ICAD Inc (ICAD) 2020 Q4 法說會逐字稿

Greetings. Welcome to iCAD, Inc. Fourth Quarter and Full Year 2020 Earnings Call. (Operator Instructions) Please note, this conference is being recorded.

I will now turn the call over to your host, Jeremy Feffer. You may begin.

Thank you, Stacy, and good afternoon, everyone. Thank you for participating in today's call. Joining me from iCAD are Michael Klein, Chairman and Chief Executive Officer; Stacey Stevens, President; and Scott Areglado, Chief Financial Officer. Earlier this afternoon, iCAD announced financial results for the 3 and 12 months ended December 31, 2020.

Before we begin, I would like to caution that comments made during this conference call by management contain forward-looking statements, involve risks and uncertainties regarding the operations and future results of iCAD.

I would also note that management may refer to certain non-GAAP financial measures. Management believes that these measures provide meaningful information for investors and reflects the way that they view the operating performance of the company. You can find a reconciliation of our GAAP to non-GAAP measures at the end of this earnings release.

I encourage you to review the company's filings with the Securities and Exchange Commission, including, without limitation, Forms 10-Q and 10-K, which identifies specific risk factors that may cause actual results or events to differ materially from those described in the forward-looking statements.

Furthermore, the content of this conference call contains time-sensitive information that is accurate only as of the date of this live broadcast, February 24, 2021. iCAD undertakes no obligation to revise or update any statements to reflect events or circumstances after the date of this conference call.

With that said, it's my pleasure to turn over the -- turn the call over to Michael Klein. Mike?

Thank you, Jeremy, and good afternoon, everyone. I'd like to begin by highlighting a few financial highlights with a particular focus on iCAD's top line revenue momentum. iCAD's fourth quarter total revenue of $10.5 million represents a 47% sequential growth over the third quarter.

We are pleased to report that revenue growth of our flagship, high-margin ProFound AI offering grew by 70% over our third quarter revenues. To further punctuate the performance of ProFound AI in the U.S., our most profitable market, total sales of AI and services collectively grew on a sequential quarterly basis by an even greater amount, 72% from Q3 to Q4. It is also noteworthy that this substantial quarterly growth also represents a 21% increase in Detection product revenue over the fourth quarter of 2019, which was the prior high watermark for ProFound AI sales.

In our Detection business, we recently achieved a critical milestone, the installation of our 1,000th ProFound AI system since our product launch in early 2019. We believe it is particularly noteworthy that nearly 40% of these 1,000 installs have been installed during the course of this pandemic. We see ourselves currently in an enhanced position to offer a steady flow of new AI offerings, product offerings that we anticipate bringing to market with an accelerating cadence.

iCAD's newest offering, ProFound AI for risk assessment, is a product that looks into the future and provides a unique and personalized probability score for potential future breast cancers. ProFound risk assessment can find cancers that are not visible or discernible today, yet may still exist in a subvisual range. Our cutting-edge AI risk algorithm could see these cancers 2 years before they enter a physician's visual field and, in doing so, provides a unique future risk score for patients' examination in the future.

With our new risk product, this would be in addition -- which would be in addition to a score for breast cancer today, we result -- we wind up with 2 products, 2 scores versus 1 today, 1 for the future and 1 that examines risk today -- I'm sorry, that examines detection today. We now have a total of 7,500 2D and 3D installations. We see this along with our unique ability to integrate with all mammographic systems, workstations and PACS providers as a significant barrier to entry.

As I've said in the past, there are over 1,800 different combinations of technologies when all put together that we've had to solve over many years to be able to apply our solutions into the market. In addition, in the area of risk assessment, we believe that we are years ahead of foreseeable competition. Our risk assessment software took 8 years of intense effort and literally hundreds of thousands of images. This launch into the U.S. and OUS breast cancer screening market could not have occurred at a more opportune moment. We are now well positioned with a novel product offering optimally suited for a pandemic, where it can assist with the scheduling and prioritizing of patient screening for breast cancer.

Now moving in parallel with the growth of our AI offering, we also experienced significant quarterly growth in our Therapy business with Xoft new product sales and installations in Q4 2020. Xoft new equipment revenues and the installations of new systems at sites are lead indicators for follow-on recurring revenue sales. Post-install sales tied to procedure volume represent the majority of Xoft's quarterly sales volume. Recurring disposable sales and X-ray source usage volume also have a significantly better profit margin and profile than the initial equipment sale. We deploy a classic razor-blade model at Xoft.

While, as a total business, iCAD achieved the above-mentioned sequential revenue growth of 47% Q4 over Q3, iCAD also grew 11% over Q4 2020. Simultaneous with this, we decreased our operating expenses in Q4 by 3% from prior year Q4. Thus, we have realized both leverage and a surge in productivity while our development teams have achieved growth of products and release of products at an accelerated pace. Sales growth, combined with cost controls and the above productivity gains, allowed us to reduce iCAD's pretax loss to $1.5 million in Q4, 16% lower than Q4 -- Q3 losses and 50% lower than Q4 2019.

We also ended the year with cash and cash equivalents of $27.2 million. These numbers are the result of a very deliberate effort and planned trajectory to achieve positive cash flow and EBITDA in our core Detection and Therapy businesses. It is also worth noting that we are achieving a goal that I'd like to refer to as frictionless procurement. This is a dynamic where sites can access iCAD technology by their choice of either an OEM mammographic system, their local distributor, mostly OUS, through tax providers or through our own direct sales force.

In addition, we have also introduced what I'll refer to as frictionless pricing. This will increasingly characterize our go-to-market approach as customers are now able to purchase via: one, upfront licenses with recurring annual service fees; or two, monthly subscription pricing; or three, purchasing on a SaaS basis, which is fundamentally a payment for each patient analyzed. We are excited about the continuing emergence of subscription and SaaS pricing.

Though still in the nascent yet growing stage of adoption, ProFound AI systems secured through these vehicles will increasingly generate recurring and predictable revenues, revenues that will move our already high-margin products to an even more attractive level. We anticipate that over time, predictable recurring revenues will also mitigate seasonal fluctuations and screening -- as screening procedures flow more evenly than health care procurement cycles.

As we reflect on 2020, we effectively lost a full quarter of selling. Even when customer doors were open, we experienced distracted clinical and administrative decision-making. It took almost 4.5 months until mid-summer before sites returned to full screening with 100% of sites open. Active screening has sustained throughout the fall and winter months and now well into 2021. As you may recall, Q3 Detection AI product sales grew 44% over Q2 and, as stated above, Q4 Detection product sales grew sequentially 70% over Q3.

With this as an overview of our recent performance and sales momentum, let me provide some brief commentary about our new risk assessment offering. iCAD's new ProFound AI Risk offering is now available for the large 2D mammography market. As we've said in earlier calls, in many parts of the world, the 2D market is larger than the 3D market. In the coming months, our risk offering will be available for 3D mammography systems as well.

This means that we will shortly have not 1, but 2 breakthrough and novel first-mover AI offering for both 2D and 3D mammography. Both will be offered in the U.S. and OUS, and both will be sold to all channels and utilize the aforementioned range of pricing vehicles. Through the use of frictionless customer purchasing, we are increasingly using the best demonstrated practices of leading software companies, practices that have successfully been pioneered to achieve dramatic growth and value creation.

With that, let me highlight 2 large deals we closed in Q4, Solis Mammography and Wake Radiology. Our 5-year deal with Solis represents the single-largest deal in iCAD history. It includes both ProFound AI for detection and ProFound AI for risk assessment, and it covers 160 mammography systems. Solis operates in more than 80 branded centers in 11 states as well as 7 offices in the D.C. capital vicinity.

Further, Solis has stated its intention to triple its installations end-market size and reach over the course of this agreement. Solis is uniquely skilled at raising patient awareness of their unique mammographic capabilities. This approach attracts and increases patient flow in their 80 served markets. We see this agreement accelerating the adoption of ProFound AI by other screening centers adopting within these 80 catchment areas.

It is our belief that a deal with this magnitude with a clear industry leader validates iCAD's broad-based clinical utility, workflow capability and economic value proposition. The fact that Solis was a flagship site for Hologic's 3D mammography system now vastly expands iCAD's footprint as a provider of AI for all mammography vendors and health care systems.

Now our agreement with Wake Radiology, North Carolina is the largest provider of 3D mammographic services in North Carolina, and it covers 22 systems in many new and important markets. It is another example of the cascading impact of ProFound in essential markets and key catchment areas and where we see the adoption at one notable site influencing or prompting the sale and adoption in other sites or at other sites in local markets.

Wake's use of ProFound Detection AI provides them with a significant differentiator against market competitors. They highlighted ProFound AI in their own self-generated media campaign and press release once about a month ago. This is yet another Hologic mammography site and further expands our well-balanced sales reach across all available mammography systems.

Finally, I'd like to review some developments in intraoperative radiation therapy, or IORT, related to the treatment of glioblastomas on a recurrent basis, also referred to as GBM. As we've discussed at our well-attended Innovation Day in the fourth quarter and in our last earnings call, the updated data on glioblastoma showed continued and, in fact, rather dramatic improvement compared to the control group. In the endpoint areas of overall cancer survival as well as progression-free survival time, dramatic results were seen.

Our new study is now enrolling and is designed to further validate these outcomes at prestigious sites in both the U.S. and in Europe. Dr. Santosh Kesari, a nationally prominent neuro-oncologist at the John Wayne Cancer Center will serve as the principal investigator of our trial and in the trials of IORT, will be used to treat current GBMs following surgical excision of the malignancy. It will be compared to the current radiation therapy standard of care, and the trial is now underway and sites are now recruiting for patients.

The trial design and details can be found at clinicaltrials.gov. The primary outcome of this 80- to 100-patient trial will be an assessment of overall patient survival. The secondary endpoints would be to assess the pattern of disease progression, potential adverse events and quality of life. This is not an overly long or expensive trial since most sites already have access to Xoft FDA-cleared technology. We anticipate the first patient treatments in the weeks ahead. We also anticipate that preliminary data will be submitted for presentation by the end of this year, 2021.

Xoft therapy can be delivered real time at the same time as the surgery and offers an important advantage, especially during COVID. Other forms of radiation therapy typically require the need to wait 4 to 5 weeks post-surgery before commencing treatment. Xoft offers real-time, on-demand treatment for a disease that typically grows as rapidly as 1% per day. Our technology offers a unique clinical advantage and a chance to impact the disease progression that typically moves quickly and do so at a significant and earlier stage.

So in summary, we are now -- we now see ourselves operating with high momentum in a vastly improved clinical and economic environment in spite of COVID-19 and perhaps, in some ways, because of it, we are making significant progress and making important leaps toward throughout our business. Both our Detection and Therapy segments are at a clear and palpable inflection point. With continued growth rates anticipated, we have an increasingly strong balance sheet and continue -- and will continue our steady march towards positive EBITDA and cash flow.

And with that, I'd like to turn the call over to our Chief Financial Officer, Scott Areglado. Scott?

Good afternoon, everyone, and thank you, Mike. I'll now summarize our financial results for the fourth quarter ended December 31, 2020. As I have mentioned previously, we believe it is useful to compare our sequential revenues from the prior quarter in addition to our year-over-year comparisons.

Fourth quarter 2020 total revenues were $10.5 million, representing a sequential increase of $3.4 million or 47% as compared to $7.1 million in the third quarter of 2020. Detection revenues were $8.1 million in the fourth quarter of 2020, an increase of 53%, driven by a 70% increase in Detection product revenue. Looking at our Q4 revenue on a year-over-year basis, total revenues for the fourth quarter of 2020 increased $1.1 million or 11%, with Detection revenues increasing $1.3 million or 18%.

Moving on to gross profit. On a percentage basis, gross profit was 71% for the fourth quarter of 2020 compared to 76% for the fourth quarter of 2019. On a pure dollar basis, gross profit for the fourth quarter of 2020 was $7.5 million as compared to $7.2 million in the fourth quarter of 2019. Total operating expenses for the fourth quarter of 2020 were $8.9 million, a $0.2 million or 3% decrease from $9.1 million in the fourth quarter of 2019. Importantly, our expense management resulted in a net loss of $1.6 million or $0.07 per share as compared to a Q4 '19 loss of $3.4 million or $0.17 per share. We remain committed to managing expenses while strategically investing in our ongoing initiatives to drive growth.

Moving on to profit metrics. GAAP net loss for the fourth quarter of 2020 reflected an improvement of $1.8 million to a loss of $1.5 million or $0.07 per diluted share compared with a GAAP net loss of $3.3 million or $0.17 per diluted share for the fourth quarter of 2019. Non-GAAP adjusted EBITDA for the fourth quarter of 2020 was a loss of $0.9 million, which represented an improvement of $0.5 million compared to the fourth quarter 2019 non-GAAP adjusted EBITDA loss of $1.4 million. Non-GAAP adjusted net loss for the fourth quarter of 2020 was $1.4 million or $0.06 per diluted share as compared to a non-GAAP adjusted net loss of $1.9 million or $0.10 per diluted share for the fourth quarter of 2019.

Moving on to the balance sheet. As of December 31, 2020, the company had cash and cash equivalents of $27.2 million compared to cash and equivalents of $22.6 million at September 30, 2020. During the fourth quarter of 2020, the company received $6.1 million in net proceeds from the sale of common stock to select institutional investors. In addition, as of December 31, the company achieved the revenue milestone as set forth in the Western Alliance loan agreement, and as a result, we have deferred repayment of the note from September 1 to March 2022.

This concludes the financial highlights of our presentation, and I would now like to turn the call over to Stacey. Stacey?

Thank you, Scott, and good afternoon, everyone. Although COVID-19 continued to impact many companies in Q4, including ours, we continue to be encouraged by the overall performance of our company. Despite unforeseen challenges brought on by the pandemic in 2020, our team continued to advance our innovative solutions in a rapidly evolving marketplace. We believe our dedication to providing precise, powerful health care solutions that are expertly engineered to optimize operational efficiency, clinician confidence and patient outcomes will continue to enhance our competitive position in both segments of our business.

While the COVID-19 pandemic introduced practice-changing challenges on many facets of health care, including cancer detection and treatment, there has never been a greater need for our cancer detection and therapy solutions. The World Health Organization recently announced that breast cancer has overtaken lung cancer as the #1 diagnosed cancer globally. This is compounded by the fact that medical care was severely disrupted in many areas at the height of the pandemic, resulting in an enormous backlog of mammograms, which is now estimated to be as high as 8 million to 10 million as a result of decreased cancer screening last year. Given these circumstances, we firmly believe that our technologies offer a unique value proposition that is exceptionally timely and positions the company and our technologies for success as we look ahead in 2021 and beyond.

Let's begin by highlighting the success of our company's latest advancement, ProFound AI Risk. This first-in-time technology uniquely combines age, breast density and subtle mammographic patterns, including those that may not be apparent to the human eye, to yield a highly accurate and personalized 2-year risk assessment. We are excited about the benefits this software can offer to both clinicians and, most importantly, patients. And we are continuing to drive forward several key initiatives relative to expanding our clinical validation on a global basis, including developing KOL research sites and clinical centers of excellence with multiethnic, multi-geographic data diversity, all aimed at driving adoption globally and ensuring product efficacy across a diverse range of patients.

We are working to finalize the risk algorithm for 3D breast tomosynthesis. This process requires a collection of 3D cases for both training and validation. We have made significant progress in collecting the cases needed for each of the DBT systems to be supported, and the preliminary performance results with the 3D images are very promising.

ProFound AI Risk will be particularly well positioned for success in the years ahead as mammography begins to transition from what is an age-based screening paradigm today to a risk-adjusted screening paradigm. This technology offers a practical solution that empowers physicians to offer more personalized screening, truly individualized for each woman. We believe this technology may ultimately lead to supplemental screening being applied to a smaller, but correctly targeted percentage of women and to lowering total cost per patient spend over time.

Expanding the body of clinical evidence supporting our technologies with high-impact clinical studies remains an important focus area for us. ProFound AI Risk is already supported by a recent study published in radiology. And in Q4, we continued efforts to initiate a multicenter European initiative to conduct a global retrospective analysis of 2D risk. We are continuing to work with a number of highly influential thought leaders in Italy, France, Spain and Germany to refine the study protocol and ensure multi-geographic data diversity.

As we have previously stated, a U.S. retrospective analysis study of Risk 2D and 3D will also be led by Dr. Emily Conant, Professor of Radiology at the hospital of the University of Pennsylvania, whose data also reflects an ethnically diverse group of women with a special focus on genetically predisposed women at high risk, such as younger African-American women. We are currently identifying additional sites for inclusion of other ethnic groups, including Southeast Asian-American and women of Hispanic origin. It is our hope that this research will contribute to the growing body of evidence supporting our technology and potentially pave the way towards more personalized screening recommendations in national clinical guideline organizations such as National Cancer Institute and the National Comprehensive Cancer Network.

High-impact clinical studies also have the potential to feed health care economics and payer relations studies, which could validate economic benefits of the model, improve savings to the health system. In summary, we continue to see tremendous interest for Risk in the market and remain very positive about the impact our innovative Risk solution will have on our business moving forward. I look forward to providing further updates on ProFound Risk in the coming months.

In addition to our focus on ProFound Risk, we are continuing to advance ProFound AI for DBT, our flagship breast cancer detection algorithm. In late 2020, we finalized the development of the third-generation release, which demonstrates improved clinical performance, especially related to specificity or false positives of the algorithm versus the current commercial product. The new version was submitted to the FDA and our notified body in Europe for CE Mark and is currently under review by both bodies. And we anticipate receiving clearance to market and sell in the U.S. and Europe in the coming weeks.

The new ProFound AI release will also be accompanied by an updated platform release, which will further reduce the time it takes to process images as well as introduce the ProFound AI index card, which will provide our customers with a simplified summary of all iCAD AI results in a single view. Additionally, the release will offer enhancements to how our AI integrates with PACS and mammography workstations to further improve reading workflow and efficiency for our customers.

I'd now like to review our strong overall performance internationally. Our total ex-U.S. revenues of $2.8 million for full year 2020 represented a 20% increase over full year 2019. In a year significantly impacted by COVID-19, we are extremely pleased with these results.

Let me review some recent international achievements. During Q4, we sold our first ProFound AI system into Israel and shipped 2 evaluation systems to the largest medical group in that country, where 70% of Israelis are treated. For the first time, we participated in the Israeli Radiology Congress with our own KOL speakers, providing us the opportunity to introduce our suite of products to this large target market.

France continues to be a key market for us. We completed our first installation in the South of France, which generated extensive TV, radio and print coverage as well as an installation at one of France's best-known hospitals in Paris, La Pitié-Salpêtrière. In addition, we signed an exclusive commercial contract with a leading private clinic group. Our sales channel investments in 2019 and 2020 are generating results, and we expect continued growth in this geography in 2021.

Now let's quickly switch gears and discuss our Xoft business, which has demonstrated important progress in multiple areas during a period of uncertainties due to COVID-19. As I stated on our last earnings call, we have brought in new senior leadership in Jeff Sirek, and the business has seen some significant changes from an operational and strategic standpoint. We have implemented a 2-tiered strategy to drive revenues around the current applications while developing the new indications, neuro and rectal, through our clinical trials, registry, KOL sites and pre-commercial efforts, which we will talk a lot more about in a future Innovation Day.

In Q4, there were 7 systems sold worldwide, including our first neuro commercial site in Germany, [the South] Hospital, which will be using a system for multiple applications. Additionally, the skin business was restarted in Q4 due to some favorable reimbursement changes, and we received an order for several systems. These purchases for KOL users were for -- on multiple dermatology sites in both the West and Southeastern U.S. areas. In the OUS business, including China, we have seen consistency with controller placements and source usage. We had 2 milestone wins in China in Q4, including Union Hospital in Beijing, one of the most influential hospitals in China, and they will be treating multiple applications.

Moving on to the new clinical applications. The protocol for treating recurrent GBM has been approved and now resides on the clinicaltrials.gov website, as Mike mentioned. This is a milestone, and it's vital as we finalize the first treatment sites and gain IRB approval from individual hospitals. In the EU, we have focused effort and hospitals identified to join our trial and begin treating GBM patients. Some of these sites are current users and others are completely new health care systems that we are targeting. We expect to be treating GBM patients in 2 to 3 sites in the second quarter.

So in summary, we are excited about the progress and accomplishments we have driven across both sides of our business. And we look forward to providing you with further updates as we continue to advance our business forward in 2021, drive sustained leadership and create additional shareholder value.

Now we will open the call for questions. Operator?

(Operator Instructions) Our first question comes from Kyle Mikson with Cantor Fitzgerald.

Congrats on this great quarter here. So I wanted to talk about the sequential Detection business growth. It was obviously stronger than it was in 3Q, and I was kind of want to break it down to a couple of categories. So first of all, which of the sales channels was the most successful during the fourth quarter? And then secondly, were a number of those installs in Detection delay from previous quarters? And then finally, what are the main types of customers that bought the software?

So Kyle, this is Scott. Look, most of the growth in the revenue was from the direct -- our direct channel versus OEM growth, and it was primarily around 3D and Risk.

Okay. But was anything pushed out from previous quarters or anything like that? Because it was obviously really an impressive number.

So we had a big enterprise sale. We've referred to it a couple of times. Those things take a while to develop. So it's -- it happened in the fourth quarter.

Okay. Got it. And just wanted to also touch on the entrant of the new competitive products in the fourth quarter as well. I know it's been a few months, but just wanted to get an update there. So obviously, it's only going to affect customers with a certain type of camera, but just was curious if there was any changes in the competitive landscape. And how that's kind of impacted the way you're going to market or the customer response to your products? Then of course, how did Risk kind of play into that kind of your strategy there?

Mike, do you want to -- Mike or Stacey, do you want to speak to that?

Yes, sure. I'm not sure if we lost Mike there. Kyle -- is that Mike? Okay, we might have lost him. Yes, so the new competitor in the space that you were referring to, we really feel a lot of confidence in our ability to win deals in those accounts, and we have a lot of evidence so far, given that the large accounts that we have landed even in Q4 that were -- that had that imaging equipment. We really have stronger clinical claims. If you look at the claims from sensitivity to specificity to reading time, these types of things, if you look at those altogether, our clinical claims are far superior. What? Mike?

Yes. I think -- can you hear me? I think my...

Yes, we can. We can. We can. And then so from a competitive standpoint, Kyle, we feel very good about our ability to compete on the clinical performance elements. And then I would say, remember, we're on the third generation of our product, right? So as new entrants are coming into the market, they're coming in with products that are in the first generation. Not only are we already on the third generation, which we expect to come out of the FDA in a matter of weeks that will have further improved clinical performance, we also are now increasingly including Risk as part of all of these deals. And there is tremendous interest in Risk. And in fact, it's actually driving the customers' choice of ProFound AI in many cases. So the combination of superior clinical claims and having a risk that nobody else has, a short-term risk algorithm at this point, I think, puts us in a very strong position competitively.

And if I could add. Yes, I don't know if I'm live now. I think somehow my, yes, my line went cold there for a moment, but I think those points were well established. But I would just add that -- because one of the benefits or just unanticipated consequences of COVID, it's very hard to launch into a market, especially for new players out of the market. And I think we thought we were going to have that dynamic at some point this year. But the competitive dynamics we anticipated seeing, both from international and domestically, we just have not seen emerge. We continue to only see in the U.S. our 3D AI system solution for detection. And certainly, on the risk side, that sort of puts us at what we call the extra mile, where we think there are a few runners. So thanks for the question, Kyle. .

That definitely answers the question. Just a couple of questions. So I've been thinking about change, obviously, recently with the Optum acquisition, and it's our view that Change's imaging business doesn't really fit into OptumInsight's main business areas. And obviously, I mean, United could be better off divesting the business. So I just wanted to ask how iCAD would affected if that were to happen? In other words, if the imaging business was kind of separated from Change, would you be better off or worse off if one of the other PACS vendors kind of added the Change assets and had that larger scale?

We're sort of agnostic on this, Kyle, in that the business with Change is -- we have great expectations for that business as it continues to gain traction. If they merge with another PACS player, that would be fine as well. I mean we're already working with the majority of them. And I will say if they continue in the current constellation that they're in with this part of United, we think that, that is actually very exciting because it provides data to a carrier, an insurance carrier that can actually be used to make positive determinations about the economics of care and someone that gets us in the door that we are longing to get into. And they'll have data from their own sources in terms of the efficacy of care, the -- where the diseases are found, at what stage diseases are found. So we kind of like where it is now, and we're fine if that was to attract itself or attach itself to another PACS or enterprise solution. Either way, it's -- I feel like it's a win-win.

All right. Got it. And just one last question. Hopefully, it will be quick. So you actually mentioned repeatedly recently that both sides of the business should grow at really strong rates going forward over the next few years. Last week or at least recently, you mentioned, I think, 50% growth. So I know you're not providing long-term guidance, but could we maybe see that in 2021 or possibly in '22? I'm just trying to think about the road map in the near term as you kind of drive broader success in both segments.

Yes. We do see significant growth and, in fact, collectively and in both sides of the business. The -- you may have heard me talk about this frictionless procurement, which means multiple channels it can be purchased through, as well as frictionless purchasing, which allows it to be acquired either as a upfront purchase with recurring revenue service, subscription or SaaS. And because of that, there's going to be some undulations in terms of which channels that comes in. Some of these channels actually might, just as an example, take $1 million of sale instead of getting it upfront with some service. It might come in at $400,000 this year, $400,000 in the next and continue to give us $400,000.

So because of that, we've been a little bit -- we felt a little bit premature to provide firm guidance because we -- our goal is to be able to meet the market where it is. But in some ways, we know that if we take $1 million off the board and do it in a SaaS model, it's going to help margin. It's going to help recurring revenue and be predictable, but it could have some effect on this year's revenues to the extent that occurs. There is no doubt that any deal we take today will not only have an impact this year, but will have an even bigger impact for the year -- in the years to come. So back to your point, we do think that the growth can accelerate it in outer years if we take subscription deals or SaaS deals today.

Our next question comes from Frank Takkinen with Lake Street Capital Markets.

Congrats on a good quarter. A couple here on Solis to start off. Obviously, this is a giant win with it being a very, very clear indicator of it being the most superior technology on the market, given it's a Hologic account. So I wanted to ask a couple more questions here and see if we can get some more color.

First, can you give us any color into the expected rollout ramp into Solis locations? Two, any color on whether or not these will be upfront purchases or subscriptions? And then three, as you spoke to tripling, I believe you said tripling the installed base over the next 5 years, whether or not ProFound would be almost a standard option as they roll out these new installs?

Yes. So Frank, it's a perpetual license sale. So we delivered a significant portion of the system to Solis in 2020. And then the rollout and the installation of that into Solis will happen on their time line in 2021. So it's a 5-year agreement. We recognize a significant amount of revenue in 2020 and then the remainder will generally come in ratably over the remaining 4 years just tied to the service and some installation. And then as they grow and as they add more ProFound location -- more locations, they would add ProFound into those locations, and we would get additional incremental revenue when they add them. Does that help?

Yes, absolutely. I...

And I'll add to that. If I can just add to that one piece on the -- going forward, the reason -- there is a particular reason this was made as a 5 -- or structured as a 5-year agreement, which is also consistent with your point about tripling. And they do an [easy] ProFound, both Detection and Risk, as a clear differentiator. So it is integral to the growth equation in markets as they move forward. In fact, there are some anticipated additions to the Solis universe that we expect to hear in Q2 and later this year. And we believe we'll be part of those deals as per the agreement.

Got it. Helpful. On the Solis -- or sorry, on the ProFound Gen 3 that you spoke to, Stacey, is there potentially a commercialization strategy change as you're going to your already installed users with Gen 3, something along the lines of we are going to continuously roll out new generations, maybe there's an interest in converting over to a subscription model where you can be a part of these updates without having to purchase new every year? Or am I overthinking a little bit here?

No. I mean it's a great question. And we already have customers today that are buying service plans that enable them access to all of the generations or software updates that get launched in any given year. So we have some experience in doing this. So I don't think it's going to change our commercial model at all. We are seeing, at this point, some customers who are expressing an interest from moving from a capital model to a SaaS-type model or a subscription-type model. Usually, that customer set may not have budget in 2021 and may not be getting budget until next year, but really urgently want to acquire the technology this year, right?

So I think we'll see a mix of models, and we're prepared to offer the customer the ability to purchase it any way they want, whether it's a perpetual license or a subscription-based model or monthly fee financing. We have the ability to offer all of those different ways, and that's what Mike sort of referred to as frictionless sales model. We can adapt to the customers' needs from a budgeting standpoint, and we are seeing customers want to accelerate the acquisition of the technology regardless of whether they have the budget today. So that's a dynamic that has caused us to be able to offer that flexibility to customers.

Our next question comes from Per Ostlund with Craig-Hallum.

I want to come back to the notion of Genius AI since it was introduced and cleared kind of around RSNA there. Given that your 2 big announcements commercially for the quarter were sizable Hologic sites, is it a reasonable conclusion to think that in a way, Hologic's FDA clearance sort of represents a bit of a dam breaking, if you will? Maybe the market was a little held hostage while sites were sort of waiting to see what Hologic actually had. And that now that it's sort of out there and would seem to be clearly inferior, that sort of driven folks in your direction?

Yes, Per, that's a very good question. The -- there's no doubt that -- it's pretty typical for large market players to want to "freeze" the market or say, we've got a solution we're coming out with. And we got to look at the solution ourselves and side by side with other -- with some of these deals that matured in the fourth quarter. And to not get into too much granularity, I would say, there's 3 things that sites need to see in an AI solution. One is that you want to have -- you want to make sure the detection is high. You want to make sure it doesn't come at the expense of false positives. And you want to make sure it comes -- that all this happens without the expense of increase in reading time.

And we use a calculation, which is called AUC, area under the curve. It's basically the miles per gallon for AI. And in the release, which is published, you can't hide these things. The FDA publishes them on the FDA site, on Hologic sites, the area under the curve, which is sort of a by-product of sensitivity and false positives, what you catch versus what you caught but you shouldn't have. We have a number that's double that of the competitive offering of Hologic's offering.

Further to the point is that you want to make sure one of the reasons for using, that one of the reasons people bought 3D was that it reduced false positives. So you don't want a product that actually increases false positives, especially during the pandemic. The data on the Hologic site, again, I don't typically like to speak this way about competitors, but just on the site, actually increases false positives.

And the third piece, which is probably one of the more challenging elements, is that when you've gone from 4 images to 300 to 400, you want software that's a workflow productivity tool that reduces the amount of time. And we do that to the effect of 53% and almost 58% for complex synergies. And in the case of Hologic, it actually results in increased amount of time, and that's all on the site.

So I think you're correct, people sort of waited, took a look at that and then it basically -- they come back to us and we say, it is a product. And it's not only where it is today, but it's going to get -- it moves to the third generation in the weeks ahead, and it has risk. And I think given those dynamics, it's created a very stable set of circumstances for us.

Sure. Okay. That makes sense. Since GBM hasn't been brought up in the Q&A, I think I'd like to do that. Far be it for me to fast forward beyond the trial that you're just embarking upon now, but let's stipulate that the data that -- the trial that you have being led by Dr. Kesari more or less matches that of the [Gorbachev] in data out of Russia, Mike, you've talked about potentially partnering with another player in this space potentially to act as a bit of a selling arm for you for IORT. Is that still something that's on your radar? Is it something you would consider sort of a bifurcated commercialization effort where you do some direct through your existing channels and partner with somebody else? How could that look?

Well, it is interesting, Per. It's a really good question. How do we optimize the value of Xoft? We see Xoft falling into a couple of buckets. One, you just mentioned, Xoft neuro. There's also Xoft, you might say, women's health, Xoft derm with Stacey, dermatology, and then there's Xoft GI. In each of these areas, it's pretty clear to us that adding our own sales forces in each of these areas, since they're all unique markets, may not optimize the technology in that this technology will -- given the pricing of the X-ray sources and the technologies and the reimbursement in these areas, can allow for distribution partners that even if they took 20%, 25%, 30% revenue would still provide very, very favorable returns to this business, perhaps even greater than the returns today because the reimbursement profile is high.

So when you look at neuro, clearly, it makes sense. While initially, we'll have perhaps 2 or 3 reps that will be working to seed the market that this product is for sale right now. But we're not positioning ourselves as competitive to any player, but we see ourselves as additive or adjunctive. And that sets up a framework for us to be able to partner with other players out there. We may have 30 to 50 people out there and may be an excellent bet to diffuse the technology and still provide very, very handsome gross margin returns.

And by the way, we think that could be the case for not only neuro, but for other areas as well. So in some ways, you can see that we have sort of a hybrid. We nurse the products along. And then where it's a breakout opportunity, we might add some dedicated capability or rather some partner capability to diffuse it into appropriate channels and do more with the technology.

Next question is Dave Turkaly with JMP Securities.

Mike or maybe even Scott, if we look at that $6.6 million Detection product revenue, and I don't think you broke out sort of the 2D Risk from the ProFound AI, but I'd love to get any color there that you might be able to give us. And maybe on the 2D versions, even if it could be sort of a geographic split, but I know you mentioned the 3 ways you're going to price this. I assume in this quarter that most of it was licensing like the ProFound in the past. Would that be a safe assumption? And any color you could give us would be great.

Well, let me give sort of a macro picture, and I'll let Scott color in with some numbers. Keep in mind that we launched it first in international, and we trained the distributors. We didn't launch until September, then we had sales training for reps and distributors in October. So we only had a couple of months, but it was a very meaningful couple of months. And then in the U.S., we also launched it in a soft launch towards the middle of the fourth quarter.

And this was designed because our plan with Risk is to also -- we've learned from Detection to try to get it out to the major sites that are going to publish the data. And if we then have a hard launch with the data published in all these luminary key opinion leader sites already out there, it will strengthen the launch, just a stronger launch, although it would be preceded by a softer launch. Having said that, the product has had unexpected outcomes for us in terms of aligning very well with reimbursement dynamics outside of the U.S., and it actually was a key swing factor in some of the deals.

So Scott, do you want to comment a little bit on some of the elements in the mix with Risk that we saw in the fourth quarter?

Yes. So look, it was primarily license sales as we've been selling historically here. And Risk was a big part of the offering for both Solis and for Wake. So we got some significant revenue. We're not breaking it out, but I'd just say that it's an embedded piece of the overall sale, and it was certainly a big value driver in getting both of those sites.

Got it. And I guess you mentioned the license. I imagine it's a similar price point. I guess my follow-up question would be, if you sold some ProFound AI to some of these accounts now and, obviously, through 2020 as a whole, can we think about adding maybe Risk to some of your already ProFound accounts and maybe putting some sort of, I don't know, $30,000, whatever the license ASP might be? Looking ahead, even on some of the places that you've already put ProFound this year, what do you think the propensity of the upgrade is? And would it be a fee like that?

Yes. So I would say, David, that certainly, when you've got now north of 1,000 installs into a market in the U.S. that's in the high teens and worldwide could be $35 million to $40 million, and we have, as I said, [6,500] 2D sites and 1,000 ProFound sites, the first place you're going to go is to your own sites for both 2D and 3D. What I would say, however, is that -- and we're being intentionally vague on this point by design about switching over to a model where at some point this year, closer to midyear, we're going to move from a licensing model for Risk to a subscription-only model.

And there's a reason for that, and that is that this is an offering that's going to rapidly evolve in new ways. And rather than have to offer an offering 6 months, 12 months with genomics with new risk factors, we want to get people. We're seeding the market in key accounts. And it's almost like the first people that put money down on their Tesla, right? So with key accounts, move the market along, and then what we're hearing, it's not just our model, is that sites are going to want to buy on a subscription basis because the product is going to be such that it will be smarter tomorrow than it is today.

So we are indeed going to go after the low-hanging fruit in our existing accounts. We're certainly going after novel accounts. This is a big differentiator. We saw that in Q4 with these deals, and we do anticipate it moving towards a model where it will be on a subscription or a per-set basis. And obviously, we're going to work with folks that sort of present these models, how we see them laying out. It's one of our obligations to help folks, our investors build their models. So we'll get a proper introduction as we start moving in that direction. But by design, we're being a little bit on the quieter side with when that turnover will happen.

No, I appreciate that. I mean it's going to have a massive impact, positive impact on your P&L. So yes, I think we're all excited about that. I think lastly, you did get asked something about sort of sustainability here. Given -- I know the first quarter usually is a step down from fourth, but the momentum that you saw in a quarter that still had some impact, I don't know, as we look at next year -- and I don't know, I should say later this year now, and vaccines and everything kind of kicking in, in some of those $8 million to $10 million backlogs coming through, I mean I know you don't want to give guidance, but I'm just to wrap my head. I didn't have you doing a $10 million quarter until fourth quarter of '21. So I've been in a bit of a quandary as to how this is going to play out, but it would certainly seem that there could be upside or said differently, that you could sustain this momentum even -- maybe even in some of the earlier parts of '21.

I would say, look, I mean there's no doubt this market is in a buying mood and in an adoption mood. And our approach is to make it, as we've said, frictionless. So the -- so as these sites progress and then -- for example, we're dealing today with -- and I'll just call it, won't get into the account. But as an example, we're dealing with an account that could be a $1 million deal, but it could come in as $400,000 this year, next year and for 3 years thereafter, right? So while it could be an amazing deal, they could also take $1 million out of this year or this quarter, put it into $400,000 spread out over the year, but yet, we're going to love that we did this deal and that we'll have higher margins. So that makes it a little bit tricky, right?

But it's -- the important thing is our mantra is, if people want to buy the product anyway, anyhow, any way they want to pay for it, any channel, we're going to be there. Because we do feel like we have a, not only first mover, but kind of at the moment almost the sole-mover advantage. Certainly, that's the case with Risk. We're going to keep moving quickly.

But you made the point that this is one of the reasons we like the subscription model is that there are episodes during the year when people do their buying, just like people buy cars at the end of the year or the end of the month. The SaaS model allows us to smooth that out over the course of procedures, but there is this hardcoded pattern of when people buy. And it typically is at the end of the year, and they typically kind of fix to get ready in the early parts of the year. So that pattern, we look to smooth out over time.

Next question comes from Yale Jen with Laidlaw & Company.

My congrats on the very great quarters happen.

Thanks, Yale.

My first question is that given that you have 2 major accounts signed in the fourth quarter, even they are not fully -- not all the sites are purchasing, but is that -- do you consider that's a little bit lumpy type of events that make a jump on the revenue? And should we think about that way for forecasting going forward? Or this is really a starting point for already higher baseline moving forward?

There is a clear tilt that a lot of this hit in Q4, but not all of it. And in fact, as Scott will say, the first things get shipped are hardware and some software, and that often could suppress the gross margins as they have been seen in the fourth quarter. But then there are sales that kind of spill out over the next quarter or 2. And then, of course, all service revenues hit 12 months or 12 months and, let's say, the first day of month 13, where the service revenues kick in. It should be noted that these deals also have service agreements with them.

So I think that there's going to be a certain degree of lumpiness, and that will combine with a certain amount of seasonality in the procurement cycle. And some of that is -- leads us to the conclusion that we want to move quickly. And we're going to take on sites that want to buy even if it means that we have to take them in creative ways such as fast ways because in the long run, that's a great opportunity.

So we're still going to have a little bit of this lumpiness and the seasonality for this year, and it's going to get mitigated over time. But it's clearly good news. And there are other big deals in the pipeline that are triggered by these other deals. But it is important to note that all of these supertanker deals, the selling cycle, which is typically now only 90 to 120 days at most, but for these larger sites, they're more -- they're closer to 4 to 6 months. There's just so many more people that have to be involved. So those deals tend to push out a little bit and take a little bit longer. Hope I've been able to help your...

Okay. Yes, that's very helpful. And maybe follow-up a little bit on that, which is that both for the Solis and the Wake, at least on the fourth quarter that the site actually make the purchase. Among all the sites, they -- under their management, is that a very small portion of that? I know it's a 5-year deal for Solis, but how should we think about at least on this particular vendor?

I'd say, in general, you should take it. So for Solis' current accounts, you should think of it as mostly all Solis deals were captured in the fourth quarter, to some extent, in other words, not every piece of revenue, but the super majority of revenue in Q4, and certainly, the hardware and software and a big deal and a good amount of the software.

Wake has more momentum as not all of their sites will be up and running, and that will be something that will flow over the next 6 months. You may have noticed that even when we did deals 1.5 years ago with SimonMed, we announced the big $1 million deal, that took 3 quarters for that to be fully implemented. In the case of Solis, much of that has already been captured in the fourth quarter with a certain amount that will go through, as stated, in Q1, a little bit Q2. And also notice that we've not included the service revenues, which are significant, that won't kick in until 12 months after install.

Our next question comes from Gene Mannheimer with Dougherty & Company.

Congrats on the strong finish to 2020. On Wake, nice agreement there. I just wanted to clarify, they committed to buying the Risk product at this point, although -- so when it comes generally available? Or is that going to be a separate sale?

I'm going to turn this one over to Stacey because she brought this deal in with Scott. So she knows all the details. Stacey?

No, this is actually a team effort on the part of our sales team and many people in the company. But to answer your question, Gene, they are trialing our Risk product and making a decision on how much they would like to deploy that throughout their sites. So this is a little bit of a sort of half purchase, half trial scenario here, and we got the first part of the purchase. And now we have some products on trial, and then they'll make a decision on acquiring the rest of the solution. But we're very hopeful that they will see the value in Risk.

Okay. Excellent. And Scott, your OpEx was down 3% in the quarter, as you called out. I'm just curious if that's representative of the go-forward cost structure or do you expect expenses to pick up as travel and hiring kind of kick into gear this year?

Yes. Yes, look, clearly, with a bigger revenue quarter, there's onetime expenses in the quarter. So I would say kind of Q4 OpEx wouldn't be your opening run rate for 2021 for sure. And as far as travel and the rest of the stuff, Gene, we'll see how it goes. Obviously, there's not a lot of people traveling yet, and we'll just -- we'll build those expenses in as we start -- as it starts to unfold.

Gene -- thanks, Scott. Just going to add one piece to that, Gene. It's been really, really interesting. I think all of this, we always felt that we have to show up and do that because it was the norm, right? You have to show up. You have to -- when I say show up, I mean, you got to be there to finalize deals and all that. Certainly, my 90% travel supports that. But we've been amazed in how much we can get done maybe because the norms have changed just by all these Zoom and Teams, et cetera.

I would say that in this year, the -- we've looked at it, and since that most of the conferences this year are still going to be pushed out, that we've converted to far more cost-effective ways of using online tools. Some of these conferences can cost $300,000, $400,000 to attend with travel and meals and all that stuff. We've been able to get the same kind of face time through Internet tools and webinars and things such as that. So we're thinking carefully before we add back those full expenses.

And the other thing I'll just add to that is that in terms of productivity, I -- this is one quick anecdote. I said to one of our software engineers since -- as the whole software team had a very productive year. I said, so you've gotten incredibly productive this year. You'd like did like 2 years of work in like 0.5 year or 9 months. And he being a thoroughbred, self-proclaimed, answered and said, Mike, I've been waiting my entire life for this moment, where I can just code, code, code, and I have to go to business trade shows, et cetera.

So I think we're going to get the dual productivity increase and cost savings for this year. And I think we're going to have to figure out what the right blend will be going forward. For example, should we be -- we're going to be taking a look at downsizing facility sizes, travel, all these things. We're going to have a new reality of work, and it's just going to be more cost effective.

Yes. And I think, fortunately, for us, Stacey's team had -- the marketing team had the vision to be out in front of this before the pandemic hit. So we were really able to kind of capitalize on that and get through 2020 with some of the investments we had already made.

Okay. Very good. And last thing for me. On the GBM side, how far are you along in recruiting those 80 to 100 patients? I'm just curious how difficult it is to maybe identify the patients that are to be best suited for this trial.

Well, the good thing is we have 8 sites, right? And at each site, we'll probably see 5 or 6 patients that are candidates, and we only need 80 patients, right? And we think we can -- and we expect to actually treat our first patients at, at least 2, likely 3 of those sites in the next 30 to 45 days. So once we get going -- COVID slowed things down a little bit, but not much because after all, single fraction radiation therapy is great appeal during COVID. You're not going to get 6 to 8 weeks radiation therapy.

So we're recruiting. We'll have 3 sites -- up to 3 sites actively treating, we believe, let's say, late March, April, and then we'll have all stocks by in Q2. I can't exactly see how long it will take us to get to 80, but it's not a big number if you can get all sites treating, it could be 12 to 18 months. And that's pretty useful because a lot of the survival time for patients can be measured during that time. And one of the reasons we say, you'll see information towards the end of the year, is that even if we have a dozen patients done by midyear, we'll track them at 6 months. And by that point, we will have been beyond the control group's recurrence rates, and we'll be able to see some meaningful data even in the latter part of this year.

Now we're super excited about this, not really expensive trial in that they already have the equipment, and it's just balloons. Really, what we're paying for is the imaging. And people can also use it commercially. So we hope to get some commercial sales since it is an FDA-cleared product. And we can't -- they can't use it commercially and get reimbursement if we're also giving them the balloons, which are not very costly or/and paying for the imagery. So I think we're going to see a lot of developments kick off this coming year, particularly as we get through midyear on GBM.

Next question, Andrew D'Silva with B. Riley Securities.

And sorry, I was hopping between calls. So just let me know if you answered any of these, but I really wanted to delve a little bit deeper into the Solis win. Is it just fair to say that the overwhelming majority of the revenue related to that contract was already recognized in the fourth quarter? And if so, could you kind of let us know how much of the fourth quarter year-over-year growth in Detection would you say would be attributable to that?

Well, I'll add to the first point. I'd say that a very significant part of that hit Q4. And Scott, you can give the breakdown of this 21% because you heard the comment, Andrew, about 21% year-over-year Q4 product sales. So that's -- what you're looking for is the subset of that, that was Solis, right, or how much Solis represented of that?

Scott, I'll turn it to you.

Yes, we're not -- I'm not going to break it out specifically here, Andy. But it's a significant portion of the year-over-year growth, right? But remember, so just -- and just to put some color behind this, right, we had a pretty significant Q4, kind of onetime Q4 deal last year as well. I know you weren't here. So fourth quarter is certainly the time when we get both years comparatively, we've had some big deals, but this one was a bigger percentage of the quarter than it was last year.

Okay. That gives a good context on kind of thinking about how to model it sequentially and year-over-year.

Let me just add one thing, Andrew, and that is that if you look -- if you Google Solis, obviously, you'll see that they did an acquisition even in Q1 after this deal, 17 systems, right? So we don't have the exact closure date of that, and there may be a lag between that and when orders are placed. But that's an example of this, this is an account that 160 sites that wants to go to -- 160 cameras that wants to go to 500 and wants to go from 80 locations to 250 and get there with us. So there is this deal, and then there will be -- and there is the revenues that may follow. As Scott said, the majority of it is going to be in Q4, but some of it will hit in the early part of this year on a more modest amount than the service revenue. And then there is, again, the adding to the Solis Mammography universe, which is a very aggressive private equity-backed company.

No, that's a very useful context. And just given Solis' historic equipment base and how that kind of ties into the competitive landscape of Hologic, now that Genius AI has released their data and you have a very notable win subsequent to that, can you maybe talk about the sales and pipeline traction and what the cadence looks like subsequent to that? Have you seen a material change in any sort of at least inbounds or just conversations that are starting or decision-making process? Any color there, I think, would be very valuable given the equipment market dynamics that exist out there right now.

Whenever you're in a sales process and somebody says, look, I want to -- I'm obligated to check out what my mammography then just bring into market, they feel offensive having to do that. In some ways, the fact was the -- it's almost like the -- something like the misery of uncertainty is worse than the certainty of misery. Well, in this case, we have certainty, and the certainty here allows our situation to be actively compared. So to answer your question, it's helping us. It's helping us.

And it does take away the obstacle of something -- it turns into a known something that was unknown. And that, combined with any other competitive threats that we may have seen from abroad, has put us in a situation that 2 years in, we still have not seen a competitive system in the market, and we don't foresee seeing one for the near future. That could obviously change, but it is -- this -- one of the things about AI is that your results are published. It's there for everybody to see. It's black and white.

So there's no selling around that. And that's all we ask for is to educate -- our whole approach here is educating customers, and we sell through education, and that we've done from day one. And as long as we're on that path, I think we'll wind up in a strong position. And to answer your question, this scenario does, in fact, play to our advantage.

Okay. So fair to kind of highlight it as the cadence of the business has materially improved as far as it relates to at least ProFound AI traction subsequent to what happened in December. That -- I mean, if I were to highlight that as like a point, that makes sense.

It's fair to say that if a sale was taking 90 to 120 days and there was another player they were considering, then they've added another 1 or 2 months to that sales process. But if they look at the data and say, okay, I've seen it or if they look at other flagship sites, I mean this is what's so emblematic about this. Some of these sites are converting. That's sort of like, okay, I get it. I've been watching them.

So people watch other vendors, but they also see where the big movers, the big chains have gone. So with SimonMed, which is right behind Solis' side and with -- then, of course, Solis and then big groups like what we've seen in Wake and with MD Anderson, which was another big site, which we had -- at that point, we couldn't really talk a lot about, when you start getting your supertanker sites converting, it starts giving people an indication of where sort of the alpha product is.

Okay. No, very useful. Last couple of questions for me, just going to be related to the glioblastoma opportunity. So obviously, you're doing a much more substantive trial right now, but you had earlier data that was very favorable. Could you talk about what your sales and marketing initiatives are at the early data stage, considering you can actually still sell the product? Is a DRG out there that seems to be favorable when utilizing your systems? And then I would just be kind of curious, if things related to the APM benefit or impact, how that's perceived in the market right now?

Yes. Well, let me do it in reverse order. On the APM side, and I know you don't -- I know you know this very well. The APMs don't only affect breast cancer, they also affect all areas of radiation therapy. So if you're going from, let's say, $20,000 of reimbursement for some of the radiation therapy for brain, and now you're getting reduced from $20,000 down to $12,000, you are going to try to find a way to shorten the radiation therapy cycle to 2 to 3 weeks, which we expect will happen in all areas in radiation therapy, which, by the way, we call that the Canadian model because that's how they typically treat. They hypofractionate, and they do that in Europe.

The fact that you could deliver a significant portion of the dose, if not all the dose, during surgery, allows for a model where people can actually reduce external beam radiation therapy and do so in 2 weeks. And the fact that we're coming under a DRG means that you can get the radiation oncology code after the DRG. So that definitely benefits this.

I didn't know that last part. That's useful. Okay. And just last thing, as I think about the glioblastoma opportunity, is your end market going to be just overwhelmingly neurosurgeons? Or is some portion of that, should we still think about it as like the traditional radiation oncology market?

I think it's going to be neurosurgeons who are going to have to pull. It's going to be both in neuro -- so I look at the neuro-oncologist as sort of like the air traffic controller. He's going to pull everybody together. There's going to be the medical oncologist stuff as well. The neurosurgeon is going to do the cutting, and the radiation oncologist is going to be the -- the radiation, the therapy. The radiation oncologist is super interested in this because the standard we're comparing to is called RTOG 1205, which is a 6- to 8-week radiation therapy regimen that uses Avastin to prevent radionecrosis.

And basically, what the study is designed to do, using the people who are the study leaders for that radiation therapy protocol that are in this study is, can we beat the established protocol in radiation therapy, which is patients with 13, 14 months if you treat with solely radiation therapy, and they go 4 or 5 months between radiation -- between progression. These are rough numbers. So this is where -- and so that is the control arm of our study.

So the evaluation arm is, can we beat the established protocol? So the radiation therapy community is very, very excited to be able to see if this can beat that control. This doesn't mean it replaces radiation therapy nor would it replace temozolomide or the Optune helmet from Novocure, but it may make them better. Instead of having them wear that helmet for 20 hours a day, they only need to wear it 5 hours a day. You might get better results with temozolomide. You may get better results with radiation therapy in a shorter period of time.

So that's what we're looking for, very excited about it. And it's all because we're treating -- we're denaturing, producing double-strand DNA breaks real-time at the time of surgery, we actually produce an immuno response, and we're doing it 4 to 6 weeks. And if you wait 4 to 6 weeks after surgery and you're growing 1% a day, you're already trying to beat that disease that's regrown over the course of time.

I would like to turn the floor over to Michael for closing comments.

Well, I want to thank everybody for hanging in there with us, and I'll quickly summarize and thank everybody for joining us and for those who hung in this whole time, which I believe is the largest turn of people. It has been our pleasure to share our full year 2020 earnings.

And again, to summarize the highlights around our current business, and I'll just iterate a couple of key points. One is the sales momentum we are seeing driven by ProFound with significant growth in Q4 and on top of significant growth in Q3. Significant near-term commercial opportunity with the launch of ProFound AI Risk, a product that's a giant leap, gives us 2 shots on goal in accounts in a growing market of not only 3D, but 2D sites. Technology, that's been validated with major deals at Solis Mammography and Wake Radiology. And finally, a very, very significant opportunity in the high-value indication of glioblastoma with an FDA-cleared product that can be used real-time, it is Xoft IORT, and is one that's actually starting treating patients as we speak and can be found on clinicaltrials.gov.

So we look forward to providing you with further updates on our progress in the coming months, and we want to thank you for your continued interest in iCAD. And please enjoy the rest of the day, and please stay safe.

This concludes today's conference. You may disconnect your lines at this time, and thank you for your participation.