ICAD Inc (ICAD) 2017 Q3 法說會逐字稿

Good day, and welcome to the iCAD Medical Third Quarter 2017 Earnings Conference Call.

Today's call is being recorded.

At this time, I would like to turn the conference over to Bob Yedid, with LifeSci Advisors.

Please go ahead.

Thank you, operator, and good afternoon.

Thanks for participating in today's call.

Joining me from iCAD are Ken Ferry, Chief Executive Officer; and Rich Christopher, Chief Financial Officer.

Earlier this afternoon, iCAD announced financial results for the 3-month and 9-month period ended September 30, 2017.

Before we begin, I would like to caution that comments made during this conference call by management that contains 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 company's Forms 10-Q and 10-K, which identify 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, November 8, 2017.

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, I'd like to -- it's my pleasure to turn the call over to Ken Ferry, CEO.

Ken?

Good afternoon, everyone, and thank you for joining us today.

I'll begin today's call with some brief comments on the financial results for the third quarter of 2017.

I'll then turn the call over to Rich Christopher, our CFO, who will review our financial results in greater detail.

At the conclusion of Rich's remarks, I'll provide an update our key growth drivers before opening the call for questions.

The third quarter of 2017 represented our strongest revenue [generated] year-to-date.

It was showed by strong product demand in both our detection and therapy businesses.

In addition, we made good progress managing operating expenses when compared to the first 2 quarters of 2017.

We intend to continue to accelerate this progress going forward.

We believe our growth momentum with artificial intelligence, software in our detection business will be particularly important as a driver of top line growth and margin expansion in the coming quarters.

We'll provide an update on this business area, as well as with our therapy business, after Rich, our CFO, provides detailed update of our financial performance for the third quarter of 2017.

Rich?

Thank you, Ken, and good afternoon, everyone.

I'd like to review the company's third quarter 2017 financial highlights which were disclosed in this afternoon's earnings release.

As I will only be discussing the highlights of our financial results for the third quarter, I'd like to refer you to our quarterly report on Form 10-Q for more specifics and details.

The Company expects to file its Form 10-Q for the third quarter of 2017 with the SEC in the coming days.

Before I get started 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 to investors as well as better reflect 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.

With that covered, I will now summarize the financial results for the quarter.

For the third quarter of 2017, total revenue was $7 million, reflecting a $1 million or 17% improvement over the comparable prior year quarter.

The year-over-year change was driven by a 5% increase in detection revenue and a 42% increase in therapy revenue.

It should be noted that our year-over-year growth was negatively impacted by the inclusion of $600,000 in MRI revenues from the prior year quarter.

As detailed on previous earnings calls, we completed the sale our MRI assets in the first quarter of 2017.

Excluding the impact of the MRI revenues, total revenues increased by $1.6 million or 29% over the comparable prior year period.

The year-over-year change is reflective of the solid 23% increase in detection revenue and a robust 42% increase in therapy revenue.

Taking a look at our results by segment.

Total revenue in our detection segment for the third quarter of 2017 was $4.3 million, representing a $200,000 or 5% increase over the prior year period.

Again, excluding the impact of the MRI revenues, cancer detection revenue increased $800,000 or 23% versus the prior year quarter.

The year-over-year increase in our adjusted cancer detection revenues was entirely driven by $1 million or 60% increase in our product revenues.

Shifting to our therapy segment.

Third quarter therapy revenues totaled $2.7 million, up $800,000 or 42% from the prior year quarter.

The improvement was constituted by a $600,000 increase in product revenue and a $200,000 increase in service revenue.

The increase in product revenue is as a result of incremental controller sales in both our skin and IORT categories.

The increase in service revenue is reflective of the growth in a number of subscription customers that are using the Xoft system to treat cancer as well as an increase in the volume of patient treatment administered.

Moving on to gross profit.

On a dollar basis, gross profit increased from $4.1 million in the third quarter of 2016 to $4.6 million in the third quarter of 2017, a $500,000 improvement.

The gross profit improvement was as a result of top line revenue expansion in both our detection and therapy segments.

On a percentage basis, gross profit for the quarter was 66%, in line with our expectations, but down slightly from the 68% achieved in the prior year quarter.

The year-over-year change in gross profit percentage was due entirely to product mix.

As far as operating expenses go, operating expenses in the third quarter of 2017 totaled $11.6 million, representing a $4.8 million increase over the third quarter of 2016.

In the third quarter of 2017, the company recorded a noncash impairment charge estimated at $4.7 million related to the goodwill and intangible assets of our therapy business.

The company completed its annual goodwill assessment and determined that the carrying value of these assets exceeded their fair value, resulting in the impairment charge.

While our therapy business has shown improvement in recent quarters, it has over a longer period of time performed at levels below previously anticipated.

Excluding the estimated goodwill impairment charge, total operating expenses for the third quarter of 2017 was $6.9 million, relatively consistent with the $6.8 million recorded in the third quarter of 2016.

Looking at our profit metrics.

GAAP net loss for the quarter was $6.9 million or $0.42 per share compared to a net loss of $2.7 million or $0.17 per share in the comparable prior year quarter.

The $4.2 million year-over-year increase in our net loss was driven by the estimated $4.7 million goodwill impairment charge, which was partially offset by gross profit growth of $500,000.

On a non-GAAP adjusted EBITDA basis, the net loss for the third quarter of 2017 was $1.3 million, representing a $200,000 improvement over the prior year period.

The non-GAAP adjusted EBITDA improvement was also due to gross profit growth.

Moving now to the balance sheet.

We ended the third quarter of 2017 with $11.3 million in cash on hand as compared to the $8.6 million we had on hand as of December 31, 2016.

During the third quarter of 2017, we experienced a $3.6 million increase in our cash position.

The increase was due to the receipt of the proceeds from our $6 million term loan, which was partially offset by $2.4 million of cash usage in the quarter.

It should be noted that our quarterly cash usage included a final payment of $500,000 related to the settlement of a patent dispute from 2012.

We anticipate the company's cash usage will be lower in the fourth quarter of 2017 as our revenues continue to grow.

As I mentioned on our last call, in August of 2017, the company entered into a credit facility in the amount of up to $13 million with Silicon Valley Bank.

The transaction consists of a $6 million term loan and a $4 million revolving line of credit.

In addition, assuming the achievement of a specific revenue milestone, the company has an option to secure an additional $3 million in term debt.

As stated previously, we received the $6 million in cash from the initial term loan during the third quarter of 2017.

Collectively, these funds help strengthen our cash position and are expected to be utilized to support top line revenue growth as well as to advance our product development programs.

In summary, the key takeaways from our performance are as follows.

First, our top line revenue is showing strong growth.

Excluding MRI revenues, total year-to-date revenue is up $2.2 million or 13%.

The growth was fueled by a $1.5 million or 14% increase in our detection revenues and a $700,000 or 11% increase in our therapy revenues.

Second, and most specifically, product revenues within our detection segment are growing nicely.

Again, excluding MRI revenues from the 9-month period, detection product revenues were up $2.1 million or a robust 36% year-over-year.

Third, service revenues within our therapy segment are also demonstrating strong growth.

This area is representative of our subscription-based skin therapy business.

On a 9-month basis, therapy service revenues were up $300,000 or 6% year-over-year.

While that may seem modest, we view this as good growth given the fact that the subscription business was launched just over a year ago.

Lastly, we believe the new $13 million credit facility with Silicon Valley Bank will help finance the company to the achievement of profitability while, at the same time, allowing us to make critical strategic investments to further support top line revenue growth.

This concludes the financial highlights section of our presentation.

Now I would like to turn the call back over to Ken.

Thanks, Rich.

Let us begin with our cancer detection business.

We're pleased with the positive momentum of PowerLook Tomo Detection, the first artificial intelligence software product based on machine and deep learning that is commercially available today for the detection of breast cancer on 3D mammograms.

To maximize the commercial potential for this first-of-a-kind solution, we will continue to increase all of our market development efforts over the coming quarters.

Efforts over the third quarter included: continuing education of the GE sales force and clinical applications teams to most effectively position and sell PowerLook Tomo Detection; ongoing participation in the GE road show, where GE highlighted its latest breast care solutions including PowerLook Tomo Detection software.

The GE roadshow held events at over 20 cities across the U.S. over the summer and fall.

The feedback we received from potential customers following this road show was highly positive and directly led to multiple sales opportunities.

We recently launched an exclusive women's portal sponsorship with multiple promotional banners and ads on Imaging Technology News, the website focused on imaging and women's health, and is gaining traction, as our powerful tomo detection customer education webinars.

We participated in a number of targeted breast imaging meetings as well such as the 22nd Annual Mammography in Santa Fe in August; Wyoming Society of Radiologic Technologists in September, the European Society of Breast Imaging in September, and are issuing press releases to highlight the success of well-known customers using our software in clinical practice.

Also at this year's Radiology Society of North America, which begins in several weeks in Chicago, we'll have a strong presence focused on the topic of artificial intelligence.

In addition to our usual iCAD trade show booth at RSNA, with the majority of our messaging will focus on artificial intelligence, we've secured booth space in the dedicated artificial intelligence pavilion.

We will make a presentation on the speakers' agenda.

And while we remain very excited about the commercial potential for PowerLook Tomo Detection, we continue to make good progress with development and testing of the next version of the product, which will be available on all the major tomosynthesis systems currently marketed.

We hope to release the next version of the software in Europe in the first or second quarter of 2018 and sometime in the second half of 2018 in the United States.

When the next version of PowerLook Tomo Detection is available, we believe we will significantly increase our addressable market for this, our second artificial intelligence solution.

Finally as I said on our last call, we are in the process of developing a broader road map of artificial intelligence, software solutions based on machine and deep learning.

We have a robust and highly innovative platform as well as a significant advanced engineering and development experience in this area, and we look forward to augmenting our product -- portfolio of products solution in the coming years for the benefit of our customers.

Shifting to therapy.

I'd like to provide an update on our progress with our skin brachytherapy business.

We continue to achieve good progress in adding new customers and increasing treating sites as well as we start an increase in therapy product sales.

In the third quarter of 2017, we added 4 new customers who are actively treating 2 capital and 2 subscription.

Collectively, we have increased total treating sites from 37 at the end of 2016 to 52 at the end of the third quarter of 2017, an increase of 40%.

Most importantly, in the subscription category, which, as we've said previously, provides the most potential for increased recurring revenue for iCAD over time, we have gone from 6 treating sites at the end of 2016 to 27 sites at the end of September 2017.

In regards to treatment volume, we delivered approximately 7,000 fractions in the fourth quarter of 2016 and approximately 15,000 fractions in the third quarter of 2017, a substantial growth rate of over 100%.

In an ongoing effort to continue this momentum, we've continued our marketing efforts, including attending key oncology and dermatology meetings where we can educate and promote the benefits of eBx as a nonsurgical alternative to Mohs surgery in the treatment of non-melanoma skin cancer and assisting our treating sites with programs to increase patient awareness and ultimately increase treatment volume over time.

As a result, the results of our recently published multicenter match peer study for eBx versus Mohs treatment, which we have discussed previously, this is -- were presented at the well-attended oral presentation during the 2017 American Society for Radiation Oncology, or ASTRO meeting.

This was on September 27.

Moreover, this abstract was selected for presentation at the upcoming Best of ASTRO meeting, which takes place November 10 and 11, and recognizes the most relevant and highly influential research from ASTRO for 2017.

Other ASTRO 2017 presentations included data from a study of the treatment of periocular non-melanoma skin cancer using the Xoft system that showed a 99% local control rate for 2 years post treatment and no reported cases of side effects, such as long-term vision changes, dry eye, corneal injury or retinitis.

Also, I'd like to highlight the 4-part webinar we hosted in the third quarter that focused on electronic brachytherapy for the treatment of non-melanoma skin cancer.

This series features 6 different skin eBx expert physicians discussing their clinical experience, ranging from community dermatology offices to large cancer centers.

We had over 130 registrants for the live program with a consistent stream of replay viewings.

Now for an update on other therapy applications, let's turn to our intraoperative radiation therapy, or IORT, business for the treatment of breast cancer and the treatment of certain GYN cancers.

The treatment of breast cancer data was presented at ASTRO on the use of IORT with the Xoft system in the early stage breast cancers with the results demonstrating that the Xoft IORT survival rates were excellent as well as with good cosmetic results and a low rate of high-grade adverse events in recurrences in appropriate patients.

This study is the largest U.S. clinical trial in IORT to date and actually now has about 1,200 patients enrolled.

These presentations were only a part of our significant presence at ASTRO this year.

We also hosted additional presentations in the series of peer-to-peer discussion opportunities with leading experts at our booth.

In addition, we hosted a successful evening of eBx cocktail reception.

Our reception -- presentations and booth demonstrations were all extremely well attended.

Collectively, these efforts generated a significant amount of momentum for our skin therapy business.

Specifically, our presence at ASTRO in 2017 generated 222 leads, nearly 75% increase over 2016, and an all-time high over the last 5 years for us at this critical medical meeting.

Immunization standpoint, we currently have 34 U.S. centers and 48 OUS centers treating patients with IORT and GYN applications.

We placed 1 new IORT system and 1 GYN system in the third quarter.

While we continue to be challenged by a lengthy sales cycle in this segment, we anticipate stronger progress in the systems' placements area in the fourth quarter in the U.S. and internationally.

We're also encouraged by the strong IORT procedure volume growth, particularly in the U.S. and in international markets in the third quarter.

We're also pleased to recently announce that our Xoft Balloon applicators have recently received approval from the China Food and Drug Administration, or CFDA, for the treatment of early stage breast cancer.

The Xoft Axxent Balloon applicators are a key component of the eBx system.

With the CFDA approval, a complete suite of Xoft systems product are now available to clinicians and patients in China.

It's an important marketplace for iCAD, as there are more than 268,000 women in China who are diagnosed with breast cancer in 2015, with more than 69,000 whom have died of this disease.

And the incidence rate has been growing rapidly.

The use of the Xoft system in China is attractive as the ability to deliver the radiation therapy at the time of surgery offering a one-and-done treatment is a very compelling option in China and other countries where convenient access to health care facilities is more limited.

We believe this will be an important revenue opportunity for us over the coming years.

So to summarize the status of the therapy business, we continue to achieve progress in skin brachytherapy, our largest market opportunity.

In addition, we are continuing to gain traction in the U.S. with IORT and GYN, and interest remains strong in Europe and Asia.

Based on all of this, we continue to expect increased therapy momentum in the fourth quarter with an opportunity to achieve stronger growth in 2018 and beyond.

And now with that summary, we'll open up the call to your questions.

Operator?

(Operator Instructions) And we'll go first to Per Ostlund with Craig-Hallum Group.

Wanted to start with the tomo CAD side of things, Ken, you alluded in your prepared remarks to the road show, which we were happy to have participated in.

Since the road show -- or I should really say coincident with the road show, we've seen the MQSA placement data show an uptick in tomo system placements.

Is there a cause-and-effect conclusion to draw there that there is pent-up demand -- or was pent-up demand for the Pristina system?

And could you kind of speak to how the GE placements' shakeout between sort of the normal course of upgrade within the GE product suite versus competitive conversion?

Sure.

I think, Per, for certain of the fact that GE is in a healthy shipping position as well as had a healthy backlog has influenced in the near term the MQSA data.

I think as you go back in time, their Pristina system was approved in late March.

And so obviously, there's been pent-up demand from their loyal customer base, in particular, for some time, and then I think they have ramped up manufacturing nicely.

It was very evident to us in the third quarter, just based on the number of units of detection software they purchased that they were definitely shipping a lot more in that quarter than they did in Q2 and Q1.

So I think the MQSA spike does have certainly something to do with the shipping plan.

How sustainable that is over the long term is another good question, which I really couldn't speculate on, but certainly there was some pent-up demand at a minimum that has been addressed with the last couple of quarters of shipments of their system.

In terms of the upgrade market, it's kind of a mixed bag out there.

A lot of their customers are using the current first generation product, SenoClaire, and are fairly happy with it.

I think there is the opportunity to upgrade those to Pristina over time because there is some real differentiated functionality and some real significant enhancements to the entire workflow.

And so, we are certainly focusing on that marketplace with them.

And at the same time, as they roll out the Pristina systems, we're working with their sales teams, so that they can then upgrade their detection technology.

The majority of shipments, no surprise early on, are to an install base of their 2D systems, and in essence, most if not all those customers have our PowerLook server.

And so in essence in our current kind of upgraded market with GE, we are seeing a blend of orders coming in for 2D software and 3D software as it relates to PowerLook Mammo Detection and PowerLook Tomo Detection.

We think over time that's going to skew significantly towards PowerLook Tomo Detection.

And today, some of the challenges are that the customers really set their budgets approximately a year ago.

And when they were budgeting a year ago using the current products, it ended up that both the Pristina product and our Tomo Detection software now cost more than the versions of 2D for our software as well as their prior system.

So we have some timing issues as the upgrade market kind of evolves and some of the customers have chosen to just add a 2D license in the near term as the 3D machines also do 2D imaging and the majority of their exams today are largely 2D.

But we see a very, very strong interest in 3D detection software.

And I think we will see over the next 12 to 18 months a higher and higher attachment rate of our 3D software as customers are able to budget for it with proper amount of time as well as they see the obvious benefit from the growing number of sites that are using it today.

So I think the market opportunity for new placements, as well as upgrade business, is very healthy.

And I really believe that over the next 12 to 24 months we'll see a steady increase in our detection software sales, particularly around 3D.

And then hopefully in the 2018 time frame with a second version of the product, which would be our second AI solution, which will address an even larger addressable market beyond GE, who we collaborate with today, the addressable market just continues to get bigger as we bring out more and more innovation in this space.

So I think the opportunity is pretty considerable over time.

That is extremely helpful.

I think you answered my next 2 questions, actually, with that answer, Ken.

I guess, maybe just to come back to the point on the attached rates.

I don't know how precisely you can sort of bracket where the attached rates are at this point, but to the extent you can.

That would be helpful.

And then, I guess, is -- are budgetary considerations really the only sort of the primary thing that would be holding people back from buying a Pristina instrument, but not taking tomo CAD at this point?

Yes, I think the budgetary issue, Per, is probably the most significant.

And in essence, what we've learned is that the budgetary pricing for the Pristina system and the actual pricing is probably about a 10% difference.

And then if you look at Tomo Detection software, contrasted to 2D, is probably also a good 10%, maybe 20% difference in the cost of that.

The other thing that's required, which I don't think a year or so ago was as clear as the need to have an upgrade to the mammo workstation that GE provides to the latest software version, which is called Seno Iris, and that also is something that I believe costs in the $40,000 to $50,000 range.

So there were probably a lot of unknowns a year ago.

They really contributed to customers not budgeting enough, and so that has driven a lot of sales in the short term to buying a license for attachment to an existing PowerLook server.

And so the good news is that each quarter the mix of 3D sales is growing compared to 2D sales, but there are some overhang, if you will, as it relates to budgets that were set a year or so go.

The other issue I would say beyond that to your question is, the GE sales force has taken on a lot to learn about the Pristina product as well as this new disruptive technology tool for detection software at the same time.

I think there still is a learning curve.

I believe there needs to be more training, more proper positioning, and there is the competition.

The competition doesn't have this.

So they make up lots of clever responses to why they have a comparable solution or have something in development and so forth, which at least in the short term can be challenging.

Because as we know, GE doesn't typically end up being the low-priced provider in this market space.

So I think there a number of issues on adoption.

What I can clearly state is that when the product is positioned properly, customers love it.

When it is trained properly, customers love it.

We're getting more and more strong testimonials.

We press released one this morning, as a matter of fact, and there will be many more to come.

But I do think for the GE team, they need more education and training.

They need more support by our sales specialists, and I do think we're making steady progress in that area.

And what I also recognize is that, as time goes on and the attachment rate increases, there'll be a lot more business out there for us.

Because in a lot of cases, their customers are buying a 3D machine, but they're still doing screening in 2D.

So that machine, while it does 2D and 3D, is being utilized mostly for 2D screening.

So the urgent need is more for detection software that applies to 2D and that's where most of the volume is.

So that has happened in some occasions, but will -- what will happen is we don't know, as this market is moving to its 3D for screening fairly rapidly.

And so the opportunity for us to come back around in 12 or 18 or 24 months and upgrade a customer that just took a 3D Pristina to the 3D software is a big upgrade market opportunity for us and it will just only add to the total revenue momentum we'll have at that point in time.

So we're doing business with all these Pristina customers.

The mix is not quite as strong in 3D yet, but as I said, each quarter, we see a noticeable increase in the ratio of 3D sales to 2D sales.

And we expect that to continue for quarters to come.

Very good.

Let me ask you one more question quickly on that and I have one that I want to ask on therapy before ceding the floor.

On the past call, at least the last call, perhaps two, you talked about what the sales funnel I believe was for tomo CAD.

I think it was $7.5 million at the end of 2Q and $4.5 million at the end of Q1.

Do you have a sense of where that stood at the end of the third quarter?

You know I don't have an exact number, it's a good question.

We could follow up with you on that.

I'm certain it's a larger funnel that was in end of Q2, but I don't have the exact numbers.

We'll get back to you on that.

Okay, that's totally fine.

And let me just ask you one question on the skin side.

I'm wondering if we can -- if you can help differentiate between a treating site and a skin customer under contract.

We kind of went back to some commentary earlier in the year.

I think, as of the fourth quarter call, there were I think 70-ish skin customers under contract.

And then now we've got 52 treating sites.

So is that just a matter of sort of getting these other folks kind of reactivated at this point?

What sort of the cadence that we might expect out of that?

And are there still subscription customers that are kind of rolling off that service and supply line that kind of obfuscate the underlying growth there?

Right.

Yes, at one point we were talking about total contracts, but then there was a difference between total contracts and treating sites.

So you had a backlog of onboarding sites.

What we chose to do starting last quarter was to talk about the change per quarter in the number of treating sites, and then secondly, the number of fractions delivered.

And when you really think about revenue ramp, particularly in a subscription business but also to some extent on the capital side, what's most important is how many sites you are treating, as opposed to how many might you have under contract.

And then secondly to that is how many fractions are being delivered contrasted to prior quarters?

So what we try to do is really move to the most relevant metrics, uncomplicate this, so that you can track growth in a number of sites treating and the growth in a number of treatment fractions delivered.

And that really is what matters.

And we'll go next to François Brisebois with Laidlaw.

There's a lot of answers there.

So just a couple of quick ones.

In terms of IORT for tracking it down, maybe I missed it, but did you mention the number of Balloons in the quarter?

We did not mention the exact number of Balloons, but we had a very, very strong quarter in terms of overall Balloon sales.

Let me see if we have something we could share with you on that regarding Balloons.

Yes.

Our Balloon volume was actually up 29% in the third quarter of '17 versus '16.

And then on a year-to-date basis, approximately about 10% overall unit growth and more weighted towards OUS.

And as we've been saying, Frank, the number of sites treating has been growing faster OUS, so hence the balloon volume growing a little bit faster OUS.

But it actually had a record quarter in Q3, almost 500 Balloons were sold, something that really was a really strong spike in overall Balloon sales.

And again, as you look at it, it was actually an extraordinarily strong quarter in the U.S. But on an overall year-to-date basis, again, we're about 10% ahead of last year, and that is more skewed with OUS growth than U.S. growth.

Okay, great.

And then the number of treatments on the skin therapy, that was very nice to see, 15,000.

That's a nice uptick.

Is there any -- in terms of timing for maybe using tomo and the AI for -- in different disease areas, is that something you guys could give any color on, move away from breast, maybe?

Well, I think we've said before that our core competencies in image processing.

And we now have an AI platform that has the ability to provide detection software across the imaging modality opportunity and we're looking clearly in environments where there are lots of procedure volume and where using AI would be very, very helpful in detecting cancers as an example, but also validating normal exams as well.

So what we're trying to accomplish now is different than in 2D.

In the 2D world, it was all about how accurate could you find the cancers and how many more could you find when a radiologist use your tool versus when they would read independent of it.

What's interesting is in the mammography space, in a screening population, you only find 3 to 4 cancers per thousand.

So in essence, you've got over 95% of the exams that are normal.

And if they're not normal, they're negative.

And so what we're realizing with AI which is very powerful is to the extent that you can read a normal exam and not put marks on it when they're normal or negative is a huge contribution to the workflow and the productivity of the radiologist.

So what we're trying to do to broaden the value proposition is obviously increase the detection accuracy of the product contrasted to 2D.

But secondly, what we're trying to do, Frank, is to be able to really not mark the normals.

And so I'll just give you an example.

In the 2D world, you are lucky if 5% to 10% of the normal or negative cases didn't have any marks.

Very, very, low number.

There were false positives on the majority of the exams.

With the version of the product we're working on for the future with our testing to date, and there's a lot more testing to be done, we're actually finding that we are probably not marking as many as 60% of the normal or negative cases.

So as you think about broadening this portfolio of AI capabilities, it's not strictly about increased cancer detection, it's also about validating normals or negatives so that the radiologist don't have to spend a disproportionate amount of time in environment where the majority with the reading is negative.

So we're trying to make some contributions that are much broader than what we did in 2D.

And we're looking at other imaging modalities.

Simple examples would be chest X-rays.

There is enormous amount of chest X-rays that are done in this country each year at tremendously high volume, and a high percentage of those are normal.

So obviously being able to read chest X-rays and not mark the normals would be a real contribution versus trying to find a cancer as the only contribution you'd make to reading that exam.

So obviously, we're looking at places like that.

We're also looking at ultrasound.

And I do think in the area of breast ultrasound, as an example, we could make some real contributions.

And so we're making some investigations at the moment in those types of areas and other imaging modalities as well.

We already have detection software that is based in the CT world for colon cancer detection, and so we're trying to leverage a lot of the experience we've done detection software based on MRI.

So there's a long list of areas that are opportunities for us, and as we get closer and closer to concluding our second version of our AI product portfolio in the not-too-distant future, those are just some examples of places we're investigating today.

Okay, great.

It's very helpful.

And then lastly, just in terms of strategy if you just compare therapy to detection.

Should we be thinking therapy, is there any chance, is there any incoming interest maybe in terms of maybe partnering?

Or are you guys trying to climb back up the hill on your own on this one?

I think we'd love to partner with a strategic that has global reach.

When you look at the larger players in radiation therapy, they have what I believe is very, very complementary products to us.

And so in essence, most of the major players using radioactive source, they have obviously a traditional line of products and a linear accelerator-based environment.

They have isotope-based brachytherapy products.

And what we do with electronic brachytherapy or high-dose X-ray, we believe is a real complement to all of those large strategics.

So if we have the opportunity to collaborate, and we are certainly making that effort, we really believe we could provide a broader set of products to the market together.

And we also think in a lot of the emerging markets and developing countries where access to shielded vaults and linear accelerator-based technology is not that prevalent, our technology would be a great fit for certain types of cancers.

So yes, we would love the opportunity.

We're talking at ASTRO, as we do every year, to the major players, looking for where we can partner, both in terms of product complement as well as in terms of field resources.

We obviously can't put a direct sales force in all the major countries in Europe and Asia at this point, but most of the larger companies can.

So we certainly are very open and interested in working with those companies, because we really believe there's far more complementary opportunity between us in synergies than any sort of overlap in how we're treating patients today.

(Operator Instructions) We'll go next to Brian Marckx with Zacks Investment Research.

Ken, wonder if you can talk about the next-gen tomo product, and you alluded it -- to it in your prepared remarks.

But just kind of, I guess, on the reader study if there's kind of a status on that, when you think that might kickoff, and then if you can remind us how long you expect that, that would take.

Sure.

So right now, Brian, we're taking the same approach we took with the first version, which is that we would do a smaller, we call it a pilot study, to validate this product to essentially gain the experience and then power what we call the pivotal study, which would be the study we use for FDA submission.

So we have now finished, I would say, 50% of the pilot study.

We're actually in the middle of the 30-day washout, where the radiologist go off and forget what they read, to come back in a couple of weeks and read the balance of the exams.

And so at least what I would say from the first half of the reads that we've been able to examine, we're extremely encouraged with the product's performance.

But we need to get the rest of that study completed.

And in parallel, we are planning to do what we call the pivotal study, which would be the basis for an FDA submission.

And we're hopeful that we can have that started by December, which is only a month away.

We're very hopeful to have it completed and to be able to put in a regulatory submission late in Q1 of next year.

And so with that said, I think in my opening remarks, I said we hope to have something in the U.S. in the second half of the year.

And so obviously, if we meet our time lines, and again all that has to be determined for certain because there's a lot of steps we have to take, but the good news is we just did this a couple of years ago with the first version.

We've had good constructive meetings with the FDA to outline the study they were going to conduct with the second version.

And so with any luck, we're hopeful to have something in the second half, but we should hopefully, if things go according to our current plan, have a submission ready toward the end of Q1 or very early in Q2 of next year.

But again, the early results in the stand-alone testing we've already done has been very encouraging, and that's very, very important because when you think about us today, we have a very large opportunity in the near term working exclusively with GE.

However, GE only represents about 25% or 30% of the addressable market, both domestically and internationally.

So to the extent we can bring the second version of this product to the entire market, by orders of magnitude, increases our addressable market.

Ken, what have you found in terms of the typical length of the sales cycle with the current product in the U.S., that current tomo product in the U.S.?

How long is it typically on average, I guess, from kind of the first face-to-face until close of the sale?

What we're hearing, Brian, is that customers for the tomo system and the interpretive software often have budgeted a year in advance of their purchase.

And so, what we're seeing is that if the product fits into the budget, and again I mentioned earlier that's been a little challenging because our product for 3D is more expensive than our current 2D, and the Pristina product is more expensive than the SenoClaire that it replaced from a 3D standpoint with GE.

So it really comes down to how quickly can the customer get the additional funds they need.

And in some hospitals, in large radiology practice environments, they can do it pretty quickly.

And we've seen some immediate orders within weeks or months of having presented the product.

In other circumstances, they basically said it has to go into next year's budget.

And what they've done is they've added a license on the existing PowerLook servers so they can at least do 2D detection on this new 3D machine.

And again since the majority of the exam being done today are in 2D, that would make some sense for some of our customers.

And then those customers have said that next year with their new budget cycle, they'll have money to buy an upgrade to add the 3D capability.

So the sales cycle is really all over the map.

What I would say is, if there is not a funding issue, we can demonstrate this product and literally close a transaction in 30 days or less, not an unusual circumstance.

We get bogged down often quite honestly when the budget is tight.

And struggling for additional funds or the competition, which has happened a lot, comes in and low-balls the transaction because we're in there with a very differentiated solution.

No one else has 3D detection software.

And so ultimately, as you can imagine, the competition comes in with a much lower price to try to counteract that.

And depending on the what the customer's ultimate budget is, that may have an impact on what they buy and what kind of time frame the sales cycle ultimately develops to.

Ken, does that mean that there is potential for, I'll use the term bulk sales, it's probably not the best term, but in terms of when the budget cycles renew for some of these customers, have you had essentially orders that say when we get the budget approved, will pull the trigger?

We've had quite a few customers say to us that now that they understand the pricing, they will get it into 2018 budget -- capital budget.

So there's no question, we've had customers that have come forward and said this will get into next year's budget, I can't afford it at the moment.

So it's been a mixed bag of responses.

And I do think to your point as we get out to the mid -- to the latter part of '18, we should start to see some really nice momentum in terms of the mix of what's purchased being mostly 3D, if not exclusively 3D software.

And we'll see an upgrade market emerging for those customers over the last year that have -- not have the ability with the budget they had to purchase it.

So that's kind of a perfect storm for us because you're also then reaching into a period where the next version of product could be available.

And then your addressable market based on serving customers outside of the GE base could go up 2x compared to where it is today.

So I would hope by the latter part of next year, all these forces come together, which help us to continue to accelerate really, really strong growth in detection software.

And I know, Rich, reviewed it quite a bit earlier, but our detection software growth has been pretty significant this year and largely masked by the MRI software that's still in our base from 2016.

So we feel pretty good that as you take a look at our product revenue for detection as an example, year-to-date, minus MRI, we're are up 36%.

And that does not include a lot of those factors that I just mentioned that will kick in 2018.

So I think as '18 unfolds, detection and product-specific revenues could grow very, very nicely, as they're doing this year-to-date.

Yes.

I want to talk -- or ask a question about the goodwill charge to the Axxent business and why you decided to take it now versus a year ago and particularly in the context of what that implies, if it implies anything, in terms of your outlook for that business?

It would seem that you knew that there -- a year ago that there was obviously issues with reimbursement.

So just kind of in that context in terms of why you decided to take it now instead of taking it, say, a year ago?

Sure, Brian.

So when you look back on what's happened here with goodwill, the company did actually take an impairment back charge in 2015 when there was a significant change in reimbursement.

And that kind of set a new baseline coming out of 2015.

And what's happened over time here is that quite honestly although we have been experiencing growth most notably over the last 4 to 6 quarters as it relates particularly to subscription, the numbers that we've achieved have been below the numbers that were previously anticipated as we came out of 2015.

So we use October 1 as our annual impairment testing date and when we ran our numbers and went through our valuation exercise and experts, we ended up with an impairment charge.

And no, we do not feel like it's indicative of anything as we move forward here.

As the history would show over the last 4 to 6 quarters, we are showing slow and steady growth and we would anticipate that to continue as we move forward.

It's just that the numbers coming out of the old goodwill models weren't -- will keep falling short of them on a quarterly basis, which resulted in the impairment charge.

And that concludes our question-and-answer session.

I would like to turn the call back to Ken Ferry for closing remarks.

Thanks, operator.

In closing, I'd like to reiterate our priorities going forward: one, to maximize the success of PowerLook Tomo Detection software in the market while continue to make strong progress on the next version, which will significantly increase our addressable market upon launch, hopefully in 2018; two, to continue to substantially expand our customer base and treatment volume with the treatment of non-melanoma skin cancer; three, continue to expand our IORT and GYN customer base in global markets; and lastly, to manage expenses and cash use prudently to reach cash flow breakeven in the near term.

With these remarks, I'd like to thank you all for joining us for the third quarter call, and we look forward to updating you again at the end of 2017 when we report our full year results.

Good day.

Thank you, everyone.

That does conclude today's conference.

We thank you for your participation.