PMCUFPMCUF
LOADING
|||
Operator: Thank you for standing by, and welcome to the Pro Medicus Limited Full Year Results Briefing. [Operator Instructions] I would now like to hand the conference over to Dr. Sam Hupert, CEO. Please go ahead.
Sam Aaron Hupert: Thank you. Good morning, everybody, and thanks for joining us for our FY '25 full year results. As most of you know, we are leaders in enterprise imaging and radiology information systems. We work in 3 jurisdictions. Our group headquarter in Melbourne, Australia. Our R&D center and support center in Berlin. And our largest market, which is about 90% of our revenue from North America with our offices in San Diego. We have 2 product lines in Australia. It's the Visage RIS, which does more of the practice management, back office billing scheduling and Visage 7, which is a clinical product. It's the radiologist desktop that they use to call up images, enhance them and make the diagnosis, and that's the product we sell globally and particularly in the U.S. In terms of our results, I think it's fair to say that all of our key financial metrics moved materially in the right direction, revenue up just under 32%. Profit after tax, approximately up 40%, underlying EBIT around 40% and our margins, which are more than any of our competitors by multiples in the industry went up from mid-72s to 74%. As a result, we increased our retained earnings by about 36%, and we've announced a fully franked dividend of $0.30 per share for the half, up 37.5%. In terms of the summary of the year, it was another record year by any measure. We won 7 contracts totaling $520 million at minimums. We also renewed 2 large contracts worth $130 million and we had upgrade for additional products with existing clients for a total of $39 million. So in all sales areas, new sales renewals and upgrades, we made material strides. We also did 7 implementations. We signed a research collaboration agreement with new UCSF that joins those from the Mayo and NYU. RSNA, which many of you know is our key conference that's in November, December, end of '24 was our busiest to date. We made significant progress with other ologies and AR, which I'll talk about a little bit later. And all of that, we think, forms a very strong base for continued growth in FY '26 and beyond. In terms of the highlights of the year in terms of sales, our first sale came very early in the year. It was a specialist Children's Hospital in Chicago. And then just after the AGM, some 5 months later, we had a slew of contracts that came in a short period. The largest being Trinity at $330 million. We think it's most probably the biggest contract let in the industry software only. Then Duly in December for $30 million, a key academic in U.K. health care for $33 million in January, then BayCare, a large IDN in Florida for $57 million. And then some more early on in March and April with LucidHealth, Iowa. As mentioned previously, we had 2 sales of additional products. We often get asked, will archive be sold back to additional clients. And here were 2 examples, NYU Langone taking both Visage 7 OpenArchive, and worklist. Duke taking Visage 7 OpenArchive. And as part of that process for both of them, we're transitioning them from on-premise to cloud. The 2 renewals I spoke about a little earlier, Mercy and a renewal of an Australian contract. And as we said, RSNA was busy, and we were very busy on the implementation front. In terms of revenue split, for those that are new to the graph, the sand color is recurring transaction revenue. The blue is recurring revenue normally around service contracts from the older style where we used to sell upfront licenses and service. The green is professional services, so we charge for implementation and training. Those costs are spread equally amortized equally across the life of the contract. So it's a 5-year contract. We take 1/5 of that revenue every year. And the yellow is the data migration as we're doing more and more full stack, more and more archive, including NYU and Duke. The archive migration component this year did kick up, and we expect that it will continue to grow as more and more new contracts come in, particularly those that have the archive as a component. So the model that most of you are familiar with, we believe it's highly scalable with a lot of operating leverage. We don't sell hardware. Cloud fees are netted off below the line, so not part of any of these revenue figures. We charge for our training and implementation and travel once the contract is signed. We have maintained a highly contained cost base. And as a result, our margins, with industry leading -- by a long shot, as I mentioned continued to grow in the full year. So the transaction model, again, we talk about a concept of a minimum. If we have a new opportunity, they tell us how many examinations they did in the prior 12 months. We usually get them to commit to, on average, around 80% of that volume, and that's what we call the minimum. So each of these contracts that we announced has material upside built into it even if they just do the same volume as they did the previous 12 months, which invariably they do. And we've noticed that our clients as a whole grow well above the industry average. So all that growth then gets captured in additional transaction volume. The forward revenue has increased materially with the recent sales. It's a 5-year window. It's moved up from 600 -- low $600 million to $948 million. Again, these are minimums. We see upside as client examinations continue to grow, and we see this very much as an annuity style revenue base. In terms of who we service, many of you would already know, we service pretty much all the markets, all the segments of the market. In the academic medical centers or we call them Tier 1 academics, we do 11 out of the 20 -- top 20. Now these are usually all academic centers because they're rated on clinical excellence and research. But I think we do at least 2, if not more multiples or higher multiples than our nearest competitor. The other area of the market which is the biggest area of the market is the IDNs. IDN stands for integrated delivery network. They are usually regional hospital systems. The bigger ones are often multistate. And we've seen that in some of the clients. Early days, we had Mercy and Sutter then a number of years ago gained Intermountain and MedStar, then another sway with Novant, Inova and Allina and recently, obviously, our biggest one to date, which is Trinity. So a very large section of the market, very important for us, and we have some material clients in that space. The next is the risk product. As mentioned, that's largely Australia focused. We do have some implementations in Canada. We have -- that was our core product prior to acquiring Visage in February 2009. We deal with the 2 largest groups in Australia in I-MED and what used to be Primary now called Lumis. We are seeing some increased growth as new practices start to form in this market. And you will have noticed that our Australian business grew just under 5% this year, largely through new additions around the risk for not only existing clients, but some newer smaller clients. Visage 7, that's a key product we sell in the U.S. We believe we are still #1 in the 3 core areas that make the product that's so important to the product. Speed, functionality, the capability of doing 3D, 4D advanced visualization as well as 2D in the one product. We believe we're unique in that capability and scalability because some of our clients are literally dealing with petabytes of data per year, and cloud is also a key part of that equation around scalability. In terms of what's driving the industry, it continues unabated. The data sets are just getting exponentially bigger. Breast imaging is usually the canary in the coal mine. In the past, 2D mammography was a few hundred megabytes each exam. We have now 3D, which is called digital breast tomosynthesis, where we're now talking about 3 gigabytes per exam and in high resolution, they can go as high as 6 gigabytes per exam. So the data sets are increasing massively. With all of the new modalities, in other words, all of the new acquisition devices. So what makes us different? We think that all of our competitors use a standard legacy technology of compress and send, where the files come from the scanner, they're compressed as much as they can be without losing any fidelity, and that file is sent to a workstation that is highly configured, unpacks the file and does all the enhancements and manipulation locally on that workstation. We, on the other hand, have a unique and proprietary streaming technology where the files come to a central spot. Today, that's cloud. All the 3D and advanced visualization is done in the near real time, and we actually stream the pixels. We don't actually move the file. And this causes a delta, particularly around speed that can sometimes be many orders of magnitude. So we are one second in most exams, whereas in some cases, if compression and technology was used, that could be many minutes as a comparison. So it is a huge differentiator, particularly important as organizations get larger and more distributed and particularly now that pretty much every radiologist wants some capability to work at home for part of the week, and they need the same functionality and speed at home as they do within the hospital. Just going through some of the clients. I talked about Trinity that came just after the AGM. It is one of the top 10 IDNs in North America. I think they have facilities in 26 states, just to give you an idea of their size. They have committed to us for 10 years for the full stack. It will be fully cloud deployed. We will do our first go-live in October this year. It will be phased over an 18-month period purely because of the regional spread and 2 of their smaller regions are getting new electronic health records, which they want completed before we complete those. But we believe the Trinity deal is one of the largest deals announced in our industry historically. The other market that we talked about in the past was the private market. There's usually partnerships of radiologists. A lot of them have sold fully or partially to consortiums or private equity. It had been dormant because there was a massive amount of M&A activity culminating up until 2023. But we are seeing movement in that market and landed 2 deals in the year, one in December at Duly Health and one with Lucid. So a total of $70 million just at minimum for those private. So this joins some of our other private clients like BayCare and CRL that we had previously announced. So I think the main thing about this is it opens up that segment of the market, but importantly, it also shows that our platform is well suited to a broad range of market segments and not confined just to 1 or 2. We did have 2 renewals. Again, I won't go into it too much. Our run rate for renewals is -- remains at 100%. Just post financial year-end and beginning of July, we renewed the last renewal for this year, which was FMOL, that not only did a renewal but took the archive and made a cloud transition at the same time. So our next renewal will be in calendar year '26, so no further renewals scheduled for this year. We also had the Visage 7 OpenArchive upgrades. As I mentioned, 2 key clients. Again, the material implementations, material revenue, $24 million from NYU and $15 million from Duke. That is additional to the existing contracts for Viewers. So that sits on top of the contracts we announced for those clients when they originally signed up with us years ago. And we believe that this is a logical upgrade path for many of the clients that currently are Viewer-only and on-prem. They're looking to cloud, and we're able to transition them seamlessly by putting the -- converting their archive from on-prem to Visage7 OA and cloud and then bringing the whole Viewer stack into cloud as we do for all our new implementations. So the solution we talk about when we look at what is our nirvana is one cloud instance with all of the branches of diagnostic imaging being serviced, so advanced visualization, 3D breast imaging, all in One Viewer. And we're now looking at the connection -- connectivity with AI and with all the other modalities and other ologies, which I'll talk about as we progress. The other strength we have, which clearly relies on the technology is the ability to implement. We believe it's unique to us. We're able to implement even the larger scale enterprises in 1/4 or 1/5 the time of industry norm. As an example, we completed the go-lives for Baylor within a 3-month window. And prior to Trinity, they were our biggest client. So it gives you an example of how we are -- can do this seamlessly and quickly, which we believe is a significant factor in reducing the barrier to change because previously, companies would take 2, 2.5 years to complete an implementation where we do it in a matter of weeks or months. So our ROI, we position ourselves as a premium product in the market. We believe we provide significant return on investment, multiples of what the clients spend with us, in 2 areas. One is infrastructure and unparalleled clinician efficiency. We quote between north of 25% radiologist improvement, which we've been able to log in a number of clients. But as important as that is, there's also the clinical capability because it is a clinical product, and we do believe we have moved the needle in terms of what a radiologist is capable of doing with our application as compared to others, and that changes the clinical outcome, which clearly is a very important part of this -- of our technology stack going forward. In terms of our growth strategy, it is multifactorial, obviously, expanding footprint, which we did more in this last 12 months in terms of new sales than we've done previously. We have got transaction growth from our existing clients, which is we think about 2 to 3x the industry average across the client base. New product offerings, which I'll talk about in a minute. And we're leveraging our R&D capability in terms of not only new product in other ologies, but also in AI and some of the research collaborations that we have with existing clients, including the new one with UCSF. The TAM, we've updated our figure from $650 million to $670 million based on growth rate of roughly 3%. In any 1 year, the industry grows between 2% and 3% organically. We believe we're able to address 100% of that TAM from a product perspective and about 85% from a commercial perspective. That is the tail of 15%. Some of the opportunities are most probably too small to address one-on-one, but that tail is shrinking as a lot of the smaller people are being merged into some of the larger health organizations in their area. Including a recent sale of UColorado, which came just after 30th of June, we estimate our current penetration is about 10% and growing. So whilst we have a material footprint, we still think we have a very large addressable runway. We often get asked about pipeline. Clearly, the pipeline is dynamic. We're standing up at the AGM in November. I think the pipeline was the biggest it's ever been. We had all of the sales that occurred within the next 3 to 4 months still in the pipeline at that time a week later. Clearly, Trinity transition from pipeline to contract and then all the other contracts. But I'm pleased to say that as a result of RSNA '24 and other opportunities, the pipeline has been continually refreshed. And I think the UColorado opportunity was a perfect example of that, that may have just come into the pipeline around then, but obviously came from pipeline to contract in early July. The product set has expanded over the years. Archive was the first -- I won't go into it too much other than to say it's an integral part of our strategy of moving people into cloud. So that strategy like with Duke and NYU and of course, all the full stack solutions, all cloud-based, all Visage 7 OpenArchive. So it's an integral part of our go-forward strategy. And like our view, it's a transaction-based model. So there's potential upside with every deal. The last of our 3 core products, workflow, we released a few years ago. Again, it's been very popular. We've been able to sell that back to NYU and to others like Yale, but also all of our new opportunities or the majority of them take the workflow as well as archive as part of the full stack solution, clearly increases our total contract value. And not only that makes the implementations easier, so we don't have to deal with third parties. Again, it is a transaction-based model with upside. One of our key strengths is cloud. We are we believe the only company that does 100% full cloud at scale. So there is no hardware on site anymore. There is no PACS hardware that all gets decommissioned, and it's all 100% in cloud, as you would expect. We have been able to show it's suitable for all size implementations from some of the smaller IDNs that we do all the way up to the groups and the likes of Baylor Scott & White, where we're talking about millions of exams and petabytes per year. I think the important thing about cloud is the market has shifted drastically towards cloud over the last 3 to 4 years and maybe a bit longer. For us, we are agnostic. In other words, we work well in all 3 cloud environments, AWS, Azure and Google GCP. So if our clients have an existing agreement with 1 of the 3, we enable them to use that agreement because they've made commitments that they want to use, and we allow them to do that without any delta between the 3. So we see that cloud and the ability to work in all 3 as a major advantage going forward. Just rounding out it in the last few minutes, we have talked in the past about one few of 4 modalities. We did announce about 2 RSNAs ago, our cardiology product, which is not only out in the market. We had our first implementation in the year, but we are able to not only work in the other ologies, but also reflective like photo and video because we see image becoming a much bigger part of the patient's electronic health care record. So we did talk about cardiology. I think the important thing that often gets missed, it is in the same code base as the Visage 7 product. It's part of Visage 7. It's not a separate product. It's not a separate platform. So it's -- we believe it's the first offering that is fully integrated into the one platform. It has native tools clearly that allow specialist cardiology measurements that are required, and it has integration into third party because particularly in cardiac ultrasound, there's a statutory need to transfer those -- that information to a health care body. So you need an interface that allows for that to occur over and above what you normally do in radiology. But cardiology, we think, is often, as I mentioned, off and running. Our first implementation has gone well. And with UColorado, it was a material sum that was involved in the cardiology going forward. Something new, most of you wouldn't have heard from us before, is digital pathology. Again, unlike others, it's not a separate platform. It's part of the same code base. We are starting to show it to prospective clients. It is a work in progress. We believe we will have it out in the market this calendar year. It is, again, fully cloud-based and one of the few. And the other thing that's very interesting is we've also interfaced it with Visage Ease VP for Vision Pro, which means that pathologists could potentially use Vision Pro to replace both screen and microscope, which we believe no one has done before. We've also had it recently validated at the IHE Connectathon, where it's an interoperability against standard Connectathon, and we believe we were the first to satisfy the digital pathology profile. AI, a few things that have transpired during the year, and we think position us well. We have codeveloped a breast cancer detection algorithm with NYU, which is commercialization, getting ready to commercialize pending FDA clearance. This will sit next to our breast density and provide a full breast AI suite. We have made an investment in Lucid for cardiac CT to help round out some of our cardiac offering and recently an investment in the Australian-based 4DX for lung AI. We've also announced an additional research collaboration agreement with UCSF, which is one of the top academic medical centers in the U.S. and we have a growing number of third-party AI integrations that we're bringing to market. Again, Visage Ease VP for Vision Pro, without going through all the details, we were the first to release a product for this -- we are seeing some real-world use cases emerging from our client base, particularly around interventional guidance for procedures. And as I said, we see some of this potentially moving into other fields as we release our pathology offering going forward. This gives you an idea of what it looks like through the goggles. There's a 3D volume that you can manipulate using your fingers and gestures as well as being able to see what looks like a screen so you can actually do the 2 concurrently. Finally, RSNA 2024, it was our biggest to date. This is a shot of the booth that we built for -- and this is -- yours truly in front with Clayton right next to me. We had about 52 people. Staff this year, we think will be even bigger purely to meet demand. So in summary, most successful year in our company's history by any measure. We've expanded our product portfolio full stack, and that's proven to be very popular. Our implementation and support capability continues. We're refining it and are able to do even the biggest ones even quicker than we were, say, 2 or 3 years ago. Cloud is a huge strategic advantage. Our North American footprint and pipeline, we believe, continue to grow strongly. Cardiology, as I said, first site was implemented very successfully now with UC Health Colorado to follow in this financial year. We will be releasing digital pathology towards the end of this calendar year. We think the efforts we've done in the AI will be able to leverage as AI becomes mainstream, and we are seeing increased use cases for Visage Ease on Apple Vision Pro. At that point, I thank you, and we look to take questions.
Operator: [Operator Instructions] Your first question today is a phone question from Garry Sherriff with Royal Bank of Canada.
Garry Sherriff: A question on pathology and a question on AI. The first one, the pathology product, certainly exciting. In terms of a market sizing perspective, we've done a little bit of digging on the size of the market. I just wanted to try and clarify it with you. In terms of volumes for those relevant areas of pathology like histopathology, it appears to be about 2/3 the size of the radiology market. I just wanted to clarify that, that was about right. And secondly, from a pricing perspective, how are you thinking about pathology exam pricing versus radiology exam pricing?
Sam Aaron Hupert: Yes. I think the pathology market is where the radiology market was about 20 years ago. Radiology, as you know, went from film to digital. The use case and the imperative for pathology to do so is slightly different. Because the bulk of pathology is not image. It's what we call biochemistry like someone's cholesterol or hemoglobin count. Having said that, the big drivers for digital pathology are around AI and the ability -- pathology is becoming very subspecialized much like radiology has, and you can't have a melanoma- specialized pathologist at every single point that you do it. So I think the ability to remotely read has become a little bit more acute in pathology than it was in the past. I think the market, we've sized it may be a bit less than what you think. But I bet to be honest, I don't think anyone really knows, and it also will depend on whether the organizations will scan their historical slides because of these glass slides that by all, they still have to keep. Mind you, small. There's 6 small glass slides on average per test. But I think a few things about pathology. We are seeing it as an emerging market. And we think our technology, particularly with cloud and with storage because pathology files are huge because they're in color and high def. Our platform is well suited. And as I said, I think it is an expanding market. So your -- the figures we have may be slightly smaller than yours, but I don't think anyone really knows because it's so nascent in terms of overall adoption, but we think that will speed up pretty quickly from here.
Garry Sherriff: And in terms of the pricing, how are you thinking about that relative to radiology?
Sam Aaron Hupert: Again, we haven't done that. We're doing some of that analysis. We do believe it will be a similar model. In other words, how many tests you actually do. And all the hospitals know roughly how many new tests that they do, and we believe it will be on a per test basis, but we haven't determined that pricing yet.
Garry Sherriff: Understood. Next question, just moving to AI. I mean, given the advances we're seeing in AI, ChatGPT-5, people are creating CRM systems in like 16 seconds. It's seeing a democratization, I guess, of software development and algorithm creation. As a consequence, I assume barriers to entry are starting to fall a little bit. How do you defend your existing technological leadership position when these barriers appear to be falling purely from a technological perspective?
Sam Aaron Hupert: Yes. So I suppose there are 2 things and a lot of what we talk about generative AI, we're seeing more of that on what we call the reporting side because as you know, radiologists dictates a clinical impression of the images. And so we're seeing that sort of ChatGPT generative AI more in that area. The AI that we're also looking in other things like breast cancer detection with NYU is clinical. And there, you have to go through quite a rigorous clinical validation process. You can't just do it and say, well, we tried it on 5 breast mammography exams and therefore, it works. There has to be quite a rigorous clinical process behind it. And that's actually not only time consuming, but you need willing partners, okay, the hospitals to do that. So I think there are 2 streams of AI in our space, generative AI, where if it's going to help and do conclusions for report, you don't need FDA and that's an aid and an efficiency. But I think the clinical part, there's still quite material barriers to entry and FDA clearance, et cetera, that you have to go through. So I don't think -- I mean, the way of learning has changed because of large language models, but the actual key parts of the process are the same.
Operator: Your next question comes from David Low with JPMorgan.
David A. Low: Sam, if we could just start with cardiology and the addressable market there. Just wondering how you think about the back book there? I mean how much opportunity is there from the existing list of customers? And is that the first quarter call? Or is cardiology pretty different? And if you could size it for us, that would be great.
Sam Aaron Hupert: Yes. So cardiology works with the same equipment as radiology. The one that was always deemed to be cardiology specific was ultrasound or cardiac echo, which is the first option that we put forward. We are seeing a lot -- a significant increase in cardiac CT as a screening test instead of angiography that it's replacing in a lot of places. Then there's cardiac MRI and cardiac nuclear. So with all of those together, we think depending on the organization, if they took all 4, it could be between 15% and 20% of radiology. There are far fewer tests done, but each test is more expensive in terms of the -- what you can charge, the industry charges more for cardiology because they're more involved. So if you get all of them between 15% and 20%, in New Colorado, what we did was the key piece of ultrasound, and that was north of 10%. So it is material. We do think we will be able to sell it back to existing clients, and we are putting it out in conjunction with full stack, it's something that new clients can take from the getgo.
David A. Low: Any reason we should think that the uptake would be different to the experience with radiology? I mean it's the opportunity -- that the starting period is I guess, gradual, you add 1 or 2 customers and then we see a speed up in the growth rates. I mean, is cardiology likely to be different in what you've experienced so far?
Sam Aaron Hupert: Not really. You have a slightly different audience, as I mentioned, with ultrasound because you're dealing with cardiologists where image and -- acquiring and diagnosing image is part of their day, whereas with the radiologist, it's 100% of their day. But having said that, once you build the base of referenceability, then clearly, the ability to sell after that gets easier. So we think the first step is getting it out, next step getting our first client. We think we've taken another step with this new sale. And again, we think that, yes, it should have a similar dynamic to what we did with radiology.
David A. Low: Just to clarify, when you say 15% to 20% of radiology, you mean by value or by volume? Value -- dollar value.
Operator: Your next question comes from Mathieu Chevrier with Citi.
Mathieu Chevrier: I was just wondering about the shape of the revenue stream for an OpenArchive upgrade compared to a new customer that goes with full stack. For example, we had this NYU add-on $24 million over 5 years. Is that evenly distributed over time? Or is that more front- loaded?
Clayton James Hatch: Yes. Mathieu, it's Clayton here. Because of the data migration that they need to go through, we can take a lot of that -- well, we take that revenue upfront. So we bill it on milestones. So we normally bill on a quarterly basis, but we take a percentage of completion of the data migration. That revenue is taken as we do it. And then the exam revenue is spread pretty evenly over the rest of the period. So for NYU's sake, that would be about -- because it was announced separately, 10% to 15% of that total contract value. For full stack clients, it will be a smaller percentage because there's more revenue in the mix, more exam revenue from all 3 components.
Sam Aaron Hupert: Yes. But the charging model is the same. Whether they take it upfront or whether they take it as an additional module like NYU and Duke.
Mathieu Chevrier: Yes. Yes. I was just more asking from a modeling point of view. And then I was just curious to get your views on -- I don't know if you've seen this, but RadNet has come up with a solution called TechLive that received FDA approval just a few days ago for remote scanning. I was just wondering what kind of opportunities and challenges that you think those kind of solutions could bring to Pro Medicus?
Sam Aaron Hupert: Well, it's not the area we work in. That's the image acquisition area. So from what I understand and read it means that a technologist can supervise a scanner that they're not next to. We don't see that as any threat whatsoever. And it's something more on the client side. So it wouldn't impact the number of scans that we would see or do or charge for. So it's neither here nor there. But it enables groups, and I would assume it enables groups and I would assume our clients as well if they use similar technology because a lot of the equipment manufacturers are trying to bring that out so that you have 10 Siemens scanners, you don't need 10 technologists. That really is more client side and really doesn't impact us at all.
Mathieu Chevrier: Do you think it could be a tailwind, though, to volume growth over time?
Sam Aaron Hupert: It could. It could mean that you need fewer people to do more scans, which could be good for us.
Operator: Your next question comes from Andrew Paine with CLSA.
Andrew Paine: Look, just one that we get all the time, just in terms of older legacy systems. Just would be good to know how long you think these can continue to service radiology and what percentage of the market you think remains on these older systems?
Sam Aaron Hupert: Yes. It is an interesting question. I think that in certain areas, that sort of technology is cracked. And the canary in the coal mine is always in breast imaging because that was always where the files were large and have now become just exponentially larger. So some groups are just struggling under the load. And the other thing that's created a much bigger load is breast imaging mammography used to be 2D. With the advent of breast tomosynthesis, the files can be up to 10x what they used to be in 2D because it's like a 3D. It's like a scan, like a CT, turn them into a scan at a certain -- through a certain level of anatomy. And the interesting thing is in the U.S., in particular, and the rest of the world will follow, tomosynthesis 3D has overtaken 2D in screening. So it's more than 50%. So you've seen a new technology become mainstream and those files becoming huge. Now having said that, it's the same with everything. In the old days, a CT scan had a few hundred slices. Now it can be 5,000, 10,000 slices. So we're seeing the data sets getting bigger and bigger and bigger. And particularly, when you start looking at remote, a lot of these systems are starting to, what I call, crack. And we think we're going to see more and more of it where they just don't really keep up. So breast imaging, the canary in the coal mine, in reality, we're seeing this particularly with remote reading, with other modalities as well.
Andrew Paine: Okay. That's great. Just the size of the market for these legacy systems is probably pretty hard to estimate, but have you got any idea there?
Sam Aaron Hupert: Sorry, we didn't hear you well, Andrew.
Andrew Paine: Sorry, just in terms of the size of the market for these legacy systems, like what percentage of market share do you think they still have?
Sam Aaron Hupert: 90% because we have 10%. That's probably the most straightforward answer. And bear in mind, some of them have, over the years, developed more sophistication around how they push the images around. But again, that's coming to its high-water mark as well. So we really don't believe anybody else has been able to do this on demand regardless of if you push the button and the file is there in full res, stop 1 second. We don't believe anybody else has been able to crack that nut except for us.
Andrew Paine: Yes. That's great. And then just as a follow-up to that, probably the second most asked question that we get is just around any steps why these legacy systems kind of update? Or why aren't they updating and trying to remain in the market given it's quite an attractive market still?
Sam Aaron Hupert: Yes. So to do it, and people have tried in the past but have stopped, you have to go right back to the beginning. You can't upgrade them, you have to bring out a totally new platform. And that's a multiyear job. Then you have to get it FDA-approved. So is it possible? Absolutely. Is it a simple 12-month process? No. And then the last thing is our product is proprietary. So Visage was built from the ground up by the team in Berlin, and it was built without using an existing toolkit. So you can't get a sort of Microsoft toolkit and get to 70% out of the box and then refine the rest. So those that have got existing systems, it's a really, really, really big job, a, to bring out a new system and then not only that, get it FDA- approved, get it hardened, get it scaled and then transition clients to it. But as far as we know, and I think we have a pretty good view over the market, no one's been able to do that.
Operator: Your next question comes from Josh Kannourakis with Barrenjoey.
Josh Charles Kannourakis: Sam and Clayton, a couple of questions from me. First, just on pathology. Do you have a bit of a read-through of how many players in the market are using specific cardiology plus pathology solutions versus those that are sort of inbuilt to broader enterprise agreements that would be using radiology plus pathology, et cetera?
Sam Aaron Hupert: Yes. So pathology is actually just starting to come of age. There's a very small percentage that have adopted it. And in the past, the reason was around it's a small part of the enterprise, but a very data-intensive part. And if you did it on-prem, the cost of data storage would just be too high. There are a number of pathology-only players that have tried to get into the market. most pathology is sold by our radiology competitors where they package someone else's or package or do a separate platform and sell it as part of that, like a Philips or a Fuji or et cetera. But in those cases, they're all separate platforms. It's not part of their main radiology platform. It's a totally different one. It's the same badge, 2 platforms. But I think there's only been a small number of pure pathology players, particularly in the U.S., there's been more in Europe, and that's where a lot of them have come from. One or 2 have just gone FDA. The thing with pathology, selling it in its own right, it's a bit like cardiology. There's a high cost of sale and the actual dollar value of the sale is nowhere near as big as radiology, certainly not at this stage. So the competitors are 2: The sort of standard ones that we deal with on the diagnostic imaging side, and we are seeing a small number of independent pathology players, but a lot of them are hitching their wagon not so much to the platform, but more to AI around it.
Josh Charles Kannourakis: Got it. No, that makes sense. And so just to round out, like we talked about cardiology being 15% to 20% of radiology in total. How are you sort of thinking today about pathology?
Sam Aaron Hupert: Yes. It's sort of one of the questions we got asked at the beginning. We haven't really sized it. It's very hard. We thought we needed to do it to get a complete single platform across everything. So that was one incentive. And we think that market will continue to grow. I don't think it will be as big as radiology, certainly not from what we're seeing, but it could be material if it goes from the 5% or 7% adoption to a more generalized adoption.
Josh Charles Kannourakis: Got it. And just a second question, just around the funding environment. Obviously, the Big Beautiful Bill has come out. There's underlying [indiscernible] there, a bit of a crunch on funding. But also, I think if you sort of look into the detail, there's quite a bit of a shift in reimbursement to the sort of outpatient clinics, which has been happening for a while, in the U.S. anyway. How are your clients reacting to that? And if you could give any sort of context around how you think you guys could be impacted, from a positive or negative perspective, in terms of demand for the platform?
Sam Aaron Hupert: Yes. I think earlier than that, I think the COVID crunch, like a lot of industries, a lot of hospitals bulked up and then now they're standing saying, "Our profitability was impacted. We need to sort of optimize." And so you can look at optimization, particularly in our area, through 2 prisms. One is just cut cost, buy something cheaper. The other, which I think makes more sense, is buy the most efficient product because that's where you'll get the most bang for buck. So obviously, there are times, as you know, that we don't win every single opportunity, pretty much 100% of the time, it's because of the perceived cost. I think that's actually the wrong way to go because you spend the money, you don't get the full result. But clearly, we seem to be winning those arguments more than we're losing them and hence, the reason for the additional sales. But yes, there is a feeling and maybe this Big Beautiful Bill will sort of heighten that. We haven't seen that, but there has been a feeling that hospitals, particularly post-COVID, became unprofitable and now they're sort of tightening their belts and optimizing and our area is one of them. And if they use us, then we believe we give them the result.
Operator: Your next question comes from Annabel Li with Goldman Sachs.
Annabel Li: I also just have 2 questions, please, and I might ask them one by one. So first one, just a follow-up on pathology. Are you able to comment on the uplift in R&D or sales head count that might be required and potentially, any forward indications from customers that you might have already received so far?
Sam Aaron Hupert: Yes. So the actual product is in our current R&D budget. So the R&D figures for the year included pathology. I think, as the product evolves, it's probably not so much in R&D, but we'll get more domain knowledge through a product specialist, someone that's lived and breathed pathology. It's similar to radiology, but it's not exactly the same. So I don't think you should expect significant cost increases around that opportunity. So it's all pretty much what we're releasing. It has been already incorporated in the existing R&D budget.
Clayton James Hatch: Yes. The huge advantage of obviously using the same code base for both cardiology and now digital pathology or within the Visage 7 platform, the development team is already in play. As Sam mentioned, as we get people onboard, similar to what we've done in radiology, we will have product specialists, but the core development team stays relatively the same.
Sam Aaron Hupert: Yes. And all the other things we've built around radiology, we look at, saying, "How can they apply to pathology?" and an interesting thing is Vision Pro. Now this is incredibly early days, and so it's a concept. But the difference between the radiologist and a pathologist looking at an image is the radiologist has an image and they move their head to see the whole image because it's on a big screen in front of them. In pathology, a pathologist looks down a microscope. So they actually move the slide under the microscope. They never move their field of vision. So the thing you can actually do with Vision Pro is exactly that, that the goggles can sort of almost be like this immersive microscope where with your hands, you move the image in front of your eyes. So we're trying to see what other assets we have in terms of not just the core platform, but Visages, Visages Pro, that will make the pathology offering even more differentiated. Now that won't require much R&D because we already have those technological assets as part of the core stack. So as Clayton said, being able to do it as the one platform not only has huge implications for us in terms of development, it also has massive implications for the clients because currently, sometimes they have to manage 50, 60 different back images to get what we do in one. So again, it's part of the R&D, but we're also looking to see what other assets we have that will make it applicable to pathology.
Annabel Li: Perfect. And just on contract renewals. I appreciate you won't have anything come up until next year, but just wondering if you've started these conversations and maybe what the initial response has been and your expectations that they will stay on Visage?
Sam Aaron Hupert: Yes, we have started conversations. Fair to say that they're pretty much in the same format as we've had all the previous ones. So you would expect -- particularly with the bigger ones, some of the bigger ones where they want to have the conversation a little earlier, but sometimes it's quicker, sometimes it's something that they talk about, put on the back burner and then, as time gets closer, it comes to the front. They each have their different cadence. But yes, we've sort of started with some of the ones more towards the front of the year, broaching that conversation, but it's pretty much what we've had, similar sort of conversations to what we've had in the past.
Operator: Your next question comes from John Hester with Bell Potter.
John Hester: A couple of questions for you. For Clayton, can you just run through the implementation schedule for the next 6 months, please, Clayton, starting with Trinity and working through Duly and so on.
Clayton James Hatch: Look, I don't have them right in front of me at the moment, John. So I can probably take that off-line. They haven't changed too much. As Sam mentioned, someone like Trinity starts in October and is spread over an 18-month period. Duly, I believe, is early next year and over different periods. But nothing has changed much from the contract announcements, but I can certainly take that off-line with you.
John Hester: Yes, no problem. I just want to come back to the investment you made or the debt facility you established with 4DX, a $10 million facility, Sam. Tell us what was your thinking process on that? And also, is that the extent of your commitment to this company? Or is there more to come, do you think?
Sam Aaron Hupert: So for full disclosure, as you know, I was on their Advisory Board prior to them listing. And after they listed, I am no longer. So I knew the company. I think this was purely an investment made through the Board. I abstained from voting on it because of my prior association. And I think some of the rationale around it was the new V/Q product that they're looking at, which is a replacement of an existing test to determine pulmonary embolus. We think it's more in our wheelhouse in terms of our capability that could be useful on our platform. So really, we looked at it or the company looked at it, PME looked at it, through 2 prisms and investment. And with the algorithm, could that be something we could potentially add to the platform as we would, let's say, an Elucid. So that's why we did it. I wouldn't read more into it than that. And again, I can't talk about future investments at this point. I think the whole concept was that one and leave it at that.
John Hester: Okay. So this is a potential add-on on technology, but you're sort of saying that you're open-minded as to whether there's sort of more investment to come?
Sam Aaron Hupert: I'm saying that we looked at the $10 million as the investment. We haven't really discussed more.
Operator: Your next question comes from Christine Trinh with Macquarie Bank.
Christine Trinh: Sam, Clayton, congratulations on a strong result. Just two for me. Firstly, on pathology, to piggyback off of Annabel's previous question. I think you mentioned you started kind of soft-launching pathology to clients. What's the initial reaction been? And I guess, do you think the pathologists are kind of ready for this technology? Or will it be a bit of a slower uptake?
Sam Aaron Hupert: The reaction has been good. It's early days. So let me preface it. But a lot of the issue around pathology with the large data sets, and that's our strong suit. So the fact that 2 things, one, we can handle them and store them and store them efficiently in cloud because, if you don't, it breaks the piggy bank because I think each test is about 6 to 8 gigabytes because they're in color. So being able to efficiently store it is an important part. And I think pathologists realize that. The other thing is the speed because it's a streaming platform and they have all the things that -- what makes it good for radiology makes it good for pathology. But it's early days. And as I said, most institutions are thinking about digital pathology, but haven't gone that fully down that path. So we're at the beginning of a new cycle, whereas in radiology, it's a replacement cycle because everyone went digital 20 years ago. So slightly different dynamics in the market. We think there's an advantage of having it all on one platform because it then becomes a licensing issue, not an infrastructure issue. It has all the attributes of streaming cloud, all the things that our diagnostic imaging product has. So we think it has a lot of advantages, but it's very early days. And we'll update the market as we progress further. Clearly, FDA is one of the steps. And as I said, we believe we'll be putting in for that this calendar year.
Christine Trinh: Perfect. And then just quickly on the U.S. defense contract. You're probably tired of answering this one, but have you got any updates for us there?
Clayton James Hatch: Military.
Sam Aaron Hupert: Military, yes, sorry. No, there's been a fair bit of interest and activity. The military hasn't actually done anything other than they're looking and they're continuing to investigate what's out there. We don't know again of anyone other than us that's got FedRAMP for cloud, but they may. It's one of those things. But Certainly, that has positioned us well in terms of further discussion. But military is government and the government -- as you know, in America, everything has been [ involved ] with the new President. And so you would expect it would be going slowly. But we are getting the engagement, what that means longer term, we'll just have to wait and see.
Operator: Your next question is a webcast question from Michael Porter with Economic Concepts. This reads, can you quantify cardiology, e.g., ejection fraction type growth in PME business, scope for penetration in China market profitability?
Sam Aaron Hupert: Yes. So cardiology, if you're looking at ejection fraction, that is the ultrasound portion. And as we said, that's somewhere around between 10% and 13%, 14% of radiology, depending on the organization and how much they do. Certainly, we're looking to sell it back and replace what are the traditional systems that are used because the traditional systems are used by the equipment vendors. They sell 20 or 25 ultrasounds to an institution and then cardiology, ultrasound package on top, which really just does that. It's not part of a broader package, which ours is, which you need to have. So that's that. In terms of China, we don't have any current aspirations for sales in China due to a whole lot of reasons. But could it be used technically? Absolutely, it can be used anywhere. It's purely because there are nuances around the Chinese market that we currently feel we should -- it's not part of our -- not a region that we're actively looking at, at this point.
Operator: Your next question comes from [ Daniel Stewart ], a private investor. This reads, does Pro Medicus plan to take the relationship with 4D Medical further than just an investment such as distributing 4D Medical's products?
Sam Aaron Hupert: As I mentioned just in answer to one of the previous phone questions, that's a possibility. The agreement we have with them and the investment opens the possibility for that. But we don't have -- at this point, it's a bit too early to have a reseller agreement. So it's possible, but there's nothing firmly on the table yet.
Operator: Your next question comes from Michael Porter with Economic Concepts and reads, has Mayo Clinic upgraded full package?
Sam Aaron Hupert: No. Mayo Clinic is one of the ones we're speaking to about renewal next year. Currently, the whole of Mayo uses an archive called Media that they codeveloped. So at this point, the answer to that is no. What happens in the future, clearly, we have to see what plays out.
Operator: Your next question comes from [ Neil Stomak ] with Far Eastern Group. This reads, can you explain the contract and its terms, e.g., USD 15 million minimum over 7 years is USD 15 million over 5, about USD 3 million per year or USD 15 million per year? When start paying 3 months later, pay upfront, e.g., on January 1 every year or arrears. Move to subscription base. For Duke, are you a U.S.- based company selling to Duke? There's no tariffs on U.S.-based software sales.
Clayton James Hatch: Well, I'll answer the last question first. There are no tariffs at this stage on software, but we do contract through our U.S.-based customer. So for any of the opportunities, that's what we do. In terms of the contract, yes, $15 million deal over 5 years is $3 million per year at minimum. So it's at a minimum number of exam volumes. There's always inherent upside to that contract on the actual exam numbers that they do. In terms of payment, we have some professional services that we bill upfront. and pay that over time. But they get paid and spread over the length of the contract, and then we bill on a quarterly basis in arrears and get paid for it on a quarterly basis. So it's already pretty much a subscription model based on a minimum number of examinations that get topped up.
Operator: Your next question is from [ Stella Wang ], a private investor. This reads, regarding the next major renewals, when is Mayo first signed July 2016 due to be renewed? If I remember right, they only previously signed for Viewer. What's the profitability of Mayo taking up the full -- probability of Mayo taking up the full stack?
Sam Aaron Hupert: Yes. So you're right, it's Viewer only. And I just mentioned with the answer to a previous question, we are looking to renew them. The first part is around the Viewer clearly because that's the contract they have with us. In terms of other applications, often, we don't do those deals at time of renewal because the organization is just looking to get the renewal out of the way. FMOLHS, our last renewal was an exception to that because not only did they renew, they took the archive. And again, with Mayo, we can't talk about the future. But clearly, we are making them aware of all the applications we have, including archive, worklist, cardiology and now we're also making them aware of our pending pathology offering.
Operator: Your next question comes from [ Neil Stomak ] from Far Eastern Group. This reads, can you discuss your competitive landscape, please? Your financial metrics are so strong. Why doesn't GE, Siemens, Google, Tempus AI, et cetera, work to buy you or compete aggressively against you? Aren't competitors enhancing AI research collaboration? How soon before AI and cardiology reach inflection points estimated 2027, 2028?
Sam Aaron Hupert: So in terms of competitive landscape, you're right, we compete against a broad range of competitors, and they all compete aggressively against us. So whether they're Siemens or whether they're a cloud-based start-up, we see them all as competition. In terms of acquisition, I can't speak for them, but no one has offered to buy us, and we try and run our own race. Everybody is looking to enhance their AI capability. I think our strategy is multipronged. We think that the platform itself is the best platform for AI regardless of who actually creates the algorithms. So it's not just developing algorithms or partnering to create algorithms. It's the platform itself. And in terms of inflection points for AI and cardiology, we think we've taken, as I mentioned before, some big steps in cardiology. Colorado will be the next one. AI, I think it's still an emerging nascent market. I believe you'll see more and more pockets emerge in the coming year or 2. But is it going to be an avalanche? I don't think so, but those pockets will coalesce. So I think both markets are emerging, and we have keen interest in both of them.
Operator: Your next question comes from [ Ian Lee ], an investor. It reads, Sam, what percentage of your pipeline is now cardiology?
Sam Aaron Hupert: So there are 2 parts to it. Cardiology that we're looking to sell back to existing clients, so clearly, we're putting our best foot forward, showing clients what exists, trying to get in front of their cardiologists because, as I mentioned, it's a different audience to the radiologists. So we're obviously doing that. And in a number of RFPs, we're starting to see the reemergence of cardiology and radiology together. And the fact that we have a really good offering for cardiology positions us well. And I think UC Health Colorado is a perfect example of that. So yes, we're looking at it through both existing clients and clearly new RFPs.
Operator: Your next question comes from [ Tony Tan ], a private investor. Would you be able to name 1 to 2 closest competitors to Visage in terms of cloud and AI implementation? And what is PME doing to maintain the moat?
Sam Aaron Hupert: Yes. In terms of cloud, I'll be honest, you go to RSNA and everybody has those exact 2 words on every single stand, cloud and AI. Reality is we don't believe that any of our major competitors have a truly 100% cloud-based offering. They talk about hybrids where you have to have a system on site. And to me, that's non-cloud cloud. I mean, totally defeats the purpose. So as far as we know, whilst there's a lot of marketing hype and noise around it, we believe we're the only ones that can do true cloud at scale. And as evidence of that, we haven't put an on-prem implementation in the U.S. in the last 4 years. And all of the small to biggest have been cloud. So again, AI is something similar that everybody says they have AI in various bits and pieces, but we don't know anyone that's taken the march on AI and won the market by any means quite the opposite.
Operator: Your next question comes from Claude Walker with A Rich Life. This reads, congratulations on the multiple big wins this year. I just wanted to check, has the company lost any of the pipeline contracts to competitors in the last year?
Sam Aaron Hupert: Look, we will never have a 100% hit rate. I wish we did. That would be unbelievable, bearing in mind that we think we outsell most of our competitors put together. The ones that sometimes we lose them early on, they just look at price and go, "You've got to be kidding." Having said that, it's not unheard of for those people to come back a few years later. And sometimes people just look at something and they feel that their budget is x. So there'll always be some that were either excluded from fairly early on or people looked at price. Sometimes there's the concept that maybe the opposition have a broad range of product, which I don't believe is the case. So we're not going to have 100%. But clearly, our recent hit rate has been the highest we've ever had.
Operator: Your next question comes from Sinclair Currie with MA, Moelis Australia. This reads, do you have any insights about timing of when customers' contracts with alternative solution providers roll? And can it provide any information about timing of pipeline conversions?
Sam Aaron Hupert: We have some. It's very hard because some of the systems that have other vendors will sometimes just roll over a contract for a year. We tend not to do that. Our contracts and renewals, as you know, are longer term. So we have some insight, but it's not like a clear crystal ball. I don't think anybody has that. And we try and remember that and certainly, RSNA try and encourage those people to come and see us, but it's not as easy to ascertain as you might think from the outside.
Operator: Your next question comes from [ Neil Stomak ] with Far Eastern Group. This reads what are your efforts to build sales and scale in Germany and Asia, 3 to 5 years later, large? Are the TAMs for Germany and growing Asia, respectively, large relative to the U.S.? What are your obstacles, sales technicians, regulatory approval, competitors largely from North America or Europe?
Sam Aaron Hupert: So there are really 2 markets we look at as blocks, you could think of them as 3 now, so Europe, U.K. and Asia as a block, excluding China and, at this point, Japan because Japan has its own regulatory cycle and very different dynamics to maybe places like Korea and Taiwan. We have said in the past, and we still say it, U.S. is the biggest market. I think about 60% of health care spend globally is spent in the U.S. That's how big it is. Having said that, we have had some success in Europe, particularly in Germany, which is our base. And we're looking at the possibility of extending our capability there. It's not a technical thing. It product will work anywhere. It's purely market dynamics. Having said that, cost of sale is much higher in Europe. Each opportunity is smaller. Each opportunity is pretty much government run. So a lot more bureaucracy, a lot bigger tenders. So the bang for buck in terms of return is not there as it is in the U.S. But we are seeing slowly a change, particularly as cloud vendors come into the mix because currently, there's pretty much 0 cloud in health care in Europe, and that may help us. So our message, I think, we maintain U.S. first. That's where the biggest opportunity is. We've got a material foothold of 10%, a huge runway, and that's where we see the least impedance to doing business. But clearly, as things change, particularly with cloud, we can see the other markets, Europe and Asia, possibly opening up in the future.
Operator: Next question comes from Abhishek Jain with Citi. This reads, what level of OpEx growth should we be expecting for FY '26 versus FY '25?
Clayton James Hatch: Yes. I think it will be a similar level of growth in our operating costs, mainly through bringing on new head count, which will help drive that growth for newer opportunities. So it will be in similar areas that we've done in the past, both in sales, implementation, project management, development and all the key areas that we have in terms of head count. So it should be on a similar level that we've done in '25.
Operator: Your next question comes from Jacob Garfield with Avidity Partners. This reads, what is the rate limiting factor on signing more new customers?
Sam Aaron Hupert: There's not any particular rate limiting factor. It's purely in terms of what RFPs come to market. It's not that we don't have capacity or other bits and pieces, it's purely the number. And it's purely, as I said before, putting our best foot forward and showing the value that we have rather than just competing on price, which we don't do. So there are no particular limiting factors at this point.
Operator: The next question comes from [ Matthew Nicolatis ], a private investor. This reads, are you seeing any delays in implementations due to client resourcing or macro factors?
Clayton James Hatch: No, not at this stage. I mentioned, I won't go through them individually, but we haven't pushed out any of the implementations from a client level or macro level. They are pretty much on -- they are on track from what we've announced previously in the contract announcements. So all ahead in terms of implementations. We do have a number of them, but we do have a schedule that will see them coming through.
Operator: Your next question comes from Jacob Garfield with Avidity Partners. This reads, which cloud-based start-ups are you seeing compete, New Lanton, Sirona, Synthesis?
Sam Aaron Hupert: Well, every start-up, and anyone in the area we see as competition and we don't take any of them lightly. We have not seen those particular companies in major RFPs that we're currently tendering for. That's not to say that maybe some of the smaller organizations, really small ones would look at them rather than ourselves or any of our key competitors. But we haven't seen them in the mainstream of what we're currently doing, certainly not up to this point.
Operator: Thank you. There are no further questions at this time. I'll hand the conference back to Dr. Sam Hupert.
Sam Aaron Hupert: Just wanted to say thank you, everybody, for joining us and for your questions. Appreciate it very much. Thank you.
Clayton James Hatch: Thanks.
Operator: That does conclude our conference for today. Thank you for participating. You may now disconnect.