Perimeter Medical Imaging AI, Inc. (TSXV:PINK)
Canada flag Canada · Delayed Price · Currency is CAD
0.3300
-0.0200 (-5.71%)
May 1, 2026, 3:59 PM EST
← View all transcripts

2026 Bloom Burton & Co. Healthcare Investor Conference

Apr 21, 2026

Moderator

Good morning, everyone. Welcome back to Hall A. Next presenter we have is from Perimeter Medical, and we have Adrian Mendes, Chief Executive Officer.

Adrian Mendes
CEO, Perimeter Medical Imaging AI

Hello, everyone. Good morning. I haven't done anything yet. Thanks for coming. I'm Adrian Mendes, Chief Executive Officer of Perimeter Medical. We'll jump right into it. Just forward-looking statements, treat that information as you should. We've got a couple different products, as these are indications for use. All right. What does Perimeter Medical do? This is the problem we're solving. This idea that you go for cancer surgery, and the surgeon doesn't get it all, and then you have to come back in again. It's horrible for all the obvious reasons, cost, health, emotional, everything. That's what we're working on. Okay, so here are the highlights, and I apologize, I'm going to step away from this so I can kind of talk through it. I will try to project, but let me know if you can't hear me.

We have the first FDA-approved AI imaging device for cancer, for breast cancer surgery. This was just recently approved by the FDA a couple weeks ago. We're kind of in the stage now where, although we've been on market with our previous generation device, which had our imaging system, this now adds an AI layer on top of it, which helps it become much more usable for the surgeons as they look at it. That is now on market, and we're entering our commercialization phase with that. We had a large clinical trial. Step down?

Moderator

Yeah, just step down.

Adrian Mendes
CEO, Perimeter Medical Imaging AI

Yeah, the people in the back can't see me.

Moderator

Oh.

Adrian Mendes
CEO, Perimeter Medical Imaging AI

All that applause, I'm afraid, I got to put a show on. Yeah, for the FDA review, we did a large clinical trial, hundreds of patients, which showed great outcomes, and it improved surgical results from that trial, which is what the approval was backed upon. It's a large market opportunity. The largest number of cancer surgeries is breast cancer, around 300,000 a year just in the U.S., and millions if you look across the world. It's a very, very large market opportunity. Our business model is recurring revenue. There's a disposable that we sell. It's a relatively high gross margin business model. We've got about 70% gross margin on the model. Like I said, we're into the commercial launch phase right now.

We do, and I'll talk a little bit about this, we do have a very good technological moat around us, as well as regulatory moat now that we are FDA approved. This is a summary basically of what I will talk about in more detail in the presentation. If you want a one-pager, this is where everything's summarized too. A few sort of stats. With our previous generation device, we've had thousands of procedures. We're across a few dozen hospitals across the U.S. Very localized. We've got a very small sales team, so we've been out seeding the market with that first generation device. Now with AI is where we start to scale the business out. Although we're focused on breast cancer, the tech works across any types of cancer. The AI is specific to breast cancer.

The imaging system, which is proprietary and unique to us, that is applicable to any type of tissue. Okay, let's talk about that for a moment. If you're unfamiliar with cancer surgery, there's this concept of margins, and you don't want positive margins. Positive margins, counterintuitively, is bad. What the surgeons try and do is take the cancer, the tumor, out of the body, fully surrounded by about two millimeters, a few millimeters of healthy tissue. Think of like a hard-boiled egg, right? The yolk is cancer, and you want a couple millimeters of white all the way around the cancer. If not, chances are what the surgeon has done is cut through the tumor and left some of the cancer back in the body. That's how you have a successful or unsuccessful surgery. That's the goal, to have clean margins, to have negative margins.

Unfortunately, it's a really hard thing to do. You can see some of the reoperation rates there for different types of cancer. It's high, 23%. One woman in every four women or five women go for breast cancer surgery have to come back again because they left cancer in the body, and that's the problem we're trying to solve. It's not just breast cancer. Other types of cancer. Really, all cancers, it's a big, big problem. Okay, two things. If you think about Perimeter, there's two technology pools that we've got, IP pools we have in the company. This is the first one. Optical Coherence Tomography is the imaging system. Think about most simply just like ultrasound. Instead of using sound waves, OCT uses infrared light waves. Two benefits of that.

One, it's very high resolution, so you can see cellular-level detail, which is important when you're looking for cancer cells, number one. Number two, it has a couple millimeter penetration depth. You can look inside that entire depth of margin all the way through those millimeters to see if you see any cancer cells. Okay? That's the basis of the technology. That OCT imaging system is what is in our first product and is also the basis for the second product, which then takes that and then adds an AI overlay on top of it. We, over time, we're about 11 years old as a company. We've collected millions and millions of images of breast tissue, both healthy and cancerous. With that, we've obviously trained an algorithm.

You put these two things together, and it not only allows the surgeon to visualize something that they have never been able to visualize before, which is imperative to having a successful surgery, but also makes it easier to use because the AI assistant is guiding them at what to really spend their time looking at. Think about autonomous driving. Everyone can drive. Some people are not as good drivers as others, but if you have all the safety features in your autonomous car, it makes everyone a much better driver. The new product is called Claire. It's a cute name that someone came up with at our company, AI in the middle of it, Claire, clarity. It's a woman's name, so it's all nice. That's it.

It's basically the OCT system plus the AI overlay on top of it, and that's what the FDA approved a couple of weeks ago. These are some of, in more common language, some of the label claims we can now make. This is very important. With FDA approval versus what we had before, although those surgeons, the market presence we already have with the current device is primarily with breast surgeons, our label claims were such that we couldn't make claims around breast cancer or margin, reducing positive margins and things like that. Now that we ran the clinical trial and those outcomes were positive, we now can make those claims. That allows us to do a couple of things.

One, it allows us to communicate way more clearly to surgeons as well as hospital administrators and insurance companies what we do, in plain English, whereas we couldn't really do that before. The second thing, which I find the most exciting, is we can now market directly to patients with words that patients will understand. If you just think about this just intuitively as a human being, for a couple of thousand bucks, $2,000, $3,000, reducing the risk of your loved one having to come back for another surgery is. We'll start to get pressure back into the hospital system. We can do, from a go-to-market standpoint, through the surgeons, through the insurance companies for reducing the cost of a second operation, and then through the patients for all the better outcomes that a patient will want. This was our clinical trial.

We had a lot of the top cancer centers participate in that. Got good coverage. When we released that, it was successful. Okay, standard of care, top row across here. In the operating room, surgeon takes the tumor out, closes the patient up, sends the patient home, takes that tissue, sends it down to the pathology lab. Pathology looks at it under a microscope, takes a week or two to get an answer back to the surgeon, and then hopefully it's a good answer, but too often it's a bad answer, and then the surgeon has to call the patient and bring the patient back again. Okay? That's how it works right now. With our device, surgeon takes the tumor out, puts it in our machine. In about 10 minutes, you get the whole thing scanned, and you can see it on the screen.

If the surgeon sees something that looks suspicious, since the patient is still laying there, they can take a little bit more tissue and try to get those margins cleaned up. That extra tissue put on the machine, make sure they really are clean. Okay, fine. Surgeon's happy. Close the patient up, send that tissue down to pathology. They do their thing over the next week or two, and the answer comes back. With the previous product that we've had on the market, no AI, but at least the imaging system and trained surgeons, gives them the ability to actually look at those and interpret those images. The standard is 20%-25% reoperation rate. Our surgeons who use the machine are down around the 5% and less reoperation rate, even without the AI. The visualization system works. If you're well-trained, you can have very good results with it.

With AI, it allows us to really propagate this throughout many, many more surgeons. There's 8,000 surgeons in the U.S. that perform breast cancer surgeries. Okay. Like I discussed, it's a very big market. This technology, although we're starting with breast cancer surgery, can expand in multiple dimensions, can expand to other types of cancer surgeries. We have one surgeon who's using it, a thoracic surgeon who's using it for lung surgery right now. We've got surgeons who do breast cancer as well as skin cancer, so they use it for that. We've got studies on our website that show head and neck cancers, so tonsil and tongue and throat, and they show good results. We haven't got AI for those other types of surgeries, but it's a market for us, with the basic hardware, it's all going to be very the same or similar. The market expands.

One way is across different types of cancers. Another way of expansion is, of course, internationally. Another expansion dimension is through the entire care, right? We're just talking about surgery right now, but really for us, anything that examines fresh tissue is an opportunity for us. Biopsy as an example, right? There's a whole other market around biopsy that we think we can attack, both with the imaging system as well as the AI. We've got a lot of supporters out there in the surgical community, patients, patient advocacy groups. Now that we have the label claims around breast cancer, we can start to really push on the patient advocacy groups. A lot of really great endorsements. If you think about the different constituents, there's benefits to all of them.

The insurance company does not have to pay for reoperations, which can get quite expensive. The hospitals can provide better outcomes, which helps their scores, their satisfaction scores, and it also helps their top line. We see this even today pre-Claire, with the older product, is where surgeons and hospitals advertise that they use our technology to try to reduce their reoperation rates. What that does is it brings in more patients to their facilities. Every breast cancer patient that comes in on average brings about $100,000 of top-line revenue to a hospital between the mammograms and all the imaging and the chemo and the surgery, et cetera. It's a way that these hospitals attract patients into their system. As you know, normally, women are the healthcare decider of the family.

If they end up going to a hospital and they like it, they bring the rest of the family along to the hospital. The hospitals look at that and say it's a way to grow the top line. Patients, it's obvious why it's helpful. Then for the surgeons, these are surgeons. These are people who are at the top of the class, always have perfection. They want to do a great job. The worst thing in the world for them is, first of all, telling the patient that they have cancer. Secondly is telling the patient that, "I told you I was going to help take care of you, and I'm sorry I screwed up, and you need to come back again for another surgery." They hate that. It just affects them so deeply.

This is a big selling point for them as well, being able to get better at what they do. With respect to the patient, and I guess to everyone, there's a huge impact on patient outcomes if you can get clean margins on the first operation. The graph on the right-hand side. By the way, this presentation's on our website, so you can pull it down if you want it. What it shows is the five-year survival rate if you have successful first surgery versus if you don't. You can see that red bar drops for different types of cancers, sometimes drastically. For breast cancer, it drops about 10%, 10% or 11%, if you're not successful on the first surgery, but o ther types, head and neck cancer is the second bar.

Your survival rate's not that great if you have that in the first place, but it drops drastically from 55% down to 10% if they're not successful on that first surgery. It's a massive impact. Five years is not that long. It's a massive impact to outcomes. Let's talk about AI for a moment. The nice thing about AI is it gets better and better all the time for three reasons. One, the hardware keeps getting better. Nvidia keeps making more powerful GPUs. Two, the algorithms keep getting better as the researchers figure out how to make smarter algorithms. Three, as the database grows, you get to train your algorithms with better data. Our database continues to grow. For every procedure that we have patients go through, we get more and more images.

The trial we ran that completed at the end of 2024 was based off an algorithm that was created about two years before that. Create the algorithm, you have to freeze it, run your clinical trial, which for us took a few years, and then take a year through FDA. In that entire time, you have to continue to use that algorithm that you had way back then. That's what we have. It's the middle one there, Claire with AI 2.0. That's what we ran the trial in. That's what the FDA approved. One of the things that the FDA sort of has is something called a Predetermined Change Control Plan.

What that means is when you submit an application for FDA review, you can also attach to it a PCCP, that plan, which says, "Here are the changes that I have not made yet, but that I'm going to make. Please approve this now, so when I make the changes, I don't have to come back for a whole another review with you." They approved that for us across a number of things, but most importantly, for the AI. What that means is within bounds, we can improve the AI without having to go back and do another trial and do another review of six months or a year or whatever and release it.

Of course, our engineers in the lab are working on next generation AIs, and the 3.0 picture on the right-hand side is an example of the ways that we improve usability, we improve accuracy, what that means for the surgeons. We're going to be pushing out our next generation, not 3.0 picture , but an intermediary step out with better accuracy, in the next few weeks here already. We just got to market. We're already pushing out new AI. For us, it's very important because it's a new product. We're creating a brand new thing. We need surgeons to understand that it's like a generation one iPhone. It's good, but it has warts. As long as you see a trajectory of it getting better, you're willing to go along for the ride.

For us, this pace at which we can improve it is so important for our corporate strategy, and so we're able to do this with what the FDA gave us authorization to do. Okay. At the end of last year, here's some statistics of where we were at. We were selling our first-generation device. We closed last year at $2.3 million of revenue. A couple dozen hospitals throughout the U.S. Very localized markets, very small sales team, really focusing the company strategically on getting that FDA approval. Okay, that happened, now it's a different ballgame. Now we're off to the races selling Claire. This year, 2026, is a transitional year.

Moving our current customers from the older device to the newer device, getting that deployed, building more inventory, and then expanding our sales force, and then growing this within territories we're in already, but then even beyond that to other regions. I've got a map, which I'll show you. That's where we are in this year and sort of into next year. As we get into beyond that, is when we start to expand that TAM out of breast cancer surgery into other types of surgeries, start to build next generation AIs, and expand the business in that way. Here's the map of the U.S. The little red dots are where we have presence today, where we have customers, paying customers.

The number one thing that should jump out at you is some of the densest places in the U.S., we don't have anything right now. New York, Boston, Chicago, Cleveland, Great Lakes region, nothing up there. We have surgeons there that want our product that are saying, "Hey, listen, come work on my administration. Now that you've got Claire on market, we want you to come because I want to start using this." We don't have salespeople up there, and we're being very judicious about where we spend our salespeople time. Now that Claire's approved, hire the sales team, do a fundraise, hire the sales team, and then start to hit up those regions up there. Very big markets in those areas, obviously. Okay. The three pillars of our strategy is, one, drive utilization in places we already have the machine.

That's. I'll call it three different ways. One is get current surgeons using it for more and more of the procedures, and there's a learning curve that happens that we've seen that we continue to push. Two, get other surgeons in those hospitals use the machine that's already there. We have a per procedure revenue stream, so the more that they use the machine, the more money we make, so expand utilization of the machines already there through other surgeons. And then three, in that first bucket, is to move them over to Claire, which has a higher ASP on a per procedure basis. That'll increase our margin and our revenue. Okay, that's number one, drive utilization. Two is pursue system-wide sales.

We have contracts at system levels within hospital networks in the U.S. where we're in one or maybe two hospitals. Now with Claire, start to expand that out to the dozens of hospitals that one contract covers. Work needs to be done at the local level, but at least we're on the approved vendor list. We've got corporate sign-off. It makes it way easier for the OR directors to sign off on it at the local level. Then three, we've got this existing pipeline. I talked about that regions we're not in, there's a pipeline of customers there, is convert those customers or convert them into customers. Talking about our revenue model, the way it works is there's a cart, the thing you see on the left-hand side, which sits in the operating room.

We start the selling process by saying you have to buy the cart, and we fall back from that only when we run into barriers. If we sell the cart, we sell it for about $250,000, about 50% gross margin. If we fall back, we fall back in one of two ways. One is, "Okay, fine, your capital budget doesn't open for another six months. Let's do monthly payments until then." Or, "We'll place it there, we retain title, but you have to have a minimum volume commitment." Which moves to the second part, which is the consumable. This is a very high gross margin, one use per patient, and that's where we make most of our money, 90% gross margin on that product. We have service contracts, so for every machine, whether we sell it or place it, we charge $25,000 a year to service.

We've got the AI, which has an additional charge on top of it. These are our four revenue streams that we bring into the company. The opportunity, I kind of talked through a little bit this already, is about 300,000 breast surgeries a year in the U.S. Our goal is to make this standard of care where you have to use this because the risk of not using is just too great and too costly. You can see some. It doesn't take long for us to get to some pretty big numbers and the dollars that we're charging already. With a 70% gross margin falling to the bottom line or falling to the middle line, we get to a break-even point very, very quickly to cover all our overhead and R&D costs.

The top graph is just breast surgeries, and as we expand out beyond that, the numbers get even bigger. This is also just U.S., not globally. This is our financial model. We've got a large market which we're just starting to penetrate right now, recurring consumable revenue, both on the hardware as well as software, which is also very nice because there's a hardware component, too. We can track very precisely how our users are using a product, and if it starts to fall off, we can go and problem solve what's going on there. At this point, we're really targeting commercial expansion. Every dollar we bring in is being pushed towards that. This is the phase of the business we're in. High gross margin, highly leveraged, so break-even point for the business as a whole is not that far off.

In terms of insider ownership, we've got a lot. We've got, I think, around 15%, let's say, owned by management. This is our revenue. I talked about this a little bit. Nice growth curve over the course of the past few quarters. We've got patents around the imaging system, both in Europe and in the United States. We've got some patents that we've submitted for around the AI. We have the big library of images, which no one else has, and you can't replicate that unless you have the imaging system, and you can't make the imaging system without the patents. This is how this whole thing is protected. Even if you had the imaging system today, it takes you a while to create that database. It's very well protected. This is the leadership team.

The company was founded here in Toronto by Andrew Berkeley and a Co-Founder still at the company just up the street in Richmond. I'm here, the Chief Executive Officer. Sarah Bryan is in the back somewhere, our Chief Financial Officer. Dr. Ted James ran the breast surgery practice at Beth Israel Deaconess Medical Center, one of the Harvard schools, for a long time. He's now moved to Chicago and has taken over implementing breast cancer best practices at that network. Carl is our head of engineering here in Toronto, and Abbey Goodman is leading our sales efforts. She's got experience in medical devices, growing businesses from zero dollars to millions of dollars a year of revenue. It's a kind of a unique find.

She's unique in that manner of being able to do this type of business and ramp it from where we are now to where we need to be. This is our capital structure, TSXV listed. You can see the major shareholders. My name's up there. Social Capital is our largest shareholder. Rocco Chiarelli owns just under 10%, and he's one of the original investors back when the company was first funded through VC. He still is an active investor. We're followed by Leede Jones Gable, Paradigm Capital, and Raymond James. There's the summary. I'm not going to read all that because I have 46 seconds left, so maybe time for two questions. Maybe one. Yes.

Speaker 3

Sales required to break even.

Adrian Mendes
CEO, Perimeter Medical Imaging AI

I'm sorry?

Speaker 3

Sales required to break even. You're putting on all the expenses for new salespeople to get out there and land fish in the boat.

Adrian Mendes
CEO, Perimeter Medical Imaging AI

Yeah.

Speaker 3

Do you need 7,000 procedures to get it done? What's required in terms of number of procedures and/or dollars to break even?

Adrian Mendes
CEO, Perimeter Medical Imaging AI

We will get there probably in the next two years. It isn't a lot of procedures. There's a dependency on how quickly we want to spend on sales and growth, so we'll do that sort of trade-off. Right now, we're burning about $400,000-$500,000 a month net. At $1,500-$2,000 per procedure, it's not that many procedures to cover the difference. Yeah.

Speaker 3

How big is the footprint of the hardware?

Adrian Mendes
CEO, Perimeter Medical Imaging AI

It's about the size of this podium.

Speaker 3

This podium.

Adrian Mendes
CEO, Perimeter Medical Imaging AI

Yeah. It's a bit short. The part where you put the thing on, it's a bit shorter there, and then there's a screen.

Speaker 3

How does the consumable interact with the?

Adrian Mendes
CEO, Perimeter Medical Imaging AI

On the top of the cart, there's a window. The consumable is two parts. The bottom part clicks into that window. It's glass, like a photocopier. The tumor goes on top of that, and then there's a cover that goes over top of that.

Speaker 3

The tumor with the wire.

Adrian Mendes
CEO, Perimeter Medical Imaging AI

What happens? Yeah, the camera is underneath, pointing upwards, and it takes a picture of the part that's on the glass. Of the egg and the yolk exactly. You do rotate it, and you take a picture of the other side, and then you get all of that imaging on the screen.

That's right. You can't really copy. One of the questions, can somebody just copy the consumable and undercut your business? You can't really do that because of the way they interact together.

Speaker 3

Right.

Adrian Mendes
CEO, Perimeter Medical Imaging AI

Yeah.

Speaker 4

The S-Series got approved four years ago, and you have only 22 centers with 25 devices.

Adrian Mendes
CEO, Perimeter Medical Imaging AI

Yeah.

Speaker 4

How do you think the AI will help you to push more? Because you have to get

Adrian Mendes
CEO, Perimeter Medical Imaging AI

Two things. One is I was very limited in my ability to grow the sales team because we were spending so much on the trial and moving through FDA. Now those expenses are behind us. We restructured our costs towards the end of last year, took 30% out of that. It allows me now to have, from a smaller base now, redirect more of mine to expand the sales team, number one. Number two is the biggest barrier we had to sales was ease of use. I'm a surgeon. I'm not a radiologist. I have to look at this complicated thing. I don't want to do that. Make it easier. Fine, here's AI. Yeah.

Speaker 4

Can we see a few words about financing?

Adrian Mendes
CEO, Perimeter Medical Imaging AI

Yes. We're in the process of financing now. We just did a press release this morning, so you can look that up. Paradigm is the lead bank of that. Sorry. Maybe I shouldn't be here at Bloom Burton saying that. Yeah, we're raising right now. The goal is to achieve two goals over the next year. One is to get the commercial engine up and running, and the other is, we're on TSXV right now, to prepare ourselves for a NASDAQ listing. There's some accounting work that we need to do in the background for that.

Speaker 4

Yeah. Have you priced it?

Adrian Mendes
CEO, Perimeter Medical Imaging AI

We have prices in the press release. Yeah.

Speaker 4

Change the release.

Adrian Mendes
CEO, Perimeter Medical Imaging AI

I'm sorry.

Speaker 4

I didn't see it in the release. It said the terms of the warrants and so on, but.

Adrian Mendes
CEO, Perimeter Medical Imaging AI

Oh, okay. It's a unit deal, CAD 0.35.

Speaker 4

CAD 0.35.

Adrian Mendes
CEO, Perimeter Medical Imaging AI

It was in the press release, right?

Speaker 4

Yeah.

Adrian Mendes
CEO, Perimeter Medical Imaging AI

Yeah.

Speaker 4

It was there. $0.35 unit.

Adrian Mendes
CEO, Perimeter Medical Imaging AI

CAD 0.35 unit

Speaker 4

CAD 0.20 warrant, five years. Hope to raise CAD 5.6 million-CAD 7.5 million.

Adrian Mendes
CEO, Perimeter Medical Imaging AI

You got a job.

Speaker 4

It works.

Adrian Mendes
CEO, Perimeter Medical Imaging AI

All right. I think I'm being hooked off the stage now. We're around. Come find me. Sarah, myself, Stephen Kilmer, Diana Chan. We're all around. Thank you.

Powered by