina out the back of the eye What’s interesting about this is, underneath the optic disk, because it’s basically a hole in the back of your eye, there are no photoreceptors So many of you might know that you actually have a blind spot for each one of your two eyes And if you close one eye and hold your thumb out about arm’s length, about five degrees off to the side, away from your nose, your thumb might disappear if you do it properly Now, fortunately, we have two eyes And fortunately, most of us have two eyes And fortunately, the blind spot for the two eyes don’t line up in the same place So the other eye can take over where one eye might have a blind spot It’s fun to do actually I do this in faculty meeting sometimes where I can stare straight ahead and make my faculty members disappear I shouldn’t say that because many of them are here tonight If I’m winking at you in faculty meetings, I’m not trying to communicate It’s because I’m trying to make you go away [LAUGHTER] In the fovea here, this is foveal pit This is where the center of vision happens It turns out that the ganglion cells are pulled away from the very center So actually in the very center, your finger nail’s width at arm’s length, you’re not looking through the ganglion cells Light can actually hit straight down to the photoreceptors So in the intact, healthy eye, this is another diagram that’s very similar to the one I showed before showing the ganglion cell, there are different cell layers There’s photoreceptors, intermediate cells, and the ganglion cells But in the diseased eye, for example, in late stage retinitis pigmentosa, you can see what’s happening here in this diagram is the photoreceptors have degenerated And so they no longer can capture light and send signals to the brain, hence blindness But interestingly, and most important for what we’re talking about today, is that many of the remaining cells are still alive, and potentially, active And so many of these sight restoration technologies are attempts to stimulate these remaining non-photo and light sensitive cells And if you think about it, if you make a ganglion cell send a signal to the brain, the ganglion cell doesn’t know that light made it go versus an electrical signal So the brain might see a flash of light or something like that, and that’s the basis of sight restoration technologies There are three main classes of sight restoration technology I’m going to focus on retinal electrical prosthetics But I should talk about the other two first These are newer, emerging technologies We know less about their efficacy, because they’re all in clinical trial stage, rather

than actual patients coming home with these devices One is a very exciting idea dealing with optogenetics Optogenetics is the process of using a viral vector to effectively change the channel structures of cells that are not photoreceptive, like ganglion cells, and make them light sensitive Which is pretty amazing So you can actually take a cell, like the ganglion cells, the cables that leave the eye, and make them sensitive to light You’d think this would be the magic bullet This would cure blindness by simply turning parts of the eye that didn’t used to respond to light, make them respond to light That’d be great, but there are a variety of limitations for this One of the biggest ones is that this technology, as it is right now, these light sensitive cells, when they’re made light sensitive, are much less sensitive than a photoreceptor And so you need two or three more orders of magnitude of light at least to make these things fire So daylight light levels are not enough to make these work So in order to use this technology what these companies are going to have to do is put a light source pointing into the eye connected to a camera which is producing a very intense light to make these things fire So it’s going to require some sort of prosthetic device in order to be able to make the the brain see a signal Another technology which is really exciting is gene therapy You might have heard about some of this Some of these are in clinical trials trials for Leber’s disease And this is the idea of using viral vectors, usually like the rabies virus, to go and basically inject it into the retina, the diseased retina And either prevent further degeneration of disease, or maybe even allow photoreceptors to grow back again by literally altering the gene structure of cells in the retina, thereby effectively eliminating the disease This, if it works, is a true cure for blindness, for these particular kinds of blindness The drawbacks are, well, as I mentioned, there’s many, many different genetic causes of sight loss And so that means you have to develop a different gene therapy for each particular type of sight loss And also it’s not obvious if it’s going to be permanent, but this is a very promising technology But it’s still in the early stages The most developed stage of sight recovery are these so-called bionic eyes, which involve direct placing an electrode, an electrical stimulating device, on the retina Which when you turn them on and send currents through them, will induce firing activity in the remaining cells in the retina And therefore, hopefully, producing something that might resemble sight And that’s what we’re talking about for the rest of today There are a large and growing number of groups that are exploring this technology throughout the world And they range from full-fledged companies, like a company called Second Sight down in Sylmar, California– that’s the company we’re working most closely with– to just independent academic groups and universities, like the group in Melbourne, which are really just a bunch of professors and engineers that are trying to develop this system Some are for profit Some are just curious scientists, and it’s a very large and growing group For the most part, these electrical prosthetic devices are based strongly and heavily on cochlear implant technology So cochlear implants are devices where you can send an electrode down inside the cochlea of the ear for people that have had hearing loss or were born deaf And with electrical stimulation, you can stimulate the remaining cells in the ear, thereby sending signals to the brain that resemble sound-like things Now, cochlear implant technology has been quite successful, especially when implanted in people at a young age And so much of this technology has been developed to use in the retina In fact, this Company Second Sight in Sylmar, its sister company is a company called Advanced Bionics, which is a successful cochlear implant company In fact, the first test implant that Second Sight used had 16 electrodes It was a four by four grid of electrodes And it was 16, because their successful cochlear implant at that time had 16 electrodes in a line right there How does this work? Well, there’s actually a variety of places in the retina that you can place one of these devices Second Sight does their retinal implant place right here on top of the ganglion cells It’s called an epiretinal location

Another potential location is the subretinal location, which has been layered between the photoreceptors and the back of the eye And of course, there are no photoreceptors in these disease patients, so it’s somewhere between the ganglion cells and back here The advantage to epiretinal here is that these electrodes are very close to the ganglion cells which we’re trying to stimulate The disadvantage is that there’s nothing really to hold this electrode and keep it from moving around And Second Sight actually has a patented tack system, where it’s stapling this thing to the back of the eye But it can still maybe lift off a little bit and move around If you stick it in between layers of cells and other stuff, it can be wedged in there and not move around as much The disadvantage of this is it’s further away from the ganglion cells, and also there’s an issue of heat dissipation So changes in current can lead to heat, and because it’s stuck inside a bunch of stuff, it’s harder to have these things not heat up and burn out the electrodes, and even worse, burn out the retina The last one is kind of interesting This is various groups like the one we’re talking with in Melbourne where you can actually stick this electrode outside the eye, basically in the back of the eye, called the suprachoroidal location And this makes surgery really easy You don’t need to make slits in the eyeball and stick this thing inside the eye You just slide it around the back of the eye and tack it on back here It makes surgery very easy, relatively easy And also makes it easy to replace As technology improves, you can just pull it out and put a new one in relatively easily presumably It has the disadvantage, of course, that we’re farthest away from the ganglion cells as any of the other methods And the further you go, the bigger the spread of current you’re going to have And so the ability to stimulate a very localized part of the retina becomes more and more difficult But what we’ve seen from the results from this group are remarkably good I’m going to show a cheesy movie pulled straight from Second Sight’s web, site because it is probably the best description of how this technology works And so let me take a drink of water while this thing runs, and let this thing go And watch the end of this movie because there’s something important about the way they think people are going to use this [VIDEO PLAYBACK] – In Second Sight’s Argus II retinal prosthesis system, a miniature video camera in the eye glasses captures the scene [MUSIC PLAYING] The video is processed by a small portable unit and transformed into instructions, which are sent back to the glasses These instructions are then transmitted wirelessly to the implant on the eye The implant consists of a receiver and an array of electrodes Instructions are received, and corresponding signals sent to the array, which emit small pulses of electricity These pulses stimulate the retinas remaining cells, and are transmitted down the optic nerve, conveying visual information to the brain which perceives patterns of light Patients learn to interpret these visual patterns [END PLAYBACK] GEOFFREY BOYNTON: So Second Sight was the first company to develop these prosthetics and get FDA approval to do this So they are on the order of dozens of patients now that have gone home with these devices So there’s lots of data to be gathered from them I don’t know if you noticed at the end of this video was the implication that if you are staring– so here’s another image taken from their web site If a patient is staring at a doorway– so here’s the region of the visual field that they have vision loss– if there’s a doorway that’s behind that, if you just simply stimulate the electrodes in such a way that resembles the shape of the doorway, the patient will see the shape of a door They’re a little careful in the wording there They say patients learn to interpret the input, but they do have this picture right here And in this picture is the implication that if I turn on an electrode right here, it’s going to lead to the percept of a single spot of light at the corresponding location where the photoreceptors used to be, underneath an electrode We call this the scoreboard model of visual percepts It’s called the scoreboard model because if you’ve ever been to a stadium, and walked up and looked up to one of those big stadiums scoreboards really close, you’ll notice the scoreboard is actually made up of a bunch of individual light sources that can turn on and off at different intensities, different colors And it looks like a continuous image, because if you stand far enough back, the visual system doesn’t have the special resolution to see the individual lights And so with a high enough resolution, you can fool the visual system into seeing something that looks like a continuous image And of course, in fact, this screen right here is the same principle If I look really, really close, I can see individual pixels that you can’t see because

you’re sitting too far away So presumably, if you have enough electrodes, and each produces its own spot of light when I turn it on, with brute force advances in technology, all I need to do is increase the number of electrodes And I should eventually be able to restore sight So the first thing you want to know is, well, is the scoreboard model valid? Well one, way to test it, and this is something that my colleague Ione and other people at Second Sight have been doing, is to literally ask patients What do you see when you turn an electrode on? And the way they can answer the question in this particular example is to have a patient draw with their finger on a touch screen right in front of them– it’s pretty crude, but effective, a touch screen– what they perceive And so for example, if you were to have a four by four array like this, like the early versions of Second Sight’s prosthetic, and turn these two electrodes on, ideally, you should see two spots of light in those two locations But what do they actually see? Well, this is an example of one patient’s drawings One of these electrodes led to a fairly nice punctate spot of light But another one led to this really long weird banana-shaped streak These are different gray scales because what we’ve done is drawn on top of each other various replications of the same drawing over a bunch of trials So it shows that her ability to draw is very easily replicable So she sees this streak every time you turn this electrode on Well, what’s going on here? This is clearly a problem with the scoreboard model Another problem is that, over time, if you turn electrode on, the percept fades Some of you might know this about the visual system If you stabilize an image on the retina, It Tends to disappear over time It’s something called Troxler fading It’s a big problem in the retinal prosthetic industry Here’s an image showing, drawn on top of– this is now an A 60 This is the 60 electrode array Drawn on top of that are streaks drawn by a patient for one half of the electrodes and the other half of the electrodes This is divided in half, so these aren’t all lying on top of each other And you can see that there are very few spots and many streaks Here’s another patient showing also not spots but thicker streaks, different percepts Not quite sure why they’re quite different, but it might have something to do with how far away this electrode is sitting from the ganglion cell there, because it does vary from patient to patient You can see there’s some systematicity in the way these streaks are pointing They’re not all just random directions And we think the cause of this has something to do with the way the ganglion cell pathways, axon pathways, are leaving the eye So for example, suppose you’re stimulating with electrode right there, and a patient sees a streak like that Well, it turns out that underneath this streak, if you underlay this axonal pathway diagram, you might see something like this, where the streak is following these pathways And if you think about it, if you stimulate right here, there might be a ganglion cell who started its pathway right there that was connected to photoreceptors originally underneath there And so sure enough, you’re going to see a spot of light in the right place But also suppose you have a ganglion cell that starts right here, and so it represents this part of the visual field that sends its signal along this pathway underneath the electrode and out the eye If that’s the case, stimulating here was also going to stimulate that axon wire, and thereby leading to a perceptive light over here, and so on, all the way down the pathway So we believe in our modeling ideas is this is what’s causing these particular streaks If you stick a electrode right here, you might not see a long streak, because the pathways basically end right there And so we’ve been able to match up these perceived streaks with this particular kind of diagram from patient to patient with reasonable success Another issue, we’ve got lots and lots of data from these patients, measurements such as– how bright does the percept look regardless of its shape? As I change the current amplitude into the electrode, and what you’ll see is a typical non-linear relationship, where as I turn the amplitude up, I don’t get this linear increase in perceived brightness And this is a simple experiment where a subject is simply giving us a number from 1 to 20, or 0 to 20, how bright this thing looks to them There’s a whole bunch of other so-called psychophysical measures using brightening– brightness matching techniques, threshold detection techniques, which is like how much current do I need to be able to see a flash of light– huge piles of data coming from Second Sight and other companies that we’re trying to compile to try to understand what these patients are seeing And ultimately, what we’re trying to develop in our developing is something that we call a virtual patient So given some pattern of electrostimulation on the retina, a patient may see some kind

of percept Our goal is to take the same input patterns of stimulation along, say, 60 electrodes over time, pass it through a model that we’ve been developing And the output of the model is going to be basically a movie that simulates what this patient will see And so for this particular example, our model output for stimulating these two electrodes gives us that right there, which is a reasonable match of the percept of the subject to the patient we’re seeing in this example Now, this model is based loosely on retinal physiology There’s various stages of spatial and temporal processing But the parameters of the model have been constrained by the data sets that we get from behavioral data So it’s kind of a nice hybrid between a physiological model that’s constrained by data, behavioral data And so how is this coming along? Well, one of the main reasons why we’re doing this, some obvious reasons– one of them is it provides a more realistic estimate of how visual prosthetics are likely to perform And this is useful for doctors that might be implanting these techniques to be able to tell the patient how well these things might work It’s great for engineers, so they get an idea of how well these things are going to work before they actually build it And it can provide insights into what sort of visual tests are important for evaluating the devices You’ll see in a second that it’s going to be pretty obvious that these devices are not going to be very good for things like facial recognition And so there’s no point in trying to do a facial recognition test on these patients And finally, the interesting part is we’re going to try to invert this model so we can find the best pattern of stimulation to lead to a desired percept So we call this a forward model where, given some input, we have some output that is going to be some movie over time But what we really want is suppose we want to have the letter A over here at the end What pattern of stimulation do I need to produce that output? That’s actually the model working backwards And because this model is so complicated and non-linear and slow, it’s going to require some pretty tricky techniques to be able to do this And we’re actually collaborating with the groups at e science in the physics building to use some of these deconvolution learning algorithms to basically train up this network, given an input and desired output, to figure out how to make this thing work backwards in just a brute force manner So how well is this working right now? Well, here we go Can anybody guess? This is our best guess of what a patient wearing the Argus II, which is the 60 electrode array, what they might see given a very large letter presented to them Can anybody guess what that is? SPEAKER 1: A GEOFFREY BOYNTON: Very good, it’s a black A on a white background How about this one? SPEAKER 2: E GEOFFREY BOYNTON: It’s actually also an A, but it’s actually a little easier to see probably to many of you, and that’s a white A on a black background And this is something we’ve actually learned from our simulations, that it seems to be easier, at least for us sighted people, looking at the output of the model, to recognize a letter when it’s white on black versus black on white And indeed, it turns out that Second Sight, when they send these patients home with these devices, they have a series of knobs on their computer that translates the camera image into the stimulation pattern And one of the knobs allows them to do contrast reversing, which basically takes a white part of the visual field and turns into a low level stimulation, and a dark part of the visual field into bright, high levels of visual stimulation And many of these patients are said to prefer this contrast reversing mode rather than the standard mode, where dark parts of the field are low amounts of stimulation It’s kind of interesting, and it’s almost a philosophical point here, where– are we trying to restore sight, as in how we see? Or are we trying to restore functionality? So you and I who are sighted would never want to reverse the contrasts of the world to function better But for some reason because of all the weirdness of the distortions of the image, it actually is easier for patients to be able to use this device when flipped around contrast reversed So this is restoring function, but not necessarily restoring traditional sight This is a really important distinction we try to make when we’re talking about this work, especially with the engineers We can also make movies of this So here’s a stock movie of the Boston subway of people walking around Let’s see if this works There’s no sound here which is good You can see there’s a person walking from left to right, and there’s some motion right to left, and the car starts moving here This is what we have our best guess of what a patient might see while looking at this movie I’ll show this a couple of times And what I see is the basic motion of a person moving from left to right

I don’t see a lot of the aftermath of people moving right to left in the distance And you can be a judge of whether this would be a useful device for somebody that’s blind I would say that, overall, it’s better than nothing You can see whether there’s global motion going on And in fact, these patients say that these devices are useful for major global motion and not so much for fine-grained detection of movement I’ve got a few more examples Here’s a fun one This is in England, northern England, where we spent a year This is my nephew schooling his older brother in an epic soccer goal right here– boom [LAUGHTER] His older brother’s not too pleased about that I probably shouldn’t be showing this without his permission, now that I think about it, because he didn’t want this movie in a publication that we made Oh, well, just pretend you never saw this What happens here stays here But here, this is a little disappointing This is less of an epic soccer goal as seen through the prosthetic device Not a whole lot going on here, and part of what you notice is when I filmed this in the backyard in England, I actually track the camera partly with the cousin So there actually isn’t much camera motion on the image itself So there’s not much to see except until the end This movie is in black and white Some people might ask about color Restoring color vision with these prosthetic devices is, at best, a very long way off If you think about trying to excite these ganglion cells to send a color signal to the brain, we hardly understand how ganglion cells send color signals to the brain So we’re way, way off from that So this movie was originally filmed on my iPhone in color But it was northern England in the winter, so it didn’t really matter anyway Here’s another one, northern England, of my then 7-year-old practicing for the X-Games on his scooter This is a good one because it really has nice, clear, right to left motion, and it’s pretty easy to see that in the movie You wouldn’t know that this is a kid on a scooter, but it’s definitely something moving right to left, and you might want to avoid it if it’s making a lot of noise So finishing up here, where are we with electrical prosthetic technology? Well, Ione has a really nice analogy where we like to think of where we are with electrical prosthetics is sort of where human flight was maybe at the beginning of the last century So back then, people were building kind of ridiculous looking things, and strapping it to their back, and jumping off cliffs, and trying to see if these things worked And there’s a lot of trial and error And some things worked, and if they did, they adapted them and moved on And eventually, we are where we are now But it was a lot of trial and error Where are we now with flight? Well, here we are in Seattle The Boeing 787 was developed almost entirely with computer aided design in simulations for mechanical engineering, aerodynamics, and heat dissipation And by the time the first 787 was built, they pretty much had a good idea of its flight characteristics And there weren’t a whole lot of surprises, except I think they were worried about the wings falling off or something like that But aside from that, the idea is that if you know enough about how the system works, you should be able to design something and build it and have an idea of how well it’s going to work before you actually try it out And what we’re doing with our group with these simulations is try to get retinal prosthetics to help them get from here to there so that an engineer will know what size electrodes, the array of electrodes, what pattern of stimulation– what’s going to work best before you even build this thing and stick it in somebody’s eye And as you know, as I pointed out, these visual percepts are not ideal They’re quite distorted And if you read articles on this sort of thing from these groups, typically in the discussion section of these articles there’s some talk about– well, in the visual system in the human brain is very plastic We can learn stuff And so perhaps patients will learn how to use this complicated and distorted visual input And that may be true We learn a lot You learned something about the octopus eye today, and you’ll probably know that for a long time Your brain has changed because you listen to somebody tell you something Well, the visual system can learn too And as you know if you meet somebody new, you’ve seen a new face, you can recognize that face later on Something in your brain has changed when you saw that face So neuroplasticity and learning can be very helpful On the other hand, there are limitations to learning

And this is something we’re going to talk about and learn about in the second half of tonight, where Mike May is going to come up after our question and answer And he’s going to talk about his experience of having his sight restored as an adult, and the successes and maybe lack of success that he’s had on using this brand new sensory input So thank you And I think there’s a microphone back there ANGIE: OK, my name’s Angie And my question is in two parts First one is with the surgery for the microchip to be embedded the optic of the eye in the retina Is that invasive laser, or is it actually going to involve a scalpel? And then second question is with the neurological system so close to the eye orbit Would it interfere at all with the brain transmitters that we have, with, say, endorphins and our natural sensors? I mean kind of like a short circuit type of thing, because of it’s an actual metal inside the eye GEOFFREY BOYNTON: Yeah, I can see it you’re saying Well, the first question is I think you’re talking about the super suprachoroidal technique, in which you’re embedding this prosthetic behind the eye That doesn’t require any opening up of the eye or heavy surgery like that As far as I know, it’s a relatively easy surgery You just go around the outside, underneath, around the socket of the eye The second question is whether these electrical signals that you’re sending into the retina can interfere with other kind of neural activity I think is what you’re asking And the efficacy of these electrical currents that you’re setting up that drive these cells really drops off quickly It’s an inverse square law with distance from the neurons So I seriously doubt that an electrical signal in your retina is going to affect any brain activity right behind that, say, which is important, because your frontal lobe is right behind your eyeball, and your supraorbital cortex and all that So I don’t think that is a problem But it’s an interesting question Thanks Go ahead SPEAKER 3: Is any of this applicable to a detached retina? GEOFFREY BOYNTON: I don’t believe at the moment So a detached retina is when basically all the layers come off from one part and it requires some surgery to correct the retina and put it back down on top of the sclera As far as I know, there would be no need to, because the retina still functions It’s just in the wrong place So I don’t think it makes any sense to stick an electrode on top of a functioning retina, even if it is somehow moved or lifted off the back of the eye Interesting question Yes? SPEAKER 4: What are the major roadblocks that you see between you and where you want to be? Is it the electrodes, the electronics, the processing, the data? GEOFFREY BOYNTON: Great question, yeah, what are the major roadblocks? So I think it’s clearly the case that exciting a given part of the retina with a small electrode is going to lead to something related to the scoreboard model So this streakiness that we’re seeing is a big problem This fading over time is a big problem, because you can’t just keep turning up the current level to compensate for the fading, because eventually you run into current level problems where you overstimulate and can cause heating in the retina But I think a lot of these problems can be compensated with what we’re doing, coming up with more clever techniques of stimulating patterns to try to produce a desired output rather than just focusing on simply a relationship between, say, the brighter part of the camera image is going to relate to a higher current for that particular electrode But I think the major limitations are the spatial blurring and the temporal fading that we’re talking about [MUSIC PLAYING]