Next on the Horizon: Frederik Ceyssens on Scaling Cortical Implants with ReVision

Frederik, how did you find your way into the bionic vision space and what motivated you to take on the challenge from the cortical side?

[Frederik Ceyssens (FC)]: I was a researcher at KU Leuven in Belgium, and had already done work on flexible electrodes together with Bart Nuttin, a neurosurgeon and DBS pioneer, for a basic science research project. After that postdoc I looked for a space to contribute my experience to, and it turned out there was still a lot to do in visual prostheses. I obtained a grant on this, together with Peter Janssen, a neuroscientist with a lot of experience in brain implants and the visual system, and also found some early-stage investors and experienced business people wanting to join for the ride. This is how the ball started rolling…

ReVision is advancing at a time when several major European bionic vision efforts (like Retina Implant AG and Pixium Vision) have faced tough headwinds. What lessons have you drawn from those earlier ventures, and how have they shaped your approach at ReVision?

[FC]: There are many—we try to speak with experienced people and as it turns out, there is experience around that goes even 30 years back now. The main lesson is of course that it’s not as easy as people thought to electrically project a useful image in the retina, and the primary visual cortex of the brain is very likely a better spot to implant. The available space in the retina is very limited, axons coming from a range of areas are passing over each other, many different cells which are separately wired are there, the retina must still be relatively healthy in the first place, etc.

This was also the conclusion of Second Sight, when they went from their Argus II to the Orion device. But, in my opinion, their electrode count was still too low for a real breakthrough. Still, for the same electrode count, the performance of the brain implant was a lot better as we learned at the Eye and the Chip conference. Another typical one (also something you could learn from YC) is that you need to make sure your concept is completely viable before you scale up. At a later stage, going back to the design table is very costly, even more so for a medical device.

Let’s talk about your implant, Occular. What makes this cortical device different from what others have attempted - and why now?

[FC]: Well, there is the choice to go to the visual cortex—and in our case, only the primary visual cortex, as beyond that the phosphenes that you elicit are too large. As most of that primary cortex is not laying on the surface, that means that there will be a lot of electrode arrays that will go up to a few cm deep, unlike other implants that focus on the brain such as Orion and ICVP, which remain at the surface or a few mm below. So we go to the brain, use electrodes not just on the surface but also deeper, as Martin Bak has shown that the obtainable resolution is more than 5x better then. The currents you need are also much smaller, just tens of microamps, compared to surface stimulation where you need milliamps.

On top of ‘just’ going to the brain, we did add some extra features that are unique to our approach, including those ‘virtual electrodes’ to improve resolution further. And then there’s the patented insertion method that we’ve developed, based on a resorbable coating that temporarily stiffens the implant during surgery.

The numbers are bold: over 1000 electrodes, flexible biocompatible materials, a scalable insertion system. What’s the single biggest technical innovation you’re most proud of - and what trade-offs did you have to make to get there?

[FC]: Well, using lithographic techniques it’s not so difficult to make a 1000 electrodes in the form factor required. However, you also want them to be flexible to cause minimal brain scarring over the long time. And then, actually inserting them in the brain becomes the real challenge, as well as the lifetime, as a few micrometers of insulation material suddenly have to do the work normally done by hundreds of micrometers of silicone in standard electrode arrays. I’m really glad we found a way to make that work. In terms of trade-offs, that means the implants still could be a little thinner and a little more flexible than the record numbers found elsewhere, but this is offset by the +15 years of lifetime (and counting) that we now have according to accelerated aging tests.

Where do you see cortical stimulation fitting in as optogenetics, gene therapy, and retinal approaches continue to mature? Is Occular complementary, or does it represent a clean break?

[FC]: That’s a lot of questions in one! Since Occular completely bypasses the eyes and optic nerve, it offers a universal solution that could help with virtually any cause of blindness. The other approaches you mention are typically less invasive, but also more narrowly targeted toward specific conditions.

That’s even true for retinal implants. The nerves in the eye need to be relatively healthy, which often isn’t the case—especially in forms of blindness that affect younger people. For example, the Pixium implant (now acquired by Science Corp. after Pixium’s bankruptcy) was designed specifically for patients with macular degeneration, where photoreceptors die but the inner retina remains intact. It restores a small region of central vision, and was tested in patients who still had some peripheral vision left. That combination can be useful, but only for a narrow group of patients.

In our experience, performance trumps invasiveness. Every completely blind person we’ve spoken to about visual prostheses has said they don’t mind whether the implant goes in the eye or in the brain. They just want something that works. And if a cortical device offers better performance (and we have good reasons to believe it will) people will choose the brain implant.

Can you give us a timeline check? Where are you now with pre-clinical work, and what milestones are you targeting over the next 12–18 months?

[FC]: We now have long-term pre-clinical data in monkeys showing excellent biocompatibility in the brain, along with proof of efficacy based on eye-tracking responses. This was done by the lab of Prof. Peter Janssen at KU Leuven in Belgium. Our next step is an acute intra-operative test in collaboration with our Hungarian partners Lucia Wittner and Richárd Fiáth (TTK): the implant will be inserted during resection surgery in epileptic patients, specifically into brain tissue that is scheduled to be removed in the same operation because it contains an epileptic focus. This allows us to validate the insertion procedure in live human subjects without additional risk.

In parallel, we're preparing a longer-term trial of the electrode arrays in a blind volunteer, together with our research partner Prof. Eduardo Fernandez in Spain. He has already implanted several Utah arrays in blind individuals with promising results, so this would be a natural next step—drastically increasing the cortical area we can interface with.

At the same time, we’re working on a fully wireless system prototype, since the earlier tests will still rely on a wired electrode array.

What does your regulatory roadmap look like? Will you prioritize MDR approval in Europe, or are you planning to engage the FDA as well?

[FC]: We’re aiming to launch in both Europe and the U.S., and plan to conduct our pivotal trials in the U.S. as well to support that goal. We’ll be submitting an FDA pre-submission soon to gather early feedback and help shape our regulatory trajectory. That said, the FDA has gone through some internal shake-ups recently, so we’re anticipating possible delays on that front.

If we fast-forward to 2030, what does success look like for Occular? Number of implants? Level of visual function? Something else entirely?

[FC]: Success by that point would mean restoring at least 3/20 vision (roughly equivalent to 20/140 on the US Snellen scale), and being ready to launch pivotal trials that position us for market access shortly afterward.

If you could instantly solve one technical or biological limitation holding you back, what would it be?

[FC]: It would be nice if the brain were a bit less twisted and bent. In humans, the visual cortex is buried up to 4 centimeters deep, making it far less accessible than in the monkey brains we've tested so far. Surgical planning to reach all the necessary areas while avoiding major blood vessels will be a real challenge. If we can access even 80% of the target region safely, we'll consider that a success.

And finally, what does ReVision need right now? Are you looking for trial partners, engineers, collaborators? How can the community help push this forward?

[FC]: We could really use a few more early-stage investors—specifically those willing to get involved based on the early human data that we have thanks to Eduardo's work. Scaling up at this stage is critical to making it all work. So if you can help spread the word, that would be greatly appreciated!