Matteo, the first time you witnessed an implanted electrode for deep brain stimulation being activated was at the University Hospital of Würzburg in 2012. What did you experience in that moment?

Matteo Voleno: The device is turned on, and the trembling in the patient’s hand stops. Immediately. That’s incredibly impressive, and I remember it very clearly, even though it was many years ago. That’s the heart of this therapy—and the reason why this field has never let me go.

Swen Woerlein: That’s exactly what it’s all about, isn’t it? I studied computer science—with a minor in medicine—because I never wanted to build abstract systems for banks or insurance companies. If my code helps someone whose life is threatened or fundamentally changed by disease—like someone who just wants to eat on their own again—that's a motivation you don't find in traditional software development.

What actually happens during this procedure—how would you explain it to someone who’s never heard of deep brain stimulation?

Swen Woerlein: We try to correct a malfunction in the brain using a tiny electrical impulse delivered through an electrode slightly thicker than one millimeter. The problem is the location: We target core areas deep inside the brain as small as a coffee bean. Every millimeter off can have dramatic consequences—for patient safety, for speech, vision, personality. It's like having to hit an exact spot in a three-dimensional maze with a needle in the dark. That's the challenge every functional neurosurgeon faces—and that's what we help solve with software.

Matteo, you used to work at Boston Scientific, the company that manufactures the implantable system for deep brain stimulation. This hardware is becoming increasingly powerful—at the same time, the number of configuration options is skyrocketing. Doesn’t that also make the procedure harder to master?

Matteo Voleno: Boston Scientific (BSC) has pioneered a new era in deep brain stimulation—defined by a level of spatial and temporal control over stimulation that simply didn't exist before. That opened up entirely new options for treating clinicians. And it's also why solutions like our software are needed, because this progress comes with complexity: The old electrodes had four contacts and sent the current evenly in all directions—like a light bulb. If the target area was right next to a speech region, you hit both: The tremor was gone, but so was the patient's ability to speak clearly. The new generation is more like spotlights: directional contacts, combined with fine stimulation control, allow the tailoring of stimulation to the individual patient's needs. Such flexibility, however, means thousands of possible configurations. The potential is enormous. But no one can try all of these out blindly on a patient—that would be neither reasonable nor medically justifiable. That's why, then and now, we co-develop our software solutions together with Boston Scientific.

And with your software, you can model this configuration before applying it to the actual electrode in the patient’s brain. 

Swen Woerlein: Boston Scientific brought image-guided programming to neurologists, for the first time in history. Our software provides the neurologist with a complete digital map: Where is the electrode located, and what structures are nearby? And then we visualize the so-called Volume of Tissue Activated—figuratively a cloud that shows which area of the brain the electrical stimulation field actually reaches. The neurologist sees in advance which regions of the brain will be affected by the stimulation. They no longer have to start from scratch.

Matteo Voleno: This makes a very tangible difference for the patient. A programming session after surgery—stimulating contact by contact, observing the reaction, and adjusting—may take hours and be exhausting. If we start with a solid foundation, these sessions become much shorter and the results are better.

And does the electrode always end up exactly where it’s supposed to be?

Matteo Voleno: Regardless of the skill of the surgeon, deviations may and will likely occur. Therefore, verifying placement accuracy is best practice. With our mobile imaging robot Loop-X, we can take a 3D scan right in the operating room immediately after implantation. The software superimposes this image directly onto the original plan—the surgeon can see down to the millimeter whether everything is correct. What this means becomes clearer when you know how it used to be done: The electrode was guided through the brain along several paths while taking measurements—three, sometimes up to five passes, to confirm the position. Today, a single path is sufficient in many centers—thanks to the evolution of imaging and a growing reliance on anatomical targeting. Fewer passes mean less risk, shorter surgeries, and less stress.

Less stress, shorter surgeries—what does that mean in concrete terms? 

Matteo Voleno: Many years ago, I saw a case that lasted nine hours—the patient was conscious the entire time. In the early years of deep brain stimulation, procedures lasted two days. Today, a well-planned procedure at a well-established center takes two hours and the patient is asleep during the procedure.

This isn’t just optimization anymore—it’s simply a different patient experience. This became possible because the quality of planning has improved so much that surgeons and neurologists can often do without or with a reduced electrophysiological check-up—precisely the steps that used to keep patients awake.

Swen, what are the biggest challenges in development? 

Swen Woerlein: We always want to develop solutions that work in all hospitals—and the reality is that there isn’t always an ideal development environment in the wide variety of clinics around the world. For example, we receive images from MRI machines that are heavily noisy and technically long outdated—and the next moment, we’ll get high resolution images from the most modern scanners. Our algorithms must work in both cases.

But the technical challenges are only part of the story. The bigger hurdle is often regulatory: Getting an algorithm from a promising research result to an approved medical product is an entirely different endeavor. It takes years, extensive validation, and evidence that it works not just on clean datasets from university hospitals—but under all real-world conditions, globally. And when you're talking about software that guides an electrode through someone's brain, that bar couldn't be higher. We have no room for error. Not here.

What role will artificial intelligence (AI) play for your products in the future?   

Swen Woerlein: Brainlab has been working in the field of computer-assisted surgery for 35 years—algorithms have always been at the core. AI isn't a panacea, but it opens the door to leaps that would have been out of reach for decades. What concerns me: The hurdle to actually develop an algorithm for a regulated medical device is significantly higher than in research. An algorithm that works well on a clean dataset from a university hospital is not yet a product. We need something that works worldwide under all clinical conditions, withstands approval by regulatory authorities such as the FDA, and that hospitals will pay for because it truly helps. My conviction is that the best solutions will come from a combination of AI and classical algorithmic methodology—not from one approach alone. 

Matteo, when you look ahead—what kinds of innovations do you think will create the greatest value for clinicians and patients? 

Matteo Voleno: For me, it comes down to one principle: the best technology keeps clinicians firmly in control while taking the tedious, repetitive work off their plate. Not to replace their judgment, but to free it up for what matters. 

The information is already there. Every scan, every planning step, every anatomical detail—it's all part of the workflow. The next step is using it more intelligently: bringing the right information to the right person at the right moment. That's what reduces variability, improves reproducibility and makes outcomes more predictable—wherever in the world the procedure takes place. 

You’ve both been doing this for a very long time. What still excites you about it? 

Matteo Voleno: The breadth of this entire therapeutic field. You can start with planning, then move on to navigation, verification and the neurology that follows—and every time you really dive deep into one area, the next one opens up. And throughout all of it, there's the privilege of working directly with the clinicians—learning from their expertise and seeing how what we build genuinely contributes to the quality of their work. That's actually what engineers dream of when they graduate. I think I've found it. 

Swen Woerlein: The interplay between significance and technical challenge is as exciting as on my first days in this field. The basic questions haven't changed: How can we best support clinicians in their work? How can we contribute to better patient outcomes? But the medical and technological landscape keeps evolving—there is always something new, something complex, something that pushes us further. Then there's the team: physicists, computer scientists, biologists, clinicians. That mix isn't something you take for granted.