Good morning, everyone, and welcome again to the 44th Annual J.P. Morgan Healthcare Conference. My name is Matt McEwan, and I'm an associate here at J.P. Morgan. It's my pleasure to introduce our next presenting company, RadNet. Joining us today will be CEO Dr. Howard Berger, CFO Mark Stolper, Chief Strategy Officer Dr. Greg Sorensen, and CEO of RadNet Digital, Kees Wesdorp. They'll be running us through a brief set of materials, and we'll ask you to hold off on any Q&A until the end. With that, I'll hand it over to you guys.
Thank you, Matt, and thank you for joining us this morning. My conversation and talk will be very brief, so I can let the people who really are help driving the company get up here and give you the bigger vision. But fundamentally, what we're talking about is healthcare is in transition. It's in transition, really, in two ways. One is the enormous challenges that healthcare is facing due to the labor pool shortage. It is affecting virtually every facet of what has been the traditional way that healthcare has delivered our services, and that is with manual efforts.
This began perhaps four or five years ago with COVID and has continued not only despite the COVID being somewhat now put in our rearview mirror, but because it has changed the fundamental outlook of how people view working in the healthcare sector, and the shortage has led to an increasingly higher cost of delivering that service. The other facets that are driving this change are the continued effort on the part of both referring physicians, patients, and commercial payers to drive as much out of the hospital systems and into ambulatory care centers as possible, given the enormous difference in both cost to the patient as well as cost to the system of delivering those services in traditional hospital settings.
The last, and perhaps the most important, is how healthcare needs to change, which is nothing new, but which the time is right to do that, is from what we've called reactive detection, meaning somebody comes in with a concern or a suspicion or a complaint, to proactive prevention, where we're trying to diagnose earlier and earlier what are the issues that the physician and the patient need to work together on to try to reduce the long-term consequences of diseases which, for the most part, creep slowly but cost the patients and the system a lot of money. And driving in on the white horse to save everybody is artificial intelligence. And necessity is the motherhood of invention.
You've probably heard that expression before, but I don't think it's ever been truer in healthcare that the fundamental problems and the changes that we need to embrace are those which artificial intelligence is capable of doing. I'm very proud to say that I think radiology and imaging are the poster child for this because we've been navigating in the digital world, really, for the last 20 years as both some of the newer technologies and now some of the older technologies have embraced digitization, so everything that is done in radiology and imaging is reduced to a bit or a byte or the viewing of a pixel in order to create an impression of what might be there or what you might not see and is there.
A couple of examples of that that we're excited about that I'm not certain that the team will get into now is that for those of you that have been following RadNet, three years ago we introduced something called EBCD, early breast cancer detection, and our ability to diagnose breast cancer maybe one to two years earlier than the best of radiologists who review these breast images has been quite a success for us, both in terms of the adoption by our patients getting screening mammography as well as the improved cancer detection. This year, we're going to be adding to that program a new tool that will allow risk assessment by looking at normal mammograms and being able to predict the likelihood of a woman within the next five years getting cancer.
So this is well before even artificial intelligence can tell you whether or not an abnormality is cancerous or not. This is predicting with a high degree of accuracy whether or not somebody should be more diligent, vigilant about managing their healthcare through screening tools. And that, in fact, is what we're trying to do on almost every facet of whether it's cancer, whether it's cardiovascular disease, metabolic diseases like diabetes, all things that can be looked at with information that has always been there in our scans, but which there never has been tools to look at at a level of detail that the human eye just can't do. So both on the clinical side, which is what I described, as well as the patient experience side of how we manage our patients through our centers using artificial intelligence-guided tools to make it a less stressful experience.
I will tell you from personal experience as well as talking to many of our patients who come through our centers, nobody wants to be in a doctor's office, and so for us to provide tools that will make that experience less stressful and give better results is not only a challenge that we feel up to, but something that we think that we will be leading, and the new tagline for the company, I think, is appropriate, and that is advancing healthcare through innovation and artificial intelligence, so that's what you're going to hear about from our thought leaders here, and I'll leave you with the one thing that makes me feel that we're on the right track is that artificial intelligence, at least in healthcare, is not a bubble. It's here. It's here today, and it can make a difference.
And I'm proud of our team that's leading this for our industry in a rather unique way because while artificial intelligence is a major effort on the part of RadNet, so is continuing to expand our services division, which allows us to deploy the artificial intelligence in real-world time to assess the effectiveness of these tools and make adjustments, which is always going to be the case with artificial intelligence. The word artificial, unfortunately, I think, is not as well understood or used because there's nothing really artificial about it. Artificial intelligence, or AI, is best when you keep feeding it data.
When you broaden what these large language model algorithms are capable of doing, you need to continue to assess the value of what your models do to make certain that you're not only improving the level of quality that you're using the artificial intelligence, but also that you make certain that the habits that can develop by our employees, both on the physician and non-physician side, aren't taken for granted and it just becomes routine. So this is kind of a backdrop of the challenge that we have and the opportunity to make a real difference in healthcare. With that, I'll turn it over to Mark Stolper, our CFO.
Thank you, Howard. I'm going to briefly, for those of you who are a little less familiar with the RadNet story, take you through an introduction of who we are and what we do, and then turn the presentation over to Kees, who's going to talk a little bit about the digital health division. So we are the largest owner-operator of diagnostic imaging centers in the United States. We're not exactly an overnight success. The company was founded in the early 1980s. We divide the company into two operating and reportable financial segments, the first being our core business of owning and operating imaging centers, which is our imaging services division. It's about a $1.9 billion annual business and growing, and we'll talk about the growth a little later on in the presentation.
We also have a newer and fast-growing digital health platform, which ties into some of the things that Dr. Berger was mentioning around artificial intelligence and driving efficiencies in the workflow, both for our company employees as well as our radiologists. We've been a fast-growing business. We are projecting to grow the company on the imaging services side in the double digits, 11%-13% over the next several years on a compound annual growth rate. The digital health division is growing over 30%, and we anticipate that continuing in the next few years. We have 11,000 employees. We're based in Los Angeles. Our team members are across on the imaging services side across eight states where we operate, our fixed-site diagnostic imaging centers, and we have over 400 team members on the digital health side of our business, which are throughout the world.
So on the core business, the imaging segment, we have 407 locations, highly concentrated in eight locations, and I'll show you a map in a few slides. There's no accident as to why we operate where we do. We'll get into that. We have core operating tenets of being densely focused in densely populated geographies. We have clustered centers approach, which gives us a tremendous amount of efficiency and the ability to use our scale to be a low-cost operator.
We also have a core operating tenet of being multimodality, meaning the vast majority of our centers are what we call large centers of excellence that practice all of the different procedures within diagnostic imaging from routine studies, which are the X-ray, ultrasound, and mammography, which is roughly about 72% of what we do, as well as the more advanced studies, which are MRI, CT, and PET CT, which is the other roughly 22% of our procedure mix. We've been a fairly aggressive acquirer. You'll see in a couple of slides, our industry is highly fragmented, and as a matter of course in our business, we do look for acquisitions of smaller operators.
We also have been expanding very quickly in partnership with some of the larger health systems in the United States, where we partner with them and become their outpatient imaging partner, and we'll get into reasons behind that later on. With regards to the digital health segment, or what we call DeepHealth, which all of our products and services are branded under that brand, we're a global leader in providing radiology workflow solutions as well as clinical AI tools. So the products that Kees is going to get into today of RIS and PACS, a radiology information system and the back-end image management solution, is the backbone that RadNet's been operating on since 2009. We've taken that platform to the next level, to the cloud.
We've got over 2,000 customers as customers of both these workflow solutions as well as these clinical AI solutions that we've been developing in specialty areas that allow radiologists to become more productive and more accurate in detecting and diagnosing disease. We've been a consistent grower and had consistent results. We've grown the top line over the last 15 years at over 8.5% compound annual growth rate. EBITDA has tracked that. We continually go back to payers for pricing increases as we can prove more and more value to the healthcare delivery system and prove that we're the partner in helping them drive a lot of this outpatient volume away from hospitals into the lower-cost sites of care, and we'll get into that later on. Let me talk a little bit about the market. If you believe the research out there, it's a big and growing market.
Nationwide, on an annual revenue basis, the diagnostic imaging market is about $140 billion. It's divided between work that gets done within hospitals and work that gets done outside of hospitals. It's believed that the hospitals today perform about 40%-50% of all the diagnostic imaging in the market excuse me, 50%-60% of all the diagnostic imaging in the market. And that includes both work on inpatients, meaning the very, very sick or injured patients that are already admitted to the health systems, and those that are brought in through the trauma centers and emergency rooms, plus outpatient ambulatory patients that are referred into the hospital for diagnostic imaging.
Then the other side of the market, the market that we play in, is the ambulatory outpatient side where we work with thousands of referring physicians from primary care doctors all the way through every specialty who send us their ambulatory outpatient work. One of the themes that you'll hear today is the movement of, and Dr. Berger mentioned this in his opening remarks, the movement of diagnostic imaging out of the hospitals into lower-cost sites of care. In our specialty, meaning radiology, the hospitals charge typically anywhere between 200% and 500% of the pricing, both to the patient and to the insurance company or the health plan, as compared to the ambulatory outpatient market.
So the payers in recent years have been trying to direct more and more of the ambulatory outpatient work into the freestanding lower-cost sites of care, and they're doing it through a number of mechanisms, one being pre-authorization, where most advanced imaging today must be pre-authorized by the insurance company. And at that time, they try to direct the patient to the lower-cost outpatient facilities like the ones we run. And then secondly, they're changing plan design to provide financial incentives with lower co-payments and co-insurance for the patient that incentivizes the patient to self-direct into the lower-cost sites of care. Our industry is highly fragmented on the imaging services side. It's believed to be almost somewhere between 6,000 and 7,000 imaging centers out there. We, plus the next four players, represent less than 15% of those outpatient centers.
So this is one of the last frontiers of healthcare services that hasn't consolidated in a meaningful way. And there's tremendous opportunity for us to continue to grow the business through consolidation and through de novo centers. So this is where we operate. We operate in eight states. We're highly concentrated. There's two primary benefits from being geographically concentrated, the first being to be able to operate at scale and to be able to centralize many of the functions that we perform on behalf of our imaging centers, such as revenue cycle, pre-authorization, insurance verification, marketing, et cetera. We're also able to load balance our capacity by sending patients to various sites that have capacity at the time that they're looking for their appointment.
Secondly, the big advantage of being geographically concentrated is that we have become indispensable to the healthcare delivery systems of our markets, meaning that the payers need to have us in network with them, and it gives us a seat at the table because we're their primary partner in helping drive this business out of the more expensive hospitals into the lower-cost sites of care. So this gives us a seat at the table from a contracting standpoint and allows us to establish fair and equitable pricing for our services and even get pricing increases throughout time. So what do we do? Well, as I mentioned, we're a multimodality company. The vast majority from a procedure volume of what we do is simple routine studies, X-ray, ultrasound, mammography that represents about 72% of all of our procedure volumes.
However, the more advanced studies, which are growing much more quickly, the MRIs, CTs, and PET CTs, which represents roughly the other 22% of what we do, that drives about over 60% of our revenue. And what we're seeing in our business is that there is a flight towards higher acuity exams, and that's being driven by technology advances in advanced imaging equipment. It's happening with more advanced post-processing software on MRIs and CTs and PET CTs, advances in contrast materials, radioactive pharmaceuticals, and AI that's making the advanced imaging modalities grow almost twice as fast as the routine imaging. And from a business standpoint, they come with higher pricing and higher margins, and we think that that trend continues. From a payer mix standpoint, we have a diversified payer mix, so we're not exposed to any one payer.
About 23% of what we do by revenue is Medicare business. We have a special book of business that we're very proud of, predominantly in California, where we're at full risk for about 1.6-1.7 million lives under capitated arrangements where we get paid a per member per month fee for being the exclusive imaging provider for that patient population. That's a growing business, and as the healthcare delivery system focuses on risk-taking and value-based care, we think that that can be a bigger aspect of the company in the future. 60% of what we do by revenue is based upon negotiations with commercial insurance, both regional as well as the large national insurance. And then we have some smaller books of business like personal injury, workers' comp, and very little Medicaid exposure.
I mentioned in my opening remarks a growing part of our business is hospital joint ventures. Today, we have 26 joint ventures with some of the largest health systems in the United States, where they have partnered with us to be their outpatient diagnostic imaging partner. As these hospitals and health systems continue to lose business to the freestanding outpatient industry, many of them are looking for a strategy to participate in that trend instead of trying to fight that trend. And by partnering with us and being an equity partner in our outpatient imaging centers, they're enjoying the growth along with us of the outpatient industry. From our standpoint, we receive a couple of benefits. The hospitals have been integral in sending us additional procedure volume using their relationships with community-based physicians that we otherwise wouldn't see ourselves.
And they also give us a bigger seat at the table when contracting with some of the larger health systems. So today, 152 of our 407 locations are held within these health system joint ventures, representing about 37% of our centers. And we think that that can grow to over half of our centers in the next five years. I'm going to turn it over to Kees to talk a little bit about our digital health segment. Thank you.
Thank you, Mark. Great to have the opportunity to give you an overview of RadNet and the value we create with our digital health segment. We are a global leader in AI-powered informatics, and in the conversations this week, but also prior, where I always start is to say and to remind people that we already have 2,000-plus external customers. We obviously have RadNet as an internal customer, but many, many customers across the world as well. We have a global footprint with over 400-plus employees across four continents. We are clinically validated, and I'm proud to say we have the most comprehensive portfolio in the industry with now 22 FDA-cleared products and 15 CE-marked solutions. And we're integrating capabilities. The journey started originally with Dr. Sorensen, his company, DeepHealth in the mammography space.
But over time, we've acquired strategically assets and capabilities and teams, most recently last year with CMOT, ICAT, and CMAR, both in the AI domain as well as in the informatics and image exchange domain. When we say our digital health segment, that's equivalent to how we market ourselves externally with DeepHealth. We operate in an incredibly exciting market. And Dr. Berger already talked about everything that's going on, but this market is already vast, $5.1 billion in 2024, growing double-digit with 11% towards the $7.7 billion market across radiology, informatics, and clinical AI. Our solutions are positioned within that in the highest-growing segments of outpatient imaging, cloud-native solutions, and clinical AI. And that's why we have such tremendous tailwinds, both with our deployment at RadNet, but also with external customers as we scale our solutions. The need for these solutions, AI-powered solutions, is high.
We need to navigate clinical, financial, and operational challenges in the workflow, in health systems, in outpatient imaging. And so to name a few, today, there's a significant challenge of no-shows, so disconnected patient engagement. Up to 15%-30% of patients do not show up, which obviously has a massive burden on the execution in center operations. Strained workforce. You often have heard about burnout and staff shortages. That's only to expect to further widen. And so by 2030, there will be a shortage of up to 30% of radiologists, for instance. And then inconsistent clinical outcomes with quite significant double-digit percentage variability in what one radiologist would say in terms of interpreting an image versus the other. Within that, a highly fragmented tech data and workflow setup.
And so, for instance, any CIO in any health system or outpatient setting is dealing with over 20-plus IT vendors to make the system work. So there are too many point solutions. It's too fragmented, and it needs strong interoperability solutions to address those challenges. That yields quite a significant cost inefficiency that externally is estimated to up to $25 billion in radiology in the U.S. alone. And these challenges are obviously our opportunity. To that end, over the last couple of years, we've invested deeply in what we call the DeepHealth OS. Think of that the operating system that is cloud-native and cloud-first.
We can also do hybrid deployments that allows PACS and RIS system to be delivered more scalable, more cost-efficient, and that integrates Agentic AI and clinical AI on top to make sure that our solutions can be delivered modular, so not in one big bang approach, modular into the radiology workflow. That infinity loop represents at a very high level the radiology workflow from center operations, where we take in the patients to image acquisition, to interpretation by the radiologist, to clinical follow-up, and so on and so forth. We have five key domains that are based on the DeepHealth OS to address each of those steps in the radiology workflow. Our operations suite automates center operations to make that more efficient and to also address the workforce shortage.
We have patient engagement tools to make sure that we can address the no-shows and get more meaningful results back in the journey that a patient goes through. We have our population health suites, which are really our clinical AI suites across breast, lung, prostate, neuro, and most recently in the thyroid domain. We have our future-forward PACS solution that we labeled a diagnostic suite, which is really cloud-native PACS and reporting solutions, and then we've got TechLive, which is remote collaboration tools to make sure that we address the technologist shortage for operating, for instance, MR scanners. And that's now widened to other modalities as well, such as ultrasound and so on and so forth. With MR, we've seen in the RadNet services business a remarkable impact with our TechLive proposition, reducing the closure hours by 42%. And obviously, that's good for patients.
That's good for the operations of RadNet. We often get the question, what does the combination of RadNet and the services business and the digital health segment deliver? And what comes to mind immediately with people, well, clearly there is a unique amount of access to data to train the AI models. That's definitely true, but the opportunity that we are capturing is much, much broader than that. It's a deep symbiosis between the services business and the digital health business in the sense that we operate with the innovation and technology flywheel. So any idea, any challenge, any opportunity that resides in the services business is brought over to the digital health team to think about how can we solve that with technology tools at hand, informatics or AI, you name it.
The solution for that is co-created and then, once mature to a certain degree, operated in clinical and operational practice so that we can refine it. And once we feel that we've got a superior proposition and product based on also all the training from data at RadNet, we then have the opportunity to deploy at RadNet, but also externally commercialize. This makes our ability to prototype, co-create, and bring to market operate at light speed. I've never operated in a role before where the product development cycle was as fast as here. I mentioned 2,000-plus customers. This is across the hospital segment and outpatient segment with very significant logos. So, for instance, existing customers today are Cleveland Clinic, the NHS in the UK, and then obviously the large and sizable outpatient imaging providers such as Akumin or Solis, and then much, much, much more.
We've had very, very strong traction over the last 18 months with our new solutions. And again, you can read on the slide some of the recent wins that we've been able to deliver. The last thing I will leave with you is how we're deploying our solutions at RadNet. There's really four themes that we're jointly addressing between the services business and the digital health business. Image acquisition by remote scanning, I already mentioned that. That's now deployed over 400 MRI scanners, and we're expanding that to other modalities as well. So the deployment there is quite mature, and we're capturing the further opportunity in 2026. Similarly, so for clinical AI solutions, especially in the domain of mammography and thyroid, we've made great progress, but we still see in other clinical domains very significant opportunity. Think of X-ray, think of lung, think of brain, think of prostate.
So still a lot of value capture to come. The third domain is in the reading and report generation domain. So think of upgrading our PACS to cloud-native solutions, including reporting tools. We're in initial deployment now, and we're rolling out over 2026, 2027 to full maturity. And last but not least, for center operations, patient experience in center operations, we've started quite a few exciting pilots. Think, for instance, about smart and digital onboarding in terms of a patient coming to a center, and then through their smartphone, they can enroll immediately such that they don't have to go through the paperwork onsite, and it requires actually much less time and effort. That's in early stages, and we expect that to come to fruition over the course of 2026 and 2027 as well.
And so the point of this slide is we have made great inroads with digital health solutions at RadNet, but there's so much more that we can still do with the existing portfolio that we have at hand, and we'll continue to innovate for further value capture. Mark, with that, back to you.
Thanks, Kees. Before we get into question and answer session, I'm just going to leave you with some of our thoughts about the future outlook of the company in the coming years. I showed this slide earlier, and the one point I'd like to make on this slide again is that there's been an acceleration of our growth over the last four or five years. And one of the questions that we get, and we held an investor day in November of last year, is, do we think this acceleration can continue both on the services side as well as the digital side? And the answer to that question is yes. From an industry perspective, the trends remain strong. We continue to see diagnostic imaging becoming a more valuable and valued portion of the healthcare delivery system. It's the gateway to diagnostics.
And we're in the sweet spot of where healthcare is and where it's going in terms of the early detection of disease, preventative medicine, population health screening, non-invasive medicine. And as the technology continues to improve, which it is and will in the future, there's just simply more and more clinical indications for ordering the types of tests that we perform. We're building centers. There's so much demand for diagnostic imaging currently. There's not enough capacity. And in many of our markets, we're building de novo centers. Last year, we built seven centers. This year, we have, in various stages of development and construction, 11 new centers. And that capacity will also be able to be filled. And as that is filled, we're going to continue our double-digit growth. Tuck-in acquisitions are still available. Again, highly fragmented industry.
We just announced an acquisition on January 2nd of a provider in Florida of 13 centers with over $100 million of revenue. We're excited to have a bigger toehold in that state where we think that there's plenty more opportunities for both de novo centers as well as further acquisitions, and that's the case with all of our markets. We think that there's more opportunity for continued reimbursement benefit from commercial payers as we continue to scale and prove our worth to them regarding our being a partner in getting this business out of the more expensive hospitals into the freestanding centers. We've delivered on pricing increases from commercial insurance companies and from capitated payers, and we believe that will continue.
Additionally, we're working on a number of new joint ventures with large health systems in both our existing markets as well as new markets, and we're excited to hopefully be able to announce some of those in the coming quarters. Digital health, with its initiatives both on the clinical side in a number of different areas as well as the workflow, is growing tremendously inside of RadNet, but more importantly, with third-party customers. And we think that that 30% growth is going to continue into the future. And finally, we're in a very good financial position. We ended last quarter, meaning the Q3 of last year, with over $800 million of cash on the balance sheet. We have about one times net leverage, and so we've got a lot of capacity and a lot of capital to continue to invest in the business.
So this is the outlook we gave for the next few years at our November investor day. We think that advanced imaging on the MRI and the CT side is going to continue to grow on a same-center basis in the mid-single digits. PET CT, which is being driven by a couple of specialty procedures we do, specifically the PSMA prostate exams as well as these amyloid brain studies looking for the presence of these amyloid plaques in the hopes of getting more patients onto some of these newer drug therapies addressing Alzheimer's and dementia. That's been growing double digits in the range of 15%-20%. We think that that could continue in the coming years. And we think that routine imaging will grow kind of in the 1%-3%. From a revenue perspective, what does that all mean?
That means that we think the imaging services business is going to still have double-digit growth for the medium term in the 11%-13% range. The digital health division is growing over 30%. We think on a compound annual growth rate, and we think that that will continue. And we think that's going to result a lot of the digital health solutions that Case had mentioned, when we implement them as we are implementing them within RadNet, will result in 100-150 basis point margin improvements over the next three years. And we're very excited that with that, as that creates more free cash flow and more ability for us to continue to invest in the business. Our plan is to keep leverage low. We're at one times today.
We would not go above three times, and we would only do that in the case of a transformational acquisition that would have a lot of value for our shareholders. And we're going to keep our maintenance CapEx level to where it is today in the kind of 3%-4% range. So with that, I'd love to open up the floor to questions.
One minute.
One minute for questions.
One and a half. I have a digital clock. Okay. All right.