The Possibilities of Vision Restoration

On Tech and Vision Podcast with Dr. Cal Roberts

This podcast is about big ideas on how technology is making life better for people with vision loss.

For hundreds of years, health professionals have dreamed of restoring vision for people who are blind or visually impaired. However, doing so, either through transplanting a functioning eye or using technological aids, is an incredibly complex challenge. In fact, many considered it impossible. But thanks to cutting-edge research and programs, the ability to restore vision is getting closer than ever.

As a first for this podcast, this episode features an interview with Dr. Cal Roberts himself! Adapting audio from an interview on “The Doctors Podcast,” Dr. Cal describes his work as a program manager for a project on eye transplantation called Transplantation of Human Eye Allographs (THEA). Funded by a government initiative called ARPA-H, THEA is bringing some of the country’s finest minds together to tackle the complexities of connecting a person’s brain to an eye from a human donor. 

This episode also features an interview with Dr. Daniel Palanker of Stanford University. Dr. Palanker is working on technology that can artificially restore sight through prosthetic replacement of photoreceptors. Having proved successful in animals, Dr. Palanker and his team are working hard to translate it to humans. 

And if that can happen, then something once considered impossible could finally be accomplished!

Podcast Transcription

Roberts: For thousands of years, people looked up at birds in the sky, desperate to join them in flight. It seemed like human flight was an impossibility, only achievable in legends and myths, like the story of Icarus. And we all know how that one ended.

But as our understanding of physics evolved, scientists and inventors became determined to take to the skies. And on December 17th, 1903, the Wright brothers successfully flew over the sandy dunes of Kitty Hawk, NC. Fast forward 120 years and now we’re flying thousands of feet above the earth, crossing the globe in the span of hours.

What we now take for granted was once considered an impossibility. So, the next time someone tells you something is impossible, think again. 

I’m doctor Cal Roberts and this is On Tech and Vision. Today’s big idea is vision restoration. Restoration of sight has been the pinnacle or brass ring that researchers and scientists have been seeking for hundreds of years.

In this episode, I’m excited that we will be delving deep into two ways in which we may be able to reach that goal. One way that true vision restoration could be achieved is through a whole eyeball transplant, replacing an eye that cannot see with one that can. Right now, eye transplants are purely cosmetic. Nobody has found a way to make a transplanted eye actually see. It’s a challenge I’ve taken on personally. And I’ve been able to work on restoring vision as a program manager in the new Advanced Research Project Agency For health ARPA-H which President Joe Biden announced in March 2022.

Biden: Folks, one of the most advanced scientific agency in the world is what they called DARPA, and they’ve done everything from coming up with the Internet to global positioning, a whole range of things. For some time now, I’ve been pushing that we have a similar thing in the health area that we call ARPA-H instead of DARPA, that’s where it kicked off today.

My colleagues in the House and Senate agreed to fund it for a billion dollars. It has the ability to make an enormous breakthroughs focusing on cancer, focusing on obesity, focusing on diabetes, a whole range of diseases that, if we put the resources of the federal government and the scientific genius we have behind it,  we’re going to make significant breakthroughs. There’s so much we can do.  There’s so much we can do. Remember anything in America is possible when we work together.

Roberts: I have the pleasure of leading an ARPA-H program on eye transplantation called Transplantation of Human Eye Allografts or THEA. I spoke to doctor Robert Cykiert about it on his podcast called DoctorPodcasts. He’s kind enough to lend us portions of that interview.

———————————

Cykiert: So, tell me how you got involved in ARPA-H and tell us about this THEA project which is transplantation of an entire eyeball. 

Roberts: So, like you, I spent a lot of years doing cornea transplants and like you, I am sure, patients all the time would say to me doc, why do you just transplant the cornea? Just transplant the whole eye. 

Cykiert: I get that all the time. Patients ask me that all the time.

Roberts: Exactly. And so, we all have developed our own 20 to 30 second  explanation to dismiss this patient so we can get onto what we really wanted to talk about right? Anyway, I was at JP Morgan healthcare conference.

It’s this annual meeting of everyone in the pharmaceutical and biotech area, and on the second day on Tuesday, the speaker was Bob Califf, the Commissioner of the FDA.  And he brought with him, Renee Wegrzyn, who had just six weeks prior, taken over as the director of ARPA-H, and Renee gave this very inspiring talk explaining how this was an opportunity to really advance healthcare as we know it to be able to do things that couldn’t have done before because there was no funding mechanism to allow that. 

And as I’m listening to her and I’m getting inspired, I think about those patients who said to me, doc, doc, doc  why don’t you transplant the whole eye? So ,after she spoke, I went up and introduced myself and I said, are you looking for program managers in the vision sciences? And she says yes, we don’t have anybody in vision science. So, we exchange business cards and I was hired in September of 2023.

Cykiert: I’m sure you read about, the whole world read about last year at NYU Langone Medical Center, NYU Grossman School of Medicine, where I’m on faculty. They did an actual eye transplant in a patient who had severe injuries to the face and and lost the eye. The problem with that, as you know, is you can’t connect the optic nerve from a donor eyeball to the patients optic nerve because once an optic nerve is cut the nerves basically don’t transmit, so while this patient has a new eyeball which looks like it’s functioning, it’s actually not. 

The eye doesn’t see because in order for the eye to see, it has to be connected to the brain. How do you propose to have the optic nerves communicate with each other? The patients optic nerve and the donors optic nerve. I’ve read about growth factors and other, you know, medications. Can you tell us some more about that?

Roberts: So, I’ll give you a couple of the type therapies that people are proposing. So, one would be what’s called a nerve wrap. And so, this would be a cuff that went around the junction between the donor optic nerve and the recipient optic nerve. So, you would bring them together and this would go around it. 

And inside this wrap there would be a framework or a biostructure that promotes the growth of nerves, and there would be growth factors in there, ones that have been shown to do two things. One is they keep cut nerves from degenerating and second, promote new nerves from growing and connecting. 

Another idea –  as embryos, when babies develop the nerves from the eye to the brain follow that way because there is an electrical stimulation, there’s what we call a potential gradient to allow electrical current. It guides the growth of the nerves.

And so, there’s been work that shows that actually you could do that in adults and give electrical current, electrical stimulation that would encourage nerves to grow and point them in the right direction. 

And another area, of course, is the use of stem cells. And so that stem cells can become optic nerve cells and they can act like bridges and bridge from the donor optic nerve to the recipient optic nerve. They not only act as bridges, but they also have growth factors in there that make them grow and what we call neuroprotective or factors that keep the container from deteriorating. 

And so, these are the kind of things that people have been working on and we’re really excited about the potential of now really making happen.

Cykiert: When do you think you’ll be ready through the THEA project to actually do this surgery and give a patient one of these optic nerve regeneration procedures and medications.

Roberts: What we did is we came up with a whole road map of how we get from where we are in the knowledge that we have today to where we want to get to in terms of restoring sight to people who are blind and we broke it down into what we call 3 technical areas. 

Now the first technical area has to do with the harvesting and the preservation of the donor eyes. So different, Robert, from what you and I know doing cornea transplants, that eyes for whole eye transplant have to come from organ donors. And then we need to be able to keep these eyes alive. Because, unlike corneas, as you and I know they will last for a week in the eye bank, retinas and optic nerves last minutes and so we have to come up with a whole different way of how to preserve these eyes and keep the optic nerve and the retina and all the other structures within the eye viable from the time that the person donates their eyes to the time of the surgery. And that might be a day that might be two days if it has to be transported someplace until you get a team together or whatever it is required in order to do the surgery.

And so there will be these chambers with perfusion of artificial blood and whatever to give nutrients and oxygen to the eye. So that’s the first technical area. That’s the organ retrieval and the preservation.

The second technical area is the one we’ve already talked about. How are you going to hook up the optic nerve and restore the signal that is generated from the eye to the brain of the recipient in a way that the recipient can receive that message and do what the brain does, which is interpret the signal into what we call sight.

The third technical area has to do with the surgery itself. How are you actually going to do the surgery? How are you going to connect everything? Are you going to connect just the eye? How about the muscles? Are going to donate the muscles or keep the recipient’s muscles and where the vessels going to hook up to? Where are the arteries and the vein, and how about all the other nerves? Because as you know, it’s not just the optic nerves, other cranial nerves. There’s cranial nerve 3, 4, 5, 6. All these are involved with vision, whether it’s innervating their cornea or the iris or the muscle. How you gonna get them all connected and get working? 

And of course, when you’re dealing with any kind of organ transplantation, you gotta be concerned about rejection. You know, we know that eyes are very sensitive to even small amounts of inflammation. And so how we’re going to make sure that the recipient tolerates getting this new eye with minimal amount of inflammation and the least chance of rejection as possible?

And then finally, how you going to monitor these patients post-operatively? Because the vision is not going to be restored instantaneously. It’s going to take weeks, months for the nerves to regrow and reattach. How do you know that it’s healing? So, we need to come up with all kinds of new imaging techniques to be able to look at the optic nerve at a cellular level, watch out how the nerves are healing to make sure that things are going in the right direction so that we know that our therapies and our surgery is effective even if the patient hasn’t gotten their vision restored yet.

——————–

Roberts: Now let’s hear about some equally exciting research being conducted by Dr. Daniel Palanker of Stanford University. He’s making great strides towards restoring vision artificially through electronic means. 

So, talk to us about your work and vision and more specifically in vision restoration. Why is this an area of interest?

Palanker: Well, first of all, it’s a very important issue currently not yet addressed. A lot of different approaches are being developed. We tried couple of those approaches ourselves. One of them is transplantation of photoreceptors and the other is prosthetic vision.

So, we started actually with transplantation of photoreceptors because we saw plasticity. When you coagulate a small patch of retina photoreceptors selectively without damaging in retina, you can see that cells migrate and rewire and they reestablish connectivity between remaining photoreceptors at the edges of the lesion that then migrate inward and bipolar cells, which got disconnected from the photoreceptors due to damage. 

It was observed in rats and rabbits. And so, we wanted to see if transplanting a sheet of photoreceptor, say, from periphery to the center where we can create a model of retinal generation will enable this plasticity and vision can be restored. So, we harvested threads of photoreceptors, transplanted them and they survived. But we saw very little reconnection. So, there was very few synaptic connections we established between photoreceptors and bipolar cells and that approach didn’t succeed, at least in our lab. It may require more tricks, you know, in the field of retinal cell biology to unlock this potential of rewiring. I think one of the biggest hurdles so far is really enabling the connectivity of photoreceptors with differentbipolar cells.

The other approach that did succeed in our lab is replacing photoreceptors. We replaced them with array of photovoltaicpixels. And we do it in retinal degeneration called age-related macular degeneration, and specifically its atrophic phase called, called geographic atrophy. In this condition, the central vision is lost may be 4mm diameter scotoma, so it’s a blind spot right in the middle of the visual field, and that has a debilitating effect on high resolution central vision. So, patients cannot read, cannot recognize faces.

But peripheral vision is preserved, so they can ambulate when they walk, they don’t bump into chairs and tables, but reading ability and face recognition is the highest on their wish list.

So, for this condition, we developed a photovoltaic array that is placed under the retina instead of lost photoreceptors. These pixels like little solar panels. They basically convert light into electrical current and this current flowing through the retina polarizes bipolar cells that are sitting right above this implant,just by proximity. Polarization of bipolar cells results in release of neurotransmitter from their terminals into ganglion cells, and that leads to spiking of ganglion cells, and these signals propagate to the brain. 

So basically, a lot of normal retinal signal processing is preserved because most of the encoding of information occurs between bipolar and ganglion cells, mediated by amacrine cells, and that part of the retinal circuitry is reserved to large extent in age-related macular degeneration. 

We also saw that we can reach acuity matching the pixel size and we measure it by grading acuity like in small children. You flip a grading on the screen. If there is a response and it means that the resolution is sufficient for resolving this image. If there is no response, it means that the grading is too fine, and it basically appears as gray screen.

So, using this technique we measure acuity with different pixel sizes from 100 microns down to 55 and with that, we decided to commercialize this technology and see what patients can actually see and this was done by a French company, PixelVision. So, that company took our patterns and built a human version of our chip and conducted now already 2 phases of trial. 

The first was feasibility trial that was done starting in 2018. I think the first recruitment, so implementation probably started in 2019 and then second phase started about a year and a half ago. And the results were remarkable. The patients saw form vision, unlike all the previous attempts where patients reported phosphinesand they cannot find where the light is and by scanning figure out what they’re looking at.

Here it was pretty much instantaneous from vision, patients can recognize the pattern, lines or letters and what was also important and remarkable is that they could do it simultaneously with peripheral vision. This implant is activated by light, but ambient light is not sufficiently bright. So, we use augmented reality glasses that project images captured by a camera on the glasses using more intense light than ambient. But to make it invisible, we use infrared wavelengths.

So, in this system there is also a capability of processing images between the camera on the glasses and projector. One of the features was zoom, so patients could actually zoom in to the extent they feel comfortable from the level of 1:1 to 4:8.

And it was surprising that patients were very comfortable zooming even to a large extent, up to 8X and read much smaller font than they could without. So, first of all, the clinical trials established that they have monochromatic central vision simultaneously with peripheral vision and that letter acuity matched the pixel size. That is remarkably reproducible. Now we have more than 30 patients and we know that the results are exactly average acuity matches pixel size with zoom, so acuity responding to 100 micrometer pixel is 20 / 420 but with zoom they could get all better than 2100 and some as good as 20 / 60.

With that level of visual acuity, some patients can read directly from a page of a journal or a book. They can use it at home. They can read signs in a subway, find directions and names of stations, and so on.

So, it’s remarkably successful. So, currently the size of a pixel is 100 micrometers, and our next step is to make pixels smaller. 

Roberts: So, making the resolution higher, is that just an electronics issue or is that a biology issue as well?

Palanker: It’s both. Smaller pixels will create more confined electric fields. Basically, it’s like little fountains. If you make it smaller and smaller, penetration depth of that fountain into the tissue will also shrink and we will not be able to reach a target neurons which are tenths of microns away from the chip. So, when you make pixels smaller, you limit yourself to the amount of charge you can safely inject to the amount of light that you will need per pixel, because there are many more pixels now in terms of thermal safety. And these are limitations on device side, but you also limit penetration depth so you cannot polarize cells or bipolar cells, and that’s a limitation on the biology side. So, they can kind of come as a package.

The way we solved this problem is by going into the third dimension. So instead of flat array we now have 3 dimensional arrays that we either have pillar electrodes or honeycomb electrodes, where cells migrate into honeycombs. Or we use current steering to control penetration depths.

Roberts: So, in the past you have given me the best explanation that I’ve ever heard as why it is that patients with ocular prosthetics such as these only see in grayscale, and I once asked you how do you get patients with a prosthetic to see in color.

Palanker: Yeah, color vision requires color opponency. So, it’s inputs from different cones. We have three types of cones, short, medium and long wavelengths or blue, green and red. They provide antagonistic input to bipolar cells, red and green, and this difference is an input that then propagates through the retina.

So, to reproduce this system we need single cell selectivity and stimulation. We need to be able to actually stimulate not only one bipolar cell, but with two different antagonistic inputs into bipolar cell, which requires not only smaller pixels, which we may be able to do, but also much fine interface where you can guarantee that your electric field is confined to 1 cell which is about 10 micrometers in size for bipolar cells and that currently is not the case. 

Currently, our pixels are at least 20 microns and the electric field spreads and we certainly affect more than one cell. So, that’s why currently its average color, which is kind of white yellowish. I think for humans though if you can provide even monochromatic vision at the level of you know 20/80 and 20/40 it will be very useful and that will bring them back into our society to a large extent and that is a viable product. 

So, it’s like with cochlear implant. You know, cochlear implants don’t restore ability of people to hear music. But it restores hearing to the extent sufficient for speech recognition, and that is good enough to bring people into society and make them productive and improve quality of life significantly. And that’s our goal as well. So, first step, at least to make it useful product with the grayscale central vision.

Roberts: So, you’ve spoken to us about two different ways to restore vision. First, you talked about transplanting cells, photoreceptor cells, living cells, either from the periphery of the retina to the to the central area, or growing new photoreceptor cells and implanting them.

And then you talk to us about the artificial approach or the electronic approach to doing this, compare the two for me, because of the theme of this podcast, is really looking at transplantation versus electronic stimulation. What’s the advantage of each?

Palanker: So ideally in medicine at least, the ideal therapy is restoration of full functionality and if we can grow back photoreceptors and make them reconnect to bipolar cells and undo all the rewiring that the retina underwent during degeneration and restore full extent of vision, that would be the ideal outcome. I don’t know if it’s possible people are certainly working on it.

The replacement of photoreceptors with electronics seems to work, at least in our case, and again by analogy with cochlear implant, people didn’t find any better solution than cochlear implant for restoration of hearing. Now, for probably four decades, it’s a standard of care, even in small children that really enables the deaf children to develop their speech capability and not only hearing, but also speaking and live pretty much normal life. 

So same with vision, if one day the full restoration will be enabled, then yes, it will be probably beneficial compared to limited grayscale vision that we can provide to this electronic substitute of photoreceptors.

Roberts: One of the biggest challenges when it comes to electronically unlocking vision restoration is getting the brain to communicate properly with that technology.

I asked Doctor Palanker more about it.

So, the location of the brain machine interface is a topic that that we’ve discussed in in previous episodes here. And where do you put the electronics? In your situation we’re putting them at the location of the retinol photoreceptors. We hear about from the Elon Musk group about putting them right into the brain.

Why would somebody want to put it directly into the brain? Or why would somebody want to put it where the photoreceptors are?

Palanker: So, for us, the consideration for restoring central vision in an AMDpatient is that we wanted to preserve retinal signal processing to the extent possible. And that means that you want to start as slow as possible in a chain of signal processing and stimulate the first cells basically after photoreceptors because the encoding is relatively simple there and location is preserved and so that’s exactly what we did. 

We replaced photoreceptors with photovoltaicpixels and we indeed see that we have many features of signal processing preserved this way. There are groups working on a retinalapproach where it is more demanding but still in the retinathey want to skip one level of processing between bipolar and ganglion cells mediated by amacrine cells, and hopes that still the signal encoding is not too complex and you can do stimulation of ganglion cells. But in this case it’s already spiking neurons, so you have to stimulate one spike in, one pulse in, one spike out. Kind of encode exactly bursts, which we don’t have to do with bipolar cells because they’re not spiking. It’s an analog cell.  

But if you want to restore functionality which is controlled by cortical activities, such as for example, for paraplegic patient, don’t have the capability of controlling their muscles, but they can think of a motion they would like to execute. Then you need to read out that activity and that has to be in the brain because there is no other place to read out this activity. So, in this interface there are two or three options. You can place electrodes outside the skull, but then the quality of signal is maybe sufficient for controlling one or two degrees of freedom only, like moving a cursor maybe.

For high fidelity, you can place a ray on top of the brain under the skull, but not penetrating. Several companies are working now on, for example speech restoration, where you read signals from the areas controlling speech and then recognizing the intended words and basically pronouncing using an electronic synthesizer using the voice actually of that person recorded prior to injury.

So, that’s done with the ECOG arrays, which are on the surface and also high fidelity of that sort of interface you can achieve with penetrating electrodes.  Penetrating electrodes as smaller they come closer to neurons, they can pick up signals in a much higher fidelity but they’re dangerous because you can injure the brain and the brain may respond by creating what is called glialor fibrotic seal around the electrodes, rejecting basically the foreign body of the electrode. And very often it means that these electrodes become deactivated. They cannot read the signals anymore after some time. 

One way to overcome this limitation is by miniaturizing these electrodes, making them very thin, solar scale and very flexible. That’s what Neuralink is doing. These electrodes, they implant electrode arrays that are, I think, 30 plus electrodes on a single shank, which are about 20 micrometers in width thin strips of polymer with electrodes posed along it and they don’t seem to create a scar, which means that they should be able to stay there and remain functional for longer. 

However, recently there was a report that these electrodes in the first Neuralink patient, actually were withdrawing on their own from the brain. Probably because of some sort either shrinkage – we don’t know exactly what’s happening, but they are moving out and that one of the problems, Neuralink should address. So, this is not a done deal. There is no technology that enables very high-fidelity stable interface yet.

But it’s moving along and might one day be possible.

Roberts: Doctor Palanker and I highlighted different approaches to vision restoration being pursued. I spoke about the challenges of eye transplantation versus electronic or bionic eye in my interview with Doctor Cykiert.

——————-

Roberts: So many people have asked me why I spend all this money on transplanting eyes. Why don’t you take that same money and just build a bionic eye? With a bionic eye you wouldn’t need people to donate. You wouldn’t have to worry about rejection. You could just take it off the shelf, put it in someone, and they’d be able to see. Wouldn’t that be a better alternative than doing an eye transplant?

Robert, as you well know, in 1964, sixty years ago, NIH created a task force to come up with a mechanical heart. An artificial heart.

Cykiert: Right. 

Roberts: And the first mechanical heart was implanted in 1982 and it was something called the Jarvik 2 heart. Well, now 42 years after the implantation of the first mechanical heart, there’s not a single person in the world walking around with a mechanical heart. All that work and all that research and all that effort to come up with a mechanical heart. Our transplants are still state-of-the-art. And so, while I believe that there is a role for a bionic eye or mechanical eye, what I really believe is that everything that we learn from doing an eye transplant will just make it better and easier when we do eventually come up with a bionic or mechanical. Here’s the analogy that I use.

So, you remember Morse code, there’s one person at one end with a little tap who’s going tap, tap, tap, tap, tap, and the person at the other end as a speaker and who hears the dot, dot, dash. Morse code only works if both people know Morse code, so it is with the eye and the brain.

The eye speaks in its own language of dots and dashes, like Morse code. Then the brain understands these dots and dashes like Morse code, and then has its ability to convert dots and dashes into what you and I refer to as vision. And so that the big development is to learn how to speak brain language.

——————

Doctor Palanker is addressing similar issues in his research. He thinks that if we can unlock the so-called brain language, we could unlock even more seemingly impossible accomplishments. For instance, someday we could all be telepathic.

So as we all know that our brain has two functions is one to take in information, and the other is to use that information often to send messages out, whether it’s sending messages out to our limbs, as you said, sending messages to our hearts, to our lungs to say “breathe,” those type of things, and so that sending them messages out from the brain to the periphery as you describe, is a task that is potentially less challenging than actually teaching the brain how to receive messages.

Palanker: Yeah, so far I think finding correlation between the signals that brain generates and intended action like if you read the signals from the motor cortex, you can usually decode the intended motion because these signals are supposed to move a few muscles.

Same with speech. They’re supposed to move, you know, muscles controlling vocal cords and rather few of them. So, this task seems to be possible, but the more complex is a task, if it’s abstract thinking and so on, the more challenging will be to find correlations even for readout and even more so more difficulties to write in the code to stimulate the right cells with right code for the brain to understand language of that stimulating pattern.

Roberts: But you painted us a picture that’s very optimistic, that if you look out into the future, so much is possible.

Palanker: Yeah, it looks like we are moving fast in this direction. There are multiple companies pursuing real medical applications now, which certainly will benefit patients. Companies in the US, in Europe, in Asia that have real products now that enable patients with injury in the spine to walk again to speak again and control functions of a computer with thinking about moving cursor or clicking on different keys. So yeah, certainly it’s moving forward.

We don’t know how fast and how far it will go. We are all encouraged by the success of cochlear implant. It is surprising how successful it is given the simplicity of the cochlear implant. Only 12 electrodes instead of 30,000 neurons, and therefore that’s encouraging that you might be able to restore certain functions in a system much simpler than real visual system, but to the extent that is limited, for example to central vision, to grayscale and so on, and step by step. I am very curious to see the first applications beyond the medical applications. What will be the first normal kind of applications where we will expand our functions rather than restore them.

We can think about other aspects of brain machine interface which takes you maybe into the realm of capabilities that humans never had. If you enable artificial sensors or enable brain to brain connectivity so you can communicate without verbalization, that would open completely new capabilities that humanity’s never had.

And from this perspective, the brain machine interface is open not only to the capability of restoring functions, but providing new functions potentially. And that makes it even more appealing. So, we are working not only in a field of medical application restoration of loss functionality, but opening up new windows into the world.

And for that reason, I think the work we are conducting may have much broader applications later on. Just when you establish high-fidelity brain-machineinterface. In one organ, you may be able to apply similar technologies into other parts of the brain and enable other functions. The sky is the limit.

Roberts: Well, this is fascinating because I think I’ve ever heard this before, so if I can pursue this a little bit more.  So, in general, we’ve been talking about inputting data into the brain, and so how can we get information into the brain that was interrupted because of disease? If I’m hearing you correctly ,you’re also talking about how can we send data from the brain and send it externally, and so that the brain, besides being a receiver of data, is a transmitter of data.

Palanker: Right. So, there are two types of interfaces, readout and write in. So, readout is a detection of signals which is considered much easier than write in because as soon as you read enough information from say motor cortex you can decode correlation of these signals with intended motion and in this way activate prosthetic limbs, for example.

So currently, most of the companies in the field ofbrain-machineinterfaces are pursuing exactly that. The restoration of motion, restoration of speech. These are readout interfaces. Write in is much more challenging. If you want to deliver to brain a code of can function like vision or hearing,  it needs to be encoded in a way the brain will recognize like signals that convey this information. As we have seen with visual prosthetics, both retinal and cortical, it’s very difficult. People tried many different approaches and pretty much ended up with eliciting phosphines, so it’s a perception of light, but not really pattern vision, not form vision. That seems to be much more challenging, requires much more proper code.

And the further away you are from photoreceptors, the more complex and more distributed is the information. Bipolar cells already perform with spatial and temporal filtering on this input. Ganglion cells converge into binary code of action potential. You can call them digital cells in language of electronics, and the code becomes much more complex. It’s a firing rate, it’s a correlation between rate in different cells. 

There are about 30 types of ganglion cells, or at least more than two dozen that deliver different aspects of information. They sample an image in parallel mosaics and deliver these 20 plus streams of data, each sampling the whole visual field and delivering them in parallel. Actually, to different pathways in the brain and then they further separate until they merge in the percept, so the further away you move, the more demanding is a task to encode it properly, to deliver to right cells in the right parts of the brain at the right moment and encoding right type of spiking.

So, the write in is much more demanding tasks. But yes, if you want to establish communication between, you know two brains, one should be sending signals and the other receiving. Both types of interfaces have to be developed.

Roberts: So, if I’ve read one book and you’ve read the other book, there should be a way that I can be learning from what you’ve read and you’ve learned from what I’ve read externally.

Palanker: Yeah, I don’t know how much information we can internalize. There is a limit probably of how many books you can remember. I probably don’t remember many of the books I read.

But it all adds up to some sort of intuitive understanding of the world. Maybe you don’t remember every word in every equation, but you remember the foundations. How to think about this phenomenon and that phenomena. What are the essentials like conservation laws, symmetry laws and so on. And then from there you can maybe find the right book to refresh your memory.

So that sort of intelligence, I think is what is education distilled to. It’s understanding how the world functions and where to go if you need more specific information.

Roberts: And it’s more and more the way I use my computer when there is an idea that I think I remember, but I don’t remember the details. I try to put in some keywords into the search and find the information that I thought I knew but I couldn’t recall.

Palanker: Exactly. So that sort of thing may be internalized. So, you will think about searching but not by typing, but thinking about that and accessing external memory more intuitively.

Roberts: I love it. I love it.

Roberts: As we’ve learned today, impossible is merely a mindset, just like the inventors and engineers who dared to believe we could someday take to the skies, there’s real work being done to achieve the so-called impossibility of full vision restoration.

We don’t yet know when or how it will be achieved, but we know one thing for sure. Moon shots are worth taking because even if you miss, you’ll still land among the stars.

Did this episode spark ideas for you? Let us know at podcasts@lighthouseguild.org and if you liked this episode, please subscribe, rate and review us on Apple Podcasts or wherever you get your podcasts.

I’m Doctor Cal Roberts. On Tech and Vision is produced by Lighthouse Guild. For more information visit www.lighthouseguild.org on tech and vision with Doctor Cal Roberts produced at Lighthouse Guild by my colleagues Jaine Schmidt and Anne Marie O’Hearn. My thanks to Podfly for their production support.