About The Innovation™…Arthrolense: Where the Idea Began

Why We Started

I’m always looking for ways to improve what we do in orthopedics — to innovate the status quo. I talk with a lot of people about the problems we face every day in the OR, and I tend to come up with what I’d call crazy ideas about how to solve them.

One of the people I talk to most is Dr. David Backstein. We kept coming back to the same conclusion: one of the biggest problems we face is visualization — how do we actually see the decisions we’re making in the moment, and the tradeoffs that come with them?

I also happen to have a strange passion for eyeglasses. I own about 80 pairs — yes 80 and everyone is prescription. I go to the same optician and one day, picking out a new pair, I said to him, “Darren, it’s funny — everything we do in orthopedics is mechanical, right-angled. Could you etch something into my lenses that would let me see the angle of a tibial cut?”

We played with that for a while — little etches, crosshairs on the lens. None of it really worked. But somewhere in that process I said, “Darren, build something kind of like a heads-up display in a car or a fighter jet cockpit.” That sent me down a rabbit hole reading about heads-up displays and mixed reality — the idea that you could see an image floating in space and use it however you needed.

That’s when I called David. “I have this crazy idea,” I told him. “What if we used mixed reality to make surgery easier and better?” This was before most people understood what mixed reality even was, let alone what it could do.

David said, “I love it.” And he said one of his closest friends, a private equity guy named Brad Nathan, might want in. We called Brad that same day. “I’ll tell you what,” he said, “why don’t we go in together and see if we can make this happen?”

An idea a surgeon has is like the tip of an iceberg — the one-centimeter that’s visible above the water. The idea is such a small part of bringing a product to the OR.

I knew nothing about the depths of the technology, and nothing about how a product goes from raw concept to market. It’s humbling. As surgeons, we sit in design meetings for a new hip or knee and say, “it should have a collar, and the collar should look like this.” We rarely appreciate what it actually takes to make something like that real and get it to market. That’s how Arthrolense started — an idea, a handful of believers, and a little bit of money to get things moving.

The Vision

I’m a surgeon. I think like a surgeon, not an engineer, and not the way a company thinks. What I wanted — what I saw other surgeons struggling with — was the ability to see the decisions and tradeoffs in front of them during a procedure, in order to get the best possible result. That starts with turning images into data, and data into decisions you can actually see play out on the patient in front of you.

The hope was better lines, better numbers, better balancing — and beyond that, a real understanding of what “balanced” even means. When a fellow or a resident comes to me, I tell them how I balance a knee. But the truth is, none of us actually knows what perfect balancing looks like. We use what we were taught, and what our own clinical experience has shown us. I would never say my outcomes are better than someone else’s — I just know what I do, because of the tens of thousands of cases behind me. That doesn’t mean I’m always right, and there’s no objective evidence either way.

If technology can give us that objective evidence, that’s the biggest reason it matters. It’s not about making the cuts. It’s about knowing, after thousands of cases, that four degrees of slope beats three for a certain kind of patient — and we don’t have those answers yet.

Objective data gets us there, and gets us there faster — which means more informed decisions, fewer complications, better outcomes across the board. That’s the real “why” behind this company. The vision was never just a tool for a single surgery. It was a platform for learning, case after case, so the next generation of surgeons inherits real evidence instead of tradition alone.

Building It: The Journey

It started with me, David, and a finance partner, and honestly, I had no idea where to begin. So I started calling software designers everywhere, looking for people current enough on the technology to be at the cutting edge of it. We found a terrific software team out of Armenia and started building on HoloLens.

We quickly learned that mixed-reality headsets — HoloLens included — didn’t have anywhere near the processing power we needed. So we decided to use a “tower”: the same kind of setup used in navigation and robotics today, sitting at the foot of the bed with cameras attached.

The pathway we chose still uses infrared tracking, similar to navigation systems on the market. But we had to ask ourselves what question we were actually trying to answer, because robotics already does a great job in this space — Mako has been around for two decades and is a dominant player. But the robot is fundamentally an execution tool. To execute, you still need a CT scan to generate a good 3D model, select registration points, and register the scan to the patient’s anatomy.

We wanted to answer a different question: What is missing? How do we know the robot is right? What can we add — to the robot, or independent of it — that solves problems we haven’t solved before?

So we introduced a spatial mapping or structured-light camera — genuinely the cutting edge of orthopedic technology right now. Interestingly, the AR/VR headset itself was the part surgeons liked least. It was bulky, and it got in the way of seeing, which was the core issue we were addressing to begin with – enhanced visualization. What other surgeons responded to instead was a sterile, draped iPad on the field, paired with structured-light imaging. That combination answered far more of the whys and hows than the headset ever did, and it’s what’s pushing us to the next stage of this technology.

We went through five full versions of this system to get where we are today. You have to be willing to change your mind without losing your conviction — stubborn enough to believe in what you’re building, flexible enough to admit when something isn’t working. We let go of HoloLens. We rebuilt the software architecture more than once, because what I liked wasn’t always what other surgeons needed. We found engineers who could actually build it. And then, once we thought we were done, we learned what it really takes to get something through the FDA — proving the tower is safe when it’s tilted, dropped, shipped, and knocked over; proving sterility holds up; proving every single component behaves the way it’s supposed to, every time.

What Structured Light Actually Does

Robots do a good job today. You register to a CT or MRI, which can be onerous - bring in a haptic arm, make the cut, and then check whether you got it right. If you didn’t, you figure out what went wrong and correct it. But here’s the thing — surgeons are generally comfortable making cuts. It was never the haptic arm they fell in love with; it’s bulky and can slow a case down. What surgeons actually want is the confidence, capability, and comprehensiveness of the brain behind it — e.g., seeing the balance is right before they cut – that improves the efficiency and effectiveness of the procedure.

Structured light gets us there. Intraoperatively, in about seven seconds, the camera throws roughly a thousand points of light onto the bone and instantly builds a sub-millimeter-accurate 3D model of exactly what’s in front of you in that operating room. We call it a digital twin — a virtual twin of the real knee, visible on a sterile tablet, accurate to a fraction of a millimeter. Additionally, if this technology is integrated into a robotic platform, it can register that preoperative image in seconds.

Because it’s virtual, you can do anything with that digital twin — including layering in CAD models of the actual implant before a single cut is made- all from an image taken instantaneously in the OR.

We’re currently working with Waldemar Link/LinkBio Corp. in the United States, using the LinkSymphoKnee. This knee offers several features that work exceptionally well with the system we’re developing. Beyond its size range and implant options, the Titanium Niobium Nitride coating- which I use in almost every patient- is a major advantage. The implant design itself is also in harmony with native anatomy and lends itself well to individualized alignment. For example, the trochlear groove is uniquely engineered to accommodate anatomic patellar tracking while also helping prevent overstuffing during early to mid-flexion, an area where many other systems haven’t focused. With the CAD models from the LinkSymphoKnee integrated into the Arthrolense system, you don’t just see where the cuts will be — you see exactly where the femoral and tibial components will sit, and how a trial will rotate. You can check symmetry, stretch the ligaments virtually, and confirm the fit before committing to anything. No CT is needed, and no competing system that we know of gives a surgeon a real-time, real-life preview like that. It means a surgeon can be a surgeon before making the cut, instead of simply trusting a number on a screen from a selected group of points.

Because you can manipulate these models freely once the image is captured, we’ve built in more than standard mechanical and kinematic alignment. There’s a technique we call PIKA — patella-initiated knee alignment — where the system tracks how the patella moves and uses that to guide rotation of the femur, then varus-valgus correction, then the distal femoral cut. It’s a way of balancing all three compartments of the knee, not just the tibiofemoral joint, with the patellofemoral joint as part of the equation from the start.

The system will also flag, in real time, whether a cut risks notching the anterior femur or overstuffing the joint — and by how many millimeters. If you’re worried about notching and want to anteriorize the femur a millimeter or two, you can see the consequence immediately and decide, for instance, to trim the patella slightly thinner to keep the joint balanced. All of this is possible because the digital twin reflects exactly what’s happening on the actual patient, in real time, before a single cut is made.

I want to be careful not to overstate where we are — but this is the direction the system is built to go.

What’s Ahead

The growth potential is the part that excites me most. In this first generation, we still select sixteen to eighteen registration points by hand. But once the system is in early evaluation and we’ve gathered enough clinical samples, the system is built for artificial intelligence to be able to recommend those points—accurately enough to avoid a mistake like picking up an osteophyte instead of the true bone surface. Imagine complete registration in seconds; Manual point-picking, and the fiducials — those metal arms with infrared markers that stick out of the bone — will eventually be a thing of the past. Not in generation one, but they’re coming.

In revision surgery, even our first-generation system can already tell a surgeon exactly where the joint line sits, in both flexion and extension. Down the road, it should be able to tell you the augment size you need, the cone size, the hole diameter — rebuilding a stable knee to the joint line in an automated fashion - reducing guesswork, reducing inventory, and placing components exactly where the anatomic joint line tells you they belong.

But the part that matters most to me long-term is the data. Every point, every adjustment, every angle and millimeter change during surgery — our system captures all of it, alongside pre-op and post-op outcomes. Over time, that lets us ask real questions: what do the best 2% of outcomes actually have in common? Machine learning can tell us the answer, even if it can’t yet tell us what to do about it — “your best outcomes were cut at this angle, you may want to consider that.” Eventually, with true artificial intelligence, a surgeon could trust the system to recognize a given patient’s anatomy and ligament balance and identify the sweet spot most likely to produce the best result for that specific person.

That’s the future — technology that gives residents, medical students, and our grandchildren real, objective answers instead of inherited habit.

In the meantime, what it gives us today is feedback in the moment — a digital twin that lets a surgeon see what they’re getting into before it’s too late. Right now, that’s what matters most.

The Ultimate Motivation

There’s a moment I think about often: someone going through an airport security scanner, telling the agent, “I have a knee replacement.” And the agent asks, “Which knee?” And one day, the answer will be, “the one done with Arthrolense.”

I’ll just smile inside and think, that was my baby.

That’s the journey. It’s been a wild one, and it’s far from over — but I live for that moment, and everything that gets us closer to it.

Please note: Arthrolense 4Di Visual Guidance System is not currently cleared for use in the United States.