ICAD Inc (ICAD) 2019 Q4 法說會逐字稿

Greetings, and welcome to the iCAD, Inc.

Fourth Quarter and Full Year 2019 Earnings Conference Call.

(Operator Instructions) As a reminder, this conference is being recorded.

It is now my pleasure to introduce your host, Brian Ritchie of LifeSci Advisors.

Thank you, Mr. Ritchie, you may begin.

Thank you, Devin, and good afternoon, everyone.

Thank you for participating on 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, 2019.

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 encourage you to review the company's filings with the Securities and Exchange Commission, including, without limitation, the Form 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 27, 2020.

iCAD undertakes no obligation to revise or update any of the statements to reflect events or circumstances after the date of this conference call.

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

Mike?

Thank you, Brian, and good afternoon, everyone, and thank you for joining us today.

To repeat Brian's introduction, I am joined on today's call by Scott Areglado, our CFO; and Stacey Stevens, our President.

We are pleased to have the opportunity to provide an update on our financial performance, progress on our strategic goals and details on both current and future growth plans.

For iCAD detection business, I am pleased to report that our revenue grew 39% in the fourth quarter of 2019 as compared to the same period of 2018.

Of the 39% growth in the detection business, software license sales grew an impressive 59% year-over-year.

On a sequential basis, detection revenue increased 13% over the third quarter.

2019 detection revenue was up 32% over 2018 on a full year basis.

We clearly exceeded our goal of selling 500 new software systems in our first year of launch of ProFound AI for 3D as well as now 2D mammography.

We achieved these revenue goals while simultaneously building and dramatically expanding our commercialization infrastructure throughout 2019.

Now as I mentioned on our last call, we are beginning to compete for a growing number of large transactions.

In the third quarter, you may recall that this included our first large exclusive Hologic account, Jefferson Radiology.

In the prior quarter, we announced a nationwide enterprise deal with SimonMed, a major GE mammography account with dozens of installations in several major U.S. markets.

I am pleased at this time to report that after 6 months of intensive effort at the clinical, technical and administration levels, we have secured the largest transaction in iCAD's history with a $1.8 million order with a world-renowned academic medical center.

This was a direct sale into an account with Siemens 3D mammography technology.

We also closed another very large deal of several hundred thousand dollars with Banner Health.

Banner utilizes Fuji mammography technology.

The takeaway here is several-fold.

Not only are we selling across all channels in partnership with 3D mammography vendors or directly into our accounts ourselves, but we are capturing business at 2 points in the sales cycle.

One point is at the point-of-sale when 3D mammography systems may be purchased.

According to U.S. MQSA track data over 400 new 3D tomo mammography systems are installed each quarter.

We position ourselves to be involved at this point in the sales process.

The second sales point is our active engagement in selling into the already installed market.

This is an even more robust opportunity as there are now over 9,000 3D systems installed in the U.S., plus an additional 1,000 units installed in Europe, where we have steadily begun to build a sales and distribution capability.

This dual selling point approach has resulted in an 87% increase in our direct sales in 2019 over the same period in 2018.

These direct sales are typically more profitable than sales to either OEM or distribution channels.

With over 500 new ProFound AI installations in 2019, our own installed base of software systems now exceeds 7,000 systems.

In total, 6,500 of these installations are earlier versions of 2D CAD technology and are obvious targets for sales upgrades to ProFound AI and for ProFound AI 2D, along with 3D.

From a product road map standpoint, we are continuing to invest to drive towards risk adaptive screening from the currently prevailing age-based screening model.

The product manifestation of this capability and a clear catalyst for this evolution is our introduction of ProFound Panorama technology.

ProFound Panorama was unveiled at RSNA in December.

Already, it has had an impact on enhancing iCAD's differentiation edge and our goal of continuing to increase the value of leading-edge AI solutions that we provide to health care systems, physicians and their patients.

Panorama technology has 3 elements.

The first is the viewing of current images and generating a cancer probability score.

The second feature is the ability to look back at prior images to compare cancer probability changes over time, typically over several years.

And the third capability is comparing a woman's historical images, probability scores and the trajectory of change to women with similar risk characteristics.

Therefore, each women, we would incorporate a discrete comparative subset for her individualized images from the millions of images that we gather to create a personalized comparative risk predictive score for that woman.

As part of this key iCAD initiative, we continue our efforts in Q4, with further acceleration of efforts with the researchers at the Karolinska Institutet.

This is the institute -- this is rather the foundation of our ability to incorporate patient risk factors into our AI algorithms.

You may recall that our exclusive agreement allows access to a massive collection of breast images and individual risk characteristics captured from over 70,000 women, images and risk factors were gathered annually for 8 years.

Therefore, we have access to nearly 600,000 cases and over 2 million images along with the participants' associated risk scores.

Our knowledge of the most predictive risk factors in combination with both today's and prior year's images enables us to produce a unique and personalized risk prediction score for a woman, one that is based on her own history and compared with great precision to other women with similar risk factors.

This is indeed the core value proposition of ProFound Panorama.

Now we continue our collaborative efforts with our Risk Advisory Board comprised of world-renowned mammographers and radiology thought leaders to develop clinical care pathways and protocols for how best to use this new and extremely useful patient information.

Their medical expertise is invaluable to us in both ensuring clinical adoption and early beta testing of Panorama technology.

We believe iCAD and ProFound Panorama technology can serve as the vanguard for what may be an increasingly rapid movement to risk adaptive screening.

Now let's switch to our Xoft radiation therapy business.

We are continuing to focus on innovation, clinical studies and the gradual adoption -- or rather addition of sales personnel late in 2020 when we anticipate an initial lift in our Xoft commercial potential.

Our 2019 revenues were up 3% over prior year, but Q4 had a doubling of revenues, much of which was driven by a previously slower Q3 revenue.

iCAD's $9 million financing in mid-2019 allowed us to accelerate the development of new Xoft applicators for new cancer treatment indications.

With FDA clearance for radiation treatment anywhere in the body and with established reimbursement for all the areas we are pursuing, we are continuing and now accelerating clinical trials to support a staged rollout of several new and exciting intraoperative radiation therapy, or IORT, offerings for cancer treatment.

As just one example, just several weeks ago, we were excited to announce that the first metastatic brain cancer patient was treated in the United States.

The surgical team at the University of Louisville deployed Xoft technology as part of a clinical trial using IORT for patients with large metastases treated with neurological resection.

The strong interest in this new application was driven by clinical results that were reported in the September 2019 issue of World Neurosurgery.

Included in this publication were the details of a matched pair study of 30 patients who are treated for recurrent glioblastoma, or GBMs.

The IORT group of 15 using iCAD's Xoft technology was treated with a single fraction of Xoft radiation, immediately following surgical reception and without chemotherapy and without the use of TMZ.

The comparison group of 15 was treated with routine postoperative adjuvant chemotherapy and with concomitant or sequential external beam radiation therapy.

Medium overall survival for the group with Xoft IORT was 24 months as of September 2019.

The overall survival for the other group of 15 patients not receiving Xoft IORT was 21 months.

As of December 2019, 9 of the 15 patients treated with Xoft IORT -- of the Xoft IORT line were still alive whereas none of the 15 patients in the other match peer group survived beyond September of 2019.

Treating sites are still capturing results as they are being submitted for societal review and abstract.

Xoft IORT for brain cancer is an emerging groundbreaking procedure that has the potential to extend life and address the significant unmet need in the treatment of metastatic brain tumors as well as primary and recurrent glioblastoma.

Worldwide, almost 297,000 cases of brain and nervous system tumors are diagnosed each year.

GBM is the most common and aggressive type of malignant primary brain tumor with a medium survival of 10 to 12 months.

Treatment of brain cancer is just one clear example of out Xoft's approach, which involves actively working with center of influence physicians.

We see these physicians as the gatekeepers of patient care and able to uniquely drive the adoption of novel technologies.

We are pursuing a similar approach as that in neurosurgery by working with surgeons in both urology and in GI.

Therefore, we are similarly initiating studies with recently developed novel IORT applicators for the treatment of prostate as well as colorectal cancers.

Center of influence physicians have indicated to us that they do not see Xoft's unique soft x-ray technology as simply a product nor as a simple product platform or even a system of care.

Increasingly, Xoft Electronic Brachytherapy soft x-ray capability is seen as an emerging core technology.

Some have recently indicated to us that Xoft may have similarities in terms of broad uses to other core technologies such as RF ablation, cryotherapy or laser frequency ablation.

These core technologies spawned dozens of companies and scores of valuable medical indications and treatments.

We believe the adoption of Xoft for the treatment of prostate and colorectal cancers, along with the aforementioned brain cancer applications, derived its innovative roots and potential from this inherent core technology capability.

We look forward to accelerating clinical study efforts this year and reporting updates as we progress.

Clinical results for Xoft IORT technologies has been submitted for presentation at upcoming shows as well as abstracts to be reviewed at key society meeting.

The still pending alternative radiation oncology codes mentioned on prior calls would be a helpful accelerant to the adoption of these new offerings.

The codes have the potential to enhance payments in key indications and in certain geographic locations.

While we obviously welcome any final announcement or updates on these codes in the near future, we see them more as an accelerant to the growth of IORT, not, however, as a requirement for adoption.

As a final comment on both therapy and detection businesses, we are embarking on a campaign to dramatically increase patient and clinician knowledge of our AI detection capability as well as our IORT treatment capabilities.

We want every woman to know that they can have their current images read today with detection probability scores included.

Also, later this year, we want women to know that their prior breast images can be similarly read.

And along with both current and prior images, women will be able to access their personal risk prediction profile.

Similarly, in terms of Xoft IORT treatment for breast cancer, we want all women to know about our ability to treat breast cancers during the course of a lumpectomy.

We want women to know about our ability to deliver care in just 8 to 10 minutes of IORT without radioactivity and without 6 to 8 weeks of follow-up external beam radiation therapy.

Let me conclude by indicating that we have a new member of the iCAD team.

I'd like to welcome Nathaniel Dalton to our Board of Directors.

Nate is one of the founders of the global asset management firm, Affiliated Managers Group, Inc., or AMG, where he remains a director and a senior adviser.

He has held a range of executive positions at AMG, including General Counsel, Chief Operating Officer, President and Chief Executive Officer.

Nate has decades of experience in both capital markets and in building significant public companies.

We look forward to his continued and valuable contributions to our corporate and capital market strategy as we continue to build iCAD.

With that, I will now turn the call over to Scott, and he can provide a more comprehensive review of our financial results.

Scott?

Good afternoon, everyone, and thank you, Mike.

Before I begin with the financial highlights, I would like to note that during my comments today, I will be referring to certain non-GAAP financial measures.

Management believes that these measures provide meaningful information for investors and reflects the way that we view the operating performance of the company.

You can find a reconciliation of our GAAP to non-GAAP measures at the end of our earnings release.

I will now summarize our financial results for the fourth quarter ended December 31, 2019.

As Mike mentioned, we are pleased with the results of the fourth quarter 2019 with total revenue of $9.4 million, representing an improvement of $2.4 million or 35% as compared to the fourth quarter of 2018.

Our top line results continue to be driven by the strength of our artificial intelligence business, which grew by $1.9 million or 39% to $6.9 million as compared to $4.9 million in the fourth quarter of 2018.

As Mike noted, ProFound AI has been now commercially available in the U.S. for 1 year.

Therefore, while our results versus the prior year quarter are impressive, we are also pleased with another sequential quarter of double-digit growth, which we view as another indicator of the commercial success of our ProFound AI product.

As I have discussed on prior calls, there are several key trends we have been tracking in our AI business.

As Mike noted, for the 12 months ended December 31, 2019, direct sales grew 87% as compared to the 12 months ended December 31, 2018, and accounted for approximately 60% of the overall revenue in the AI business, reflecting our investments in our commercial infrastructure at the beginning of the year.

Our OEM business grew 34%, with 3D sales representing a significant contributor to the overall growth in both the U.S. and OUS.

Our U.S. business continues to be the significant driver of our overall revenue increase and has been primarily 3D sales.

We continue to expect OUS will remain the primary markets for 2D and 3D as both are growing OUS.

Moving on to gross profit.

On a pure dollar basis, gross profit for the fourth quarter of 2019 was $7.2 million as compared to $5.4 million in the fourth quarter of 2018, representing a $1.8 million year-over-year improvement.

On a percentage basis, gross profit was 76% and 78% for the fourth quarter of 2019 and 2018, respectively.

In our implementation of ProFound AI in 2019, we prioritized an improved customer experience, which translated to slightly higher costs, resulting in detection gross margins of 84% in the fourth quarter of 2019 as compared to 87% in the comparable quarter of 2018.

However, moving forward, we expect margins to stay consistent with 2019.

Operating expenses in 2019 reflect the ongoing and planned future improvements to our AI algorithm, development of new products such as Panorama as well as the investments in the commercial infrastructure responsible for driving the top line growth in the detection segment.

These investments translated to a $0.9 million or 11% increase from the $8.2 million in the fourth quarter of 2018 to $9.1 million in the fourth quarter of 2019.

Looking ahead, we expect operating expenses in the first quarter of 2020 to trend similarly to the fourth quarter of 2019 as we continue to focus on both near-term opportunities to build market share as well as invest in sustainable long-term growth.

Now summarizing our profit metrics.

GAAP net loss for the fourth quarter was $3.3 million or $0.17 per share loss compared to a net loss of $3.3 million or $0.20 per share loss in the fourth quarter of 2018.

Non-GAAP adjusted net loss, which primarily excludes the impact of the fair value of the convertible debentures, was $1.9 million or $0.10 per share loss for the fourth quarter of 2019.

On a non-GAAP adjusted EBITDA basis, the net loss for the fourth quarter of 2019 was $1.4 million, essentially flat as compared to the fourth quarter of 2018.

Moving on to the balance sheet.

As of December 31, 2019, the company had cash and cash equivalents of $15.3 million compared to cash and equivalents of $12.2 million as of December 31, 2018.

As we disclosed in our 8-K filing last week, we executed the forced conversion of all the outstanding principal amount of the convertible debentures.

As noted, the company issued to the investors approximately 1.8 million shares of common stock, reflecting repayment of the outstanding principal amount of the convertible debentures and the make-whole interest provision.

This initiative eliminated $13.6 million of debt from our December 31, 2019, balance sheet and strengthened our overall financial position, which will be reflected in our first quarter 2020 balance sheet.

In closing, we are pleased with the momentum built in our business during the course of 2019 and look forward to continued success in 2020.

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

Stacey?

Thank you, Scott, and good afternoon, everyone.

We are excited about the strength demonstrated in our business in 2019, which we expect to continue in 2020, and I'm pleased to provide you with updates around our initiatives aimed at accelerating our growth and expanding our global footprint.

Importantly, our performance continues to be particularly strong in detection as ProFound AI further penetrates the market.

The qualitative feedback we continue to receive from our customers about the clinical and workflow impact this solution is having on their practices and their patients continues to be highly positive.

It's truly gratifying to know that more cancers are being found earlier and faster with this revolutionary and first-of-its-kind solution that is truly changing the landscape of breast cancer detection.

We are finding that customers are achieving real-life clinical results that even exceed our FDA claims.

For example, Regional Medical Imaging, the largest independent imaging group with 10 locations across Michigan, is realizing a 30% reduction in biopsies since implementing ProFound AI, just a remarkable impact when considering the impact on patients from not having to undergo the stress and anxiety of unnecessary procedures.

On the therapy side of the business, we're encouraged by continued positive momentum on the clinical data front with new data and multiple abstracts being accepted across several clinical applications at various upcoming events in the spring.

In addition, as we have discussed previously, the new proposed bundled payment radiation oncology model in the U.S. could be an accelerant for our IORT business and provide even greater access to this technology to patients by the end of the year.

With that, I'd now like to provide you with an update on our specific activities on both our cancer detection and cancer therapy segments.

Let's begin with cancer detection.

We have just completed the initial 12-month sales cycle for ProFound AI, and we are thrilled with the key trends that are emerging from the marketplace.

We are closing larger-scale accounts, and our pipeline for these deals continues to grow.

As we continue to pursue large contracts for ProFound AI, our message is crystal clear: we are focused on improving clinical outcomes, patient satisfaction and clinician experience while decreasing the overall cost of care.

We believe this strategy is well aligned with the Institute for Healthcare Improvement's Triple Aim initiative, a framework our customers are using for optimizing health system performance with a 3-dimensional approach that includes: one, improving the patient experience, including quality and satisfaction; two, improving the health of populations; and three, reducing the per capita cost of health care.

Importantly, we are delivering this message to the C level at our targeted customers.

We believe the success we are achieving in selling ProFound AI indicates that this message is resonating well.

Moving on, we're beginning to actively engage with the largest integrated delivery networks and group purchasing organizations, which over time will serve as a channel for volume-based transactions, which will help us accelerate revenue growth.

We continue to achieve important progress towards securing these national contracts, and we are expecting to secure our first nationwide GPO contract this year.

With regard to our strategy with the leading PACS players, we were pleased to announce in Q4 a formal relationship with Nuance where ProFound AI will be available on the largest open AI marketplace.

This also creates an enabling architecture for the PACS companies to interface with.

Nuance is a radiology reporting system used by over 80% of U.S. radiologists in over 6,500 health care facilities.

We remain in active negotiations with a number of PACS companies.

And while these contracts do take time to materialize, we remain very confident that these companies will represent an excellent complementary distribution channel as we begin to add this SaaS-based component to our business with its predictable recurring revenue.

Now turning to Europe.

We made substantial progress in 2019 in building our commercial strategy for long-term success.

We established a stronger direct presence in the geography and built a distribution network that now includes coverage for 19 countries.

We opened up a French entity and secured business with the VIDI Group, the largest private radiology network in France.

On the marketing front during the fourth quarter, we hosted a symposium at the French radiology congress with key opinion leaders from France and Belgium, and we had well over 200 attendees.

Our efforts in Europe delivered orders in late 2019 from Greece, Austria, Belgium and Ireland.

We believe this increased visibility, along with the strong positive response to the launch of ProFound AI for 2D, will continue to drive further ProFound AI sales in Europe and we already have a robust European sales pipeline for 2020.

We believe that Panorama with its prediction capability set to launch later this year is particularly well suited for the European market as this geography is already ahead in terms of risk adaptive screening programs.

I'd now like to briefly highlight our success at RSNA, the industry's biggest trade show of the year, which took place in early December in Chicago.

Our very strong presence at the show included hosting numerous well-attended events for investors and customers and featured multiple clinical presentations throughout the show on our leading advances in breast cancer detection.

As a result of these presentations and our robust promotional efforts prior to the show, we achieved the highest number of prequalified sales leads in the company's history, which we expect to power our strong pipeline momentum in 2020.

We had a number of important successes in the fourth quarter, especially around breast cancer awareness month in October.

We had more than 20 media placements that reached an audience of more than 250 million individuals across key national and local networks.

From a social media perspective, we reached over 39,000 individuals in October alone with messages about the unique value of our portfolio of breast health products.

We expect that over time, we will be able to leverage this educational momentum and retail pull-through to help drive our sales strategy.

Now moving on to cancer therapy.

In 2019, we placed 21 therapy systems globally, which included 13 IORT systems and 8 skin capital systems.

We are pleased with our growing installed base, and Xoft systems are now in use treating multiple clinical indications at well over 150 centers globally.

Looking ahead, we're very excited about the future of our IORT business and continue to work on completing research for new applications.

Mike discussed brain IORT and the very promising matched pair GBM results presented at the INS 2019 show earlier this year.

In prostate cancer, our applicator for use in robotic surgery is generating a significant amount of early interest from key international surgeons and radiation oncologists.

We've already developed a focused business plan for this exciting indication and look forward to the commercial availability of this applicator next year.

In rectal cancer, the clinical study protocol is being finalized with the goal of enrolling the first patient in this trial late in the current quarter.

An abstract detailing the study has already been accepted for presentation at ASTRO 2020.

In summary, our business is performing well, with multiple high probability ways to drive growth, and we have a number of expected key catalysts in the coming months.

We are really well positioned to continue to create a sustainable leadership position for iCAD, drive long-term growth and generate significant shareholder value.

Now we will open up the call for questions.

Operator?

(Operator Instructions) Our first question comes from the line of Dave Turkaly with JMP Securities.

It was close.

Mike, if we look at that -- the 1 order, the $1.8 million, and we consider sort of ASP, I would think you'd come out somewhere around 60 or so licenses.

I'm just curious, like is that a common -- I mean it seems particularly large, but you mentioned there's a pipeline of sort of similar opportunities.

I guess how many different facilities does that cover?

And how many opportunities like that are there out there for you to sign of that magnitude?

They are increasingly large in nature, David, in that more and more of health care systems seem to be acquiring or lining up to acquire the technology as part of integrated delivery networks, as part of purchasing groups or, in general, there's just been an aggregation with enterprise providers with some very large change.

As an example, we mentioned SimonMed, which was an order earlier this year.

That was over 100 software systems that's been rolling out since the second quarter.

So we would expect the -- to continue to increase our pipeline and at the same time, have these increasingly large deals.

There are, in fact, other deals of similar and, in fact, even greater size in the pipeline.

I do want to highlight, however, David, that this particular account did take 6 months to close, there are a lot of work to close these.

But when the pipeline is big enough, we can work on these longer-cycle sales deals, at the same time the smaller deals, and kind of keep things in balance.

Great.

And then when you look at -- you mentioned all the different camera, kind of Hologic and GE and the Siemens and the Fuji in terms of accounts.

Could you just comment on like, are they -- were they using CAD in the past?

Were they customers of iCAD specifically?

Or were they using any computer-aided detection?

Are these mainly brand-new accounts for you guys?

Yes.

I would say that 80% to 85% of the market is already using CAD in some form.

Obviously, with 7,000 installations, you can imagine that we have a significant percentage of the market already using our technology.

So these are opportunities to be able to upgrade our own installed base as well as accounts that may not have been on that 15% to 20%, that may not have used CAD or AI at all as well as converting other accounts that may be on a competitive 2D platform.

And of course, that's greatly assisted by the fact that we're the only one in the market with FDA clearance for 3D -- for AI on 3D mammography.

So part of what you're hearing is us moving as quickly as possible.

We do have a first-mover advantage, and we want to continue to move rapidly.

And just last quick one for me.

Just to be clear that some of these big accounts or specifically the one you mentioned in the press release, were they a customer of yours already?

Or is that a new relationship?

Yes.

In accounts such as that, what we're finding in a lot of these accounts is that we may have had some installations, which was the case here of 2D but then may have been partial in only certain locations.

And a further color in my answer earlier, these accounts, such as the one that specifically have sort of a hub-and-spoke approach, where there's kind of a mother ship in the middle, big university center, and they often have some secondary and tertiary sites associated with that.

And those tertiary sites may be independent imaging centers that are all part of an overall system.

So in a lot of cases, these accounts are partially installed, it's rare to find one that is 100% installed with CAD, so we have the opportunity to upgrade to all systems.

And in many cases, they have more than 1 vendor involved.

By that, I mean, they may have Hologic, Siemens, GE and Fuji or some combination thereof.

One of our advantages is that we are the -- not only the only one in the market, but we are compatible with every one of those systems.

So as we look at the future, this ability to have compatibility with all of those vendors, with all 15 workstations that they display on and actually working with virtually every PACS company that connects them, we believe that gives us a very strong competitive advantage in selling through all different sales channels.

Our next question comes from the line of Per Ostlund with Craig-Hallum Capital Group.

Wanted to look ahead to the other elements of Panorama coming later this year.

So just thinking about the risk predictive opportunity, the opportunity within prior mammograms.

Is there an updated time line?

And is there a specific cadence that you have in mind there that you expect one to come before the other and if there's some intention behind that?

And then maybe related to that, since you referenced the 500 ProFound placements in 2019, Mike, I'm curious how many of the 500 do you have a sense or aware of this pipeline already?

And are you getting a sense of a measure of pent-up demand as you look ahead to the second half of the year?

Yes.

So with respect to your first question, in terms of cadence for ProFound, we are looking very carefully at the launch, the first launch of ProFound, which will be again in the second half of the year, but in the third quarter time period for a European launch, and that's actually by design.

Europe is very active in the -- with their screening programs and is very amenable to the risk adaptive screening model.

Their economics actually even favor this model in that they're looking at the patient over the course of time.

So Europe becomes a very receptive market for us to launch.

It's no wonder why the initial work in the -- the work that we've sort of adopted and worked with, Karolinska, comes out of Europe, or I should say, Scandinavia, in Sweden.

So that will be our start.

And it will actually start on a 2D program because recall that the 2D market in Europe is actually growing equal to or, in fact, slightly faster than 3D.

And that is because in Europe, some of the market is still -- a good portion of the market is still in analog adopting -- moving to 2D digital.

And at the same time, we also have, in Europe, a market that uses 3D tomo more for diagnostic use than for upfront screening.

So what we've also heard is that the ability to have ProFound on 2D actually allows for the -- some of the diagnostic work to be done to actually accelerate the adoption of 3D.

So we actually believe that the launch of 2D in Europe actually will lead to follow-on sales of 3D mammography, which will benefit vendors, but will also allow for a follow-up sale for us in 3D.

And of course, we'll then follow with that later this year with the U.S. introduction.

So in summary, second half, first, in Europe, first on 2D, then in U.S. And in U.S., the focus will be primarily on 3D given the size of the 3D market in the U.S.

The second question, I believe, had to do with the 500 installations.

Would you mind repeating that part of the question?

Sure, of course.

Just was curious, I know you teased Panorama at RSNA -- probably more than teased, I would say, it was quite prevalent.

But in terms of the folks that you've already had adopt ProFound AI, is there -- do you think that there is a widespread awareness throughout that user base that the Panorama suite is poised to expand quickly?

Do you have a sense of kind of that clinical demand, clinical imperative maybe sense out of the people using ProFound already today?

Yes.

I mentioned in my comments there -- it's a good question.

I mentioned in my comments that has kind of led to our further differentiation.

So obviously, the majority of those 500 weren't aware of Panorama as they purchased before -- well before the end of the year because we showed it in December.

But clearly, we're going back to those accounts.

And our opportunity in -- therefore, in 2020, is to not only continue installing new accounts, again, at initial install and with the installed base, but to upgrade existing accounts that we sold last year.

I will say that what Panorama has done for us, we believe, for 2020, where most of the impact will be felt, is give an even more compelling reason to get on the ProFound platform.

So in some ways, what your question sort of gets at is our goal of not simply just introducing an AI tool but a whole platform and suite of capability with the perception and the reality of continued innovation.

I will say that since we've installed -- or since we launched Panorama, the desire on the part of sites to purchase, what we call, technology innovation programs, which is almost like insurance that you may get later generation technologies at a discounted price, has actually had a pretty big uptick, by that I mean a lot of folks want to know if Panorama will come with a purchase.

And let me just answer that quick.

Panorama will be a separate and discrete sale.

It will not be included.

However, we will be launching a new algorithm midyear based on 300% more data than we have today.

And that upgrade to that new technology would be included in an innovation program but -- and it is a big incentive for folks to purchase that.

But obviously, what that innovation program does for us, along with Panorama is keep us ahead and will raise our sensitivity and improve our specificity even beyond what it is.

So yes, cadence, the cadence here is all innovation-driven and a steady introduction into the market of logical next steps improvements.

Perfect.

Excellent.

Just quickly switching over to the Xoft side.

You alluded, obviously, a lot of folks are thinking about the bundled payment model.

Do you have any updated sense that you can communicate, of course, but any updated sense of what your consultants and that sort of thing are telling you about?

When we might expect something out of CMS, and if there's been any kind of leaning as to where they're going to go with the final rule versus proposed?

I'll make a quick comment, then I'm going to turn it over to Stacey who tracks this probably on a daily basis.

In a town or a city where there are very little secrets, in D.C. -- or poorly kept secrets, this has got to be one of the best kept secrets I've ever seen in that it's absolutely impossible to get a clear read and no matter who we speak to and what large industry parties are involved in terms of getting a consistent read.

There are different comments you can hear passing in hallways but until we see something in writing, we chuff them off to rumors.

And as you know, rumors travel halfway around the world before the truth has its boots on.

So I will turn it over to Stacey to maybe color that in a bit.

Yes, Per, to answer your question, there really is no formal documented update, right?

So the best that we can do is to circle the wagons with all of our various reimbursement consultants and our contacts in the various societies and people that we have relationships with inside of CMS and policymakers and the prevailing wins are still pointing towards a spring release of the final language, right?

We don't know precisely when, but this has been what we have been hearing now consistently for several months that there'll be some type of spring release of the final language with potentially a July or September implementation.

So that's the best we know right now.

As we've talked about in the script, so we see it as an accelerant to our business certainly, but certainly not something we're living or dying by for the therapy business.

We have so many opportunities in front of us as we're taking this core technology and expanding to other clinical application areas that the release of these bundled codes would be icing on the cake.

But that's going to happen when it's going to happen, and it's not going to stop us from pursuing the huge growth opportunities that we have in front of us with expanding into some of these other areas.

If I could just add one other point because your question really triggers something that was mentioned in the previously mentioned script.

In the area of breast cancer, we think that the codes are clearly not only an accelerant, but they are going to be a driving element for adoption.

I mean, let's face it, we've been in this business for a while.

We do have new clinical data that's come out and more that will come out.

So that's going to help.

But what has changed with the glioblastoma work and what we've referred to as our clinical influence physician focus is an area where the gatekeeper physicians often are at the top of the food chain in terms of revenue generation for the hospital, the robotic surgeons, the urologists doing the prostate cancer work, of course, the neurosurgeons.

The reimbursement in these areas is already very robust.

And the reimbursement for their overall procedures are robust.

So in these areas, while codes will help, they certainly are not going to be a gating item for our movement into these areas.

And again, we have FDA clearance in all of these areas.

So the codes are going to very much help us on the breast cancer side to move that market.

But as I said earlier, on the position of the center of influence physicians, the technology itself, the other economics of care really pull the patient through the advising radiation oncologist into areas where the economics of care and the outcomes potentially are super ordinately high and will drive adoption.

Again, we want [the code] that will impact all these areas.

But in the areas we're moving into, they are not, as you might say, essential as they might be in the breast cancer area, which is obviously very important as well.

Our next question comes from the line of Brooks O'Neil with Lake Street Capital Markets.

So having been at RSNA, in December, it's pretty clear that pretty much every vendor there was trying to talk on some level about AI.

So I'm hoping you could just help us to understand your perspective on your competitive position?

Obviously, you're the only one with an FDA approval in the breast area.

Where do you think you stand in terms of future competition coming to this business from other vendors?

Brook, it's a great question.

And obviously, I don't only hear that question about, let's say, the obvious players, but sometimes you hear some household names from big companies, Silicon Valley companies may be getting in the space.

We like when we get those opportunities from the media because it gives us an opportunity to counter and talk about how the results we see from others are on 2D, not 3D or they are based on a limited data set based on where we were maybe 3 or 4 years ago.

Here is the differentiator, Brooks, that we see.

7,000 sites sending us data every day, tens of thousands of images being generated every day.

As I've said before, in AI, data is not king, it's the kingdom.

We basically have tentacles out in 7,000 sites to generate these images.

We have the data not only to detect the images today but to look at the prior images.

By the time folks get to the market, we're going to be looking at prior images.

So if you get a score of 50% today, that's only valuable if you could know what the score was 1, 2, 3 years ago.

We're already working on technology that's probably for others, 4 or 5 years away.

And then when we add the risk factors on top of that, based on our Karolinska work, we're talking about the combination of radiomics or imaging, risk characteristics and genomics that we believe put us in the distant -- clearly [distancing].

So I like to think that there are a few runners on the extra mile, and we're going on that extra mile, and we're not looking over our shoulder in that we believe that 7,000 units, rate -- Risk Advisory Board, all channels covered, partnering with all the PACS channels that we're just going to continue to move quick.

So I think we move the -- with what I've said earlier, the economies of speed, we're moving at the speed of light.

That's our mantra.

So we're moving quick.

And let others come in to the market, we intend to be ahead.

Oh, and one last thing.

When we do launch our next algorithm, which will be in the second half of this year, that algorithm will be based on all the images from not only these thousands of images, but by that point, probably 1,000 installed 3D AI accounts, 1,000, and prior images from 1,000, from multiple vendors.

So we believe that we are creating a bit of what we might think of as an insurmountable lead, but we're not going to be -- have too much hubris about this, even though it may sound that way.

We're always going to assume that people are close.

So we continue to move forward with great velocity.

Great.

Second question I have is you mentioned that you have FDA approval for Xoft in all parts of the body.

I'm curious about a couple of things real quick, one is what exactly needs to happen for you to really open the door in the brain for Xoft in terms of commercialization?

And then secondly, what needs to happen in some of these other, take prostate, take rectal, take whatever?

And then third and last, do you see any opportunity to use the AI capability, diagnostic imaging in areas beyond the breast?

That's all that.

Well, those are some good questions.

First, FDA, we have FDA clearance via 510(k), and I want to distinguish that from approval.

And the clearance, which is very much the case for radiation therapy, in general, it's all 510(k).

I was really impressed when I mentioned Varian and played a lead role in the Varian Medical System, as we're launching intensity-modulated radiation therapy, that everything is a 510(k).

So -- and everything is essentially cleared for therapeutic use anywhere in the body.

And in fact, in our case, we don't only have the FDA clearance but reimbursement in these areas.

So the real challenge when you're moving into these new areas is making sure that if you're introducing a new capability that before you go too deeply into areas that are already cleared and already have reimbursement, they do have a successful launch.

So you want to start by doing your safety trials, your initial clinical efficacy trials, even though you have FDA clearance, you want to be able to run single and multisite institutional studies to validate your treatment because you only get one chance at an introduction.

So specifically, in the brain area, we are doing just that.

As you saw, we're taking work that was done outside the U.S. and now introducing it into at least 3 to 4 new sites in Europe, in brain, [Clio], that are in the works.

The same in the U.S., we are trying to replicate the matched pair study you saw.

That and other work, the study being done at Louisville is on metastatic lesions.

That's different than the primary and recurrent glioblastomas.

So we're going to see 2 or 3 different studies coming out of the brain area, and we intend to move forward, again, with FDA clearance and reimbursement.

We need to make sure that new applicators are biocompatible.

When you actually do that, those are sometimes things that you may need to do a 510(k) on if you're dealing with issues of biocompatibility.

Recall that the brain applicator use the exact same balloon as we're using in breast.

So it's already cleared.

For prostate and rectal, I might turn that to my colleague, Stacey, who's been working closely on our efforts -- and in fact, you mentioned in your script about rectal and also can talk a little bit about our efforts in robotics with prostate.

Yes.

I think to your question, Brooks, the nice thing about moving into these new clinical application areas is that there's very little product development that needs to go on.

And in almost all the cases, with the exception being rectal, which is a brand-new applicator that has been developed, in the case of prostate and brain, we're actually using an adaptation or modification of the same balloon that we're using to treat breast cancer, right?

There'll be some differences.

But the real work that needs to be done is in the clinical study area, right?

But where I think we're very optimistic is that if you look at historically what's held back breast from more widespread adoption, certainly, there's been the economic factor, but it's also been clinical data.

And radiation oncology is a very conservative discipline in medicine and typically wants 5 years of data before we start changing treatment paradigms.

But in the case of, say, neuro, these patients have such a poor prognosis, and they're most likely going to die in 10 to 12 months, right?

There just aren't any great treatments for them.

So the requirement for clinical data is going to be much less than would be the case for something like a breast application, right?

So -- but it's the same for prostate, right, and rectal.

It's really the clinical trials, right?

We're ready from a product standpoint to go to market, but we want to establish a body of clinical data that will help us in sort of the full commercialization of a product to the marketplace.

And that is going to be a very key focus of ours across all of these application areas, beginning now and continuing through 2020 as we get ready to launch into some of these new areas in 2021.

And on your last point, Brooks, which is a really good question, is there a way that AI could sort of have an impact or inform what's done on the therapy side.

We increasingly see, and candidly, the folks at Varian have actually done a very nice job of introducing AI-driven therapies, what AI-driven therapy primarily is, is when you capture treatment planning information about a certain category of patients, and then you deliver that care and you know the results, you begin to develop a correlation between the treatment plan, the particular type of patient and the outcome.

So as we start looking at our atlas of 36 different plans that are built into our Xoft technology for breast cancer, for us, that would mean correlating those plans with the women that those plans were delivered on and then tracking what the outcome was.

So AI on the therapy side is really a correlation analysis between the plan delivered, the particular patient and the results.

So that is -- I don't want to say low-hanging fruit, but it's an area of opportunity.

And the same would exist in almost all the areas that we could move into on the therapy side or future areas on the AI side.

Our next question comes from the line of Yale Jen with Laidlaw.

First question just on the therapies of products, you have pretty much a jump for the last quarter.

Could you provide a little bit color in terms of with some specifics of that?

Yes.

Well, as you may have heard in our -- in my remarks, some of the jump, if you look at the relationship between Q3 and Q4, you'll note that Q3 was a below-average quarter.

And what that -- essentially what that represented was a significant order flip from Q3 into Q4.

So -- which showed up in the fourth quarter.

So this is part of what the inherent lumpiness that we see when we've got capital equipment sales.

And also, the fourth quarter for capital equipment sales, in particular, can be a particularly robust quarter.

So that's one of the key elements that we see happening.

There is some -- therefore, what I basically said, there is some seasonality on the capital equipment side.

And in fact, the same is true even on the imaging side to a certain extent in that when you get to the end of quarters, you see a bump.

In those quarters, 90% of hospitals have Q4 as their final quarter or Q2.

So the pattern we see in this, you might say, cyclicality or seasonality can sometimes fall into that pattern.

So I think that's sort of the dynamic in that fourth quarter growth.

Okay, great.

That's helpful.

And also, the Panorama appears to be a very, very impressive product.

I'm just curious, by the time you launched that, would that have any sort of impact on cannibalizing at least new accounts or taking a ProFound AI instead of just directly go to Panorama?

No, we -- it's -- yes, we envision and the way we're positioning ProFound is as a platform.

And you know it's pretty well established what our sort of base ASP is out there, we see Panorama as being an additive feature set on top of that.

So when I made the point earlier about people buying an innovation program, they might just be able to get the new algorithm, such as your iPhone might get a new operating system with an innovation program, but Panorama would be an additional cost on top of the technology, of ProFound.

Now if you were to buy Panorama and ProFound together, it would be almost akin to buying, let's say, an iPhone 13 in that it would be a higher-priced product with new features included.

So as we move forward, we see that people may buy ProFound AI as a base product, which would be useful for detection today, but if they wanted the capability or if they wanted to upgrade from ProFound AI, they would buy an additional feature set or a license for Panorama.

If some, for example, wanted to buy ProFound with Panorama from the get-go, it would be a higher cost product.

But you can also migrate from a purchase of ProFound, of course, we now have 500-plus already on ProFound, we want to give them an upgrade path to include Panorama, but we do not see cannibalization.

We see it as additive, additional feature.

Okay, great.

That's very helpful.

Maybe 2 quick housekeeping questions.

The first one is, what was -- what is the margin for the -- on the therapy side for the fourth quarter, gross margin?

It's roughly 60%.

And what's typical is that the initial -- which is one of the things that many people may not realize about the business -- well, not certainly those who follow us closely like you do -- is that the business is about 60% recurring revenue.

And recurring revenues are important to us because not only it, by its very nature, makes a little more predictive and happens each quarter, but it's typically of a higher margin than that 60%.

So as we move forward, we're very interested in increasing the usage of already installed systems because one of the things that helps drive revenue is increased usage, and that is typically at a higher margin than the 60% that Scott mentioned.

Okay.

Maybe a last question here is, as you guys are retiring the convertible debenture, should we see that the GAAP and non-GAAP numbers, maybe starting from either the first quarter or second quarter of this year, start to -- the GAAP become much narrower?

It absolutely will, Yale.

Just a comment here on the convertible debentures in the first quarter, so the last thing we have to do before we force the conversion was to true up the fair value of those instruments through the conversion date, and if you look at our stock price at the end of December and our stock price at the conversion date, there's a pretty big gap.

That's going to imply that there's some additional fair value accounting and cost in the first quarter of 2020 as we convert those.

That it all flip to equity at the conversion date, so it will all be sort of transparent in the quarter, but there will be a last charge in Q1 2020 related to the debentures before we flip them to equity.

Okay, great.

And appreciate, and again, congrats on the very good year and look forward for this year to your performance.

Thank you.

Thanks Yale.

And our final question today comes from the line of Gene Mannheimer with Dougherty & Company.

I just have 8 questions.

Kidding, guys.

Congratulations on all the great progress.

I know you don't give forward guidance around the financials, but as I think about the number of systems you installed last year, 500 or so, should we expect another 50%, 60% increase in systems placements?

Well, I think that when you look at how we built last year -- and you may recall in my comments, Gene, that we had these clocks that we were trying to synchronize, which was -- first, which is getting sales reps hired , which basically took us 2 to 3 quarters.

And then we had to get accounts from sort of 0 or initial call to close.

And then we had to do all of the connectivity in those accounts.

So these were like 3 clocks of 90 to 120 days.

So with those clocks fully synchronized going into the fourth quarter, which may sort of explain things in the fourth quarter to a certain extent, that pattern kind of smooths out in terms of having our reps and the timing aligned.

The one thing that we'll mention going forward is that we will continue.

We do continue to have an increasing pipeline, we also have larger deals in that pipeline.

So the deals could come in these big chunks, as David Turkaly indicated at that account of 60 at a time, but those deals ideally will hit in in a logical way, but they do come in chunks.

We believe, however, having said that, that our pipeline is strong enough to be able to compensate for that.

So I mean it's also very, very important that as accounts begin to start reporting out their data, and Stacey indicated some of these key opinion leader talks about improvements and sensitivity and specificity, that word begins to get around and that sort of gives us a bit of a tailwind.

Okay.

That's real helpful, Mike.

And on that note, when I hear about larger deal sizes, obviously, positive overall, but I also think more lumpiness and perhaps some loss of visibility, and I realize you have a large pipeline that can maybe help smooth that over.

But can you talk to potential increases in volatility by quarter?

And maybe this is a good time to segue into how your subscription model could help alleviate this and maybe the timing for that.

Yes.

I mean we do have -- even if you look at us historically, in fact, Scott ran these numbers for me a little while back.

If you look at the last 8 to 10 -- 8, 9 years, you will see a seasonal dip from Q4 to Q1, which is the normal purchasing pattern that folks tend to buy.

Just as folks like to buy cars at the end of the year or at the end of the month, folks are trained to do their purchasing in the fourth quarter or, in some cases, in the end of the second quarter.

So that historically has a pattern of Q4 to Q1 that has a little bit of a dip.

The dynamic you referred to about the larger accounts, we believe that on that dynamic, the size of the account and the size of the funnel kind of neutralize each other.

So even though it's great to have these accounts, there's a lot of smaller deals that can get us to where we want to be.

Now on your other point about pay-per-click numbers.

Now let me talk more specifically about the difference between SaaS and pay-per-click.

There's a lot of folks that say, they want to pay as they go, and they say SaaS.

But what they're really saying is that I want to pay for a certain bundle of screens on an as-I-go basis, and they don't necessarily care if it's in the cloud.

So we are seeing that dynamic emerging, and we have previously predicted that, that was mostly going to be a second half dynamic.

I will say that we are seeing an increase in demand.

It's moderate -- modest to moderate of accounts that are saying, "Hey, instead of me buying this and spending, let's say, $300,000, I would like to buy it over 3 years at $100,000 per year." Now though what's interesting about this -- these deals is that they may not have bought at all as a $300,000 sale.

So part of what we're seeing is that there's another part of the market that may not have bought or may have had a discount significantly for another quarter or 2, but we might be able to get them started on a pay-as-you-go, let's call it, subscription model.

And we are seeing a bit more of that going into the first and in the second quarter, a bit more than we may have expected.

We don't think it's dramatic, but it is something that exists.

Now keep in mind that this is good, too.

It's just a -- it is our ability to capture revenues as a recurring revenue stream, not through at one block in time but over many years repeatedly, again and again, and often at a higher margin because the servicing cost and selling costs are less.

So we're modeling this.

We do think it will have some modest impact maybe a quarter or 2 earlier.

We do not think it's fundamentally going to change the shape of 2020.

And in fact, when we look at 2020 overall, we think over the course of the year, all this stuff kind of is a push, but it's almost like the shape of the revenue looks a little bit differently in each quarter based on these pay-per-click subscription models.

And again, we can talk more about that specifically.

And it is a little bit of -- it's a lot of what we expect to happen more in the second half.

But it -- we'd see it slightly creeping in, in the first half.

We have no further questions at this time.

I'd like to turn the floor back over to management for closing comments.

Okay.

So we believe that Q4 may be the first quarter in recent memory where revenues on a sequential quarter basis have moved in a positive direction for both our detection and our therapy business and have done so in a substantial way.

We are indeed pleased with the overall performance of the business in 2019.

Results in the detection business were particularly strong as we've reported as ProFound AI further penetrates the market.

We continue to identify a life-threatening cancer through early and more expedient detection.

This is changing the landscape of breast cancer detection and leading the charge towards risk adaptive screening.

Similarly, we envision Xoft's IORT technology being used to treat many earlier detected cancers and will do so with increasing degrees of efficacy, speed and convenience.

We will continue to use this center of influence concept to drive adoption.

We believe that the emerging breakout opportunities will begin to potentially be realized based on the use of Xoft and the consideration of Xoft as an emerging and enabling core technology.

We believe that, that will create new and compelling cancer treatment indications that we will be able to realize.

And as they are realized, we believe and we envision a considerable return in terms of shareholder value.

And with that, let me say thank you all.

Thank you for joining us.

Thank you for the questions.

And please have an enjoyable remainder of your day.

This concludes today's teleconference.

You may disconnect your lines at this time.

Thank you for your participation.