Good day, and welcome to the ENDRA Life Sciences’ Fourth Quarter and Full-Year 2018 Conference Call and Webcast. Today’s conference call is being recorded. At this time, I would like to turn the conference over to Mr. Chris Tyson, Managing Director at MZ North America, ENDRA Life Sciences’ Investor Relations firm. Please go ahead, sir.

Good afternoon. I’d like to thank you all for taking time to join us for ENDRA Life Sciences’ fourth quarter and full-year 2018 conference call. Your hosts today are Mr. Francois Michelon, Chief Executive Officer; as well as Mr. David Wells, the Company’s Chief Financial Officer; and Mr. Michael Thornton, the Company’s Chief Technology Officer. Francois and Michael will provide a business update, which will cover product updates and operational milestones, while David will discuss the financial results. The company filed its Form 10-K for 2018 this afternoon and a press release detailing these results crossed the wires today and is available on the Company’s website endrainc.com. Following management’s prepared comments, we will open the floor to questions for those of you who are dialing in for today’s call. Before we begin the formal presentation, please take note of the Safe Harbor paragraph that appears at the end of the press release covering the Company’s financial results. And that any forward-looking statements that we make only apply as of today’s date and are subject to inherent risks and uncertainties including those described in the Company’s SEC filings and should not be unduly relied upon. Except as otherwise required by Federal Securities Laws, the Company disclaims any obligation or undertaking to publicly release any updates or revisions to any forward-looking statements. We would also refer you to the Company’s website for more supporting industry information. At this time, I’d like to turn the call over to Francois Michelon. Francois, the floor is yours.

Thank you, Chris. Welcome everyone to ENDRA Life Sciences’ fourth quarter and full-year 2018 conference call. We laid a strong foundation in 2018 and we're riding excellent momentum into 2019. I'll start by highlighting our top five accomplishments in 2018 and how they've advanced ENDRA’s goal of bringing our Thermo-Acoustic Enhanced Ultrasound also known as TAEUS to market, the clinical goal of helping assess and monitor the earliest stage of liver disease, which affects over 1 billion people globally In 2018, our top accomplishments were: first, we renewed and extended our partnership with GE Healthcare, the global ultrasound market leader; second, six of our patent applications were issued by the US PTO, growing our IP portfolio to 41 assets by the end of 2018, comprised of defined, filed, issued, and licensed patents; third, approval was secured from Health Canada to begin the first in human feasible study with our TAEUS reference system that was designated as a class two device for the purposes of the study. Mike Thornton will provide detail on the study in a few minutes. Fourth, we expanded the pre-commercial market awareness of our TAEUS technology with ENDRA Booth at several key radiology and hepatology meetings, and a growing social media presence. And fifth, and finally, we further capitalized the company with approximately $10 million extending our operational runway and eliminating debt from our balance sheet. Before digging into details about some of these accomplishments, it's important to provide some context and reaffirm ENDRA’s goals and strategy. Our overarching goals are to broaden access to better health care while driving shareholder value. To achieve these goals, our strategy involves three core elements. First, introduce new point of care technologies that increase the utility of ultrasound and we will do this initially by proving out market adoption of…

Thank you, Francois. Hello everyone. We have a lot of exciting activities at ENDRA in the areas of product development, innovation, and intellectual property. But I'll begin with the clinical programs that explain what we're trying to accomplish with our first in human clinical feasibility study in Canada, what we've learned so far, and what lies ahead. I'm happy to report that the study is on track and we are preparing to report findings at the end of March. For those of you who are new to ENDRA, let me briefly review the roadmap for clinical studies and how they fit within our product development and review the four key objectives of the study. The Canadian feasibility study is the first in human use of TAEUS technology through collaboration between ENDRA and scientists at the imaging labs at the Robarts Research Institute in London, Ontario. The study utilizes our TAEUS development platform system that we call the reference design and targets healthy volunteers with ENDRA team members directly involved in data collection and analysis process, so that we can learn directly and relay ergonomic clinical workflow and signal processing feedback to the StarFish commercial product engineering team. ENDRA’s engineering team will continue to be involved in data collection for one or two additional studies in the feasibility phase of development as it is critical that we incorporate the maximum amount of learning into our early commercial product design. After this feasibility phase with ENDRA’s direct involvement data collection analysis, ENDRA will transition to supporting clinical studies during the collaborative product testing phase currently planned for Q2 of this year. And then later this year, move to independent investigator clinical product evaluation studies of the fatty liver application with little to no ENDRA involvement other than user training and standard product…

Thank you, Michael. 2018 was a lot of hard work. Thank you. I will now provide a summary of our reported full-year 2018 financial results. Revenue for the year ended December 31, 2018, totaled approximately $6,000 compared to approximately $350,000 for the year ended December 31. 2017. Revenue earned in 2018 was a result of service revenue for a Nexus 128 product line. Operating expenses increased to 8.9 million for the year ended December 31, 2018, up from 4.8 million for the same period in 2017. The increase in operating expenses year over year was primarily due to research and development costs incurred for a full year of development for our TAEUS product line, and general and administrative spending including costs associated with an increased headcount and legal fees. Note that our fiscal year 2018 results included non-cash expenses, comprised of approximately 1.4 million for equity issued for services, approximately 700,000 related to the amortization of debt discount associated with the convertible note and approximately 300,000 related to the impairment of inventory, held for our now discontinued Nexus 128 product line. Our net loss for the year ended December 31, 2018, was 9.7 million or $2.16 per basic and diluted share as compared to a net loss of 5.4 million for the year ended December 31, 2017. During the full year 2018, we used approximately $7.7 million in cash in our operating activities, which again was due mainly to the continued development of our TAEUS product line. Our spending remains on budget and on track with our internal projections. Our cash as of December 31, 2018, totaled approximately $6.5 million as compared to approximately $5.6 million as of December 31, 2017. The increase of cash as of December 31, 2018, compared to December 31, 2017, is the result of the gross proceeds of approximately $10 million from capital raises throughout that year. This included two common stock offerings and one convertible note placement, whose participants included new institutional funds, accredited investments in ENDRA’s management and board of directors. The common stock offerings were raised through an effective Form S3 shelf registration statement that was filed last year. In the first quarter of 2019, we filed a new Form S3 shelf registration statement to extend our access to the capital markets. Note that we also filed another S3 that covers common shares issued should any of our warrant holders exercise their warrants which are traded under the ticker symbol NDRAW. This was a required filing for compliance. As before and in summary, we believe that the combination of our asset light operating model and continued effective and efficient use of cash will position ENDRA to first commercialize our TAEUS liver product in the European Union and then in the United States as scheduled. I will now turn the call back over to Francois. Francois?

Thanks, David. And thanks, Michael. In closing, as we move towards the close of the first quarter of 2019, ENDRA’s position to deliver key milestone for our fatty liver application, including; number one, securing our ISO 1345 Certification, supporting our CE application in Q2, the first in human study report findings at the end of March contributing quantitative results, safety and human factors supporting our CE application, and the controlled launch of our TAEUS liver device in Europe this summer. ENDRA is riding strong momentum in 2019, we have an improved balance sheet, human studies are underway, necessary foundations for a strong CE application are close at hand, we've kicked off an artificial intelligence collaboration, and we're partnered with a market leader GE Healthcare. We look forward to updating investors on our Canadian human study in the weeks to come. At this time, I'd like to open up the call to questions from our listeners. Operator?

[Operator Instructions] We'll take our first question from Kyle Bauser with Dougherty and Company.

I'll start off with the clinical trial. So based on how quickly you're able to enroll the first 25 patients, it seems like enrollment in subsequent studies and beyond the 50 will allow for continued refinements without many delays in your timelines. Can you talk a little bit more about internal plans to continue building clinical evidence beyond this initial CE mark application and how large these trials might be?

Sure. And let me give a high level and then turn it over to Mike for more detail. As Mike pointed out, and as you reaffirmed Kyle, yeah, this Canadian study at the Robarts is only the first of several sites where we will be testing our system. And in this case, in the Robarts Institute, this is our reference engineering system. This is not the final commercial product, which is benefiting, if you will, from all the learnings that we described earlier on this call. We have paperwork in process at two other North American sites that we will announce in Q2 outside of the Canadian study. We believe that between the Canadian study and these two additional sites in Q2, we will achieve approximately 200 subjects scanned versus MRI and potentially other tools that will give us, we think, the minimum critical mass to develop that clinical value proposition, which is so important to our controlled launch mid-year. Our goal, frankly, is to continue through the process in the US as well as with the early adopters in Europe. And as Michael pointed out, gradually become naturally less engaged in the data collection and to hand this product off more independently to our collaborators and investigators in Europe and elsewhere. And so just to give you a sense of our personal or our internal targets, we really want to get to 500 subjects minimum scanned and published on before the end of the year. And we think that that critical mass is what we need by the end of the year, as we enter in the US. Michael, if you're on, perhaps you can give a little bit more color to that?

Yeah, sure. Thanks, Francois. Again, I apologize for the technical connectivity issues earlier. So, in this feasibility study phase, we expect to do one or two additional studies. And those would range between 50 and 80 subjects each. As we transition to the product testing phase where we move to US and potentially European sites, those studies will be on the scale of 70 to 80 subjects. That's what we're currently discussing with the sites and then ultimately, when the system is in independent clinical product evaluation, that will be up to the investigators as they will be initiating the studies rather than us taking the lead in organizing those studies. I hope that –

Does that answer your question?

Yeah. That's helpful. And then, Michael, you mentioned in the prepared remarks, a little bit about being able to go back and rescan patients, compliant with the trial design. I would love to hear a little bit more about that and what you're finding and just opportunities you had to go back once you've refined the algorithm in the system?

Yeah, sure. And that's a great question. An important function of the feasibility study, there are actually two really important functions. As we move from in vitro bench top testing with phantoms that are anatomically accurate to actual patient in vivo scanning, we want to see what changes. Obviously, we don't have breathing artifacts on the bench top, the temperature of the materials is different and that's one important role. The other one is just to understand what constitutes a bad measurement. So we spent quite a bit of time looking at data that looks somewhat different from what we do on the bench top. It has all the characteristics of what we do in the laboratory, but with the added features of respiratory motion and ribs and just the variation and human anatomy from different size subjects, I can tell you that just looking at the coded data, obviously everything is anonymized. But we see the BMIs and actual weight. We have everything from 100 pounds subject all the way to 250 pound subject. So that gives us a great ability to understand what the types of artifacts and nuances and features in the data are available. So that's really what we're doing in this feasibility phase along with collecting quantitative data. Does that answer your question?

Yeah, that's helpful. And just on the commercialization front now, if we fast forward to later this year, following the anticipated CE mark, can you talk about the controlled launch with respect to the number of facilities you might target and any feedback you might already be hearing following your participation in international conferences over the last year?

Sure. So you're absolutely right, those early adopters that we've been meeting and cultivating, engaging, as I mentioned in our prepared remarks, are the initial targets. They've been selected because one, in some cases, we approach them and they were very receptive or in other cases, they approached us, expressing real interest. So there's a natural shared interest. These are leading people in their field, in radiology and hepatology. As I mentioned, they’re in key markets, while we certainly aspire to growing beyond the core EU markets, it's important to have seeds planted with influencers in large markets that have a real interest and a strong infrastructure and a pedigree of publishing high quality research. So without mentioning how many specifically, I can tell you, it's a good handful of people that are leaders in their field are ready to take the product, wanted CE marked and are eager to start publishing and using it. And those are really the bedrock if you will. That's not to say that we won't make this available obviously beyond that group, but we want to make sure that we get it into the hands of those people. Commercially, to give you a sense of tactically, what we'll be doing and again, I don't want to describe too, too much from a competitive sharing point of view and I'm committed to give more detail over the course of these calls, but really getting our systems in Europe working with GE using the clinical value proposition that I mentioned, which will be developed from these clinical studies in the US and in Europe, engaging and leveraging these key opinion leaders, and then running tactical programs to build awareness, and that can include everything from lunch and learns, targeted one to two hour sessions in key markets with 15 to 20 clinicians that we have in our database, bringing them in to a show and tell where you scan and demonstrate the product and from there, Kyle, gleaning interested parties who want a demonstration of the product at their clinical facility. I've done this for most of my career. And so while it's not rocket science, it's really important to get that sequence of tactics well mapped out and executed and I don't think it is rocket science, but it does take organization and all the elements that I mentioned. So I hope that gives you a little bit of a sneak peek into what we're thinking of doing.

We'll take our next question from Brooks O'Neil with Lake Street Capital Markets.

Good afternoon, guys and congratulations on all the progress.

Thank you.

So I was hoping I don't think I heard you talk very much about what the pathway and potential timeline is for regulatory approval in the United States. But could you share just a little bit in a kind of a big picture sort of way about what the outlook is here?

Sure. I'll give an intro and then Mike, add any color that I may have missed. But following our submission of the CE application in Q2, we see ourselves leveraging that content and filing the 510(k) for a class 2 device in the US mid-year. We think based on the average turnaround cycle for class 2 applications of around 140 days, there's the potential to get the 510(k) approval by the end of the year. We think that we've unfortunately learned our lesson from Health Canada that that is probably the optimistic scenario, but very comfortably and I think much more realistically having it for the very beginning of 2020 and that would be the timeline, Brooks. Is that helpful? And Mike, is there any additional color that you'd like to add?

Yeah, just that we have a targeted date in late May for this CE mark technical file submission and then we'll obviously, as that's being reviewed by the regulatory bodies, we will shift our focus to reformatting that, which is essentially what the process is to the 510(k) submission, as Francois indicated mid-year.

Right. That's very, very helpful and frankly quite exciting in my opinion. Secondly I'm curious and it's mostly because I don't know that much about the ultrasound business itself, but I recognize GE is potentially significant partner. Are you thinking that you primarily work with GE or other potential partners that might offer complimentary or different capabilities to what GE offers?

Sure. Great question. First, we're good partners. We respect and have chosen to work with GE and they graciously decided to work with us. And so we honor that commitment and we're excited to work with them. I will say, technologically, we're agnostic to ultrasound. So while we're working to finalize the product and plan the commercial roll out of the product with GE, technologically, there's nothing to prevent us from working with others. But we're partners and we want to make this work with GE. Now, I think more interestingly, to answer your question, if you look at it from another perspective, we think there's a very natural opportunity to partner complimentarily with pharmaceutical companies in the liver space, which would complement our partnership in the ultrasound space with GE. So certainly an attractive scenario, Brooks, would be, we're partnered with the ultrasound market leader GE, focused on radiology, hepatology and those markets and we reinforce that with a partnership either with an end OEM, if you will, of a pharmaceutical company or potentially a high quality CRO, who's sort of at the intersection of that. Beyond liver, you may have heard me mention, some of the other applications for tasks, such as this real time visualization of tissue temperature change during thermal ablative procedures for cancer, cardiology, and elsewhere. There are a host of completely separate companies that I won't name here, because I don't want to infer that we have discussions going on. But there are a host of companies that our listeners can research that are focused on energy based surgical tools that use heat or cold to affect tissue in a vast range of clinical applications. And we think that that's another very interesting potential partnership that would not conflict with GE and we're keen to explore those as well. I hope that's helpful.

Yeah, that's really helpful. I think that's quite exciting. So thanks for mentioning all that and keep up all the great work.

Thank you, Brooks. Operator?

And we'll take our next question from Paul Bienstock with National Securities.

Thanks for the great update. One question I had is you mentioned a rollout of the pipeline for liver drugs. Have you approached or thought about approaching some of the pharmaceutical companies as part of their phase 2 or phase 3 trials to use your device to sort of help better quantify the effects of their drugs?

Yeah, that's a great question. Obviously, some good sensitivity that I have to be careful of there, but suffice it to say that our participation in these four major conferences, liver and radiology last year, certainly exposed us to several OEMs in both the device and in the pharmaceutical space. And I think there was genuine curiosity at a minimum, and in some cases, outright interest. I think that will accelerate, as we report out on this first in human feasibility study. So I think that that's a natural fit. We're certainly very open to that. But I also don't want to infer or overstate the maturity of any discussions with a strategic, because these things take a long time, but we're certainly very open to that. I hope that's a fair and honest response to your question.

Perfect. And then my second or last follow-up here is you’ve talked about workflow with [indiscernible] if I'm pronouncing that correctly. I'm curious what sort of things have you learned actually being in the clinic as compared to being in the layup. I'm just curious any kind of insights that you've kind of pitch for.

Well, I'll give you the lay person’s description. And then I'll turn it over to the brains of the outfit, Mike, to give you more detail. But, obviously, as you can imagine, working with either animals or phantoms, as they're called, models of tissue in the lab is very, very different than when you're applying a product to a patient who is moving, breathing, has a different anatomy, has bones and some are fat, some are thin. So it doesn't sound like a big leap. But really, as Michael pointed out, we've learned a lot about where to place our probe. We assumed it would be intercostally between the ribs on the right side of your body where the liver is, but frankly, we weren't really sure we could get a measurement there until we did it. And so as Michael pointed out, as part of our first objective, we've understood that the clinical positioning of our product aligns with existing application of ultrasound to abdominal and liver application. So we're not creating a new workflow, which is great. And I don't think that should be minimized because should be a great technology, but it has to be used in a weird way. And that's already an impediment to adoption. So we're using it and getting results in a way that is currently used today with other products, namely ultrasound. The other things that we've mentioned on the call that Michael pointed out is we observed the sonographers using our products and you have to understand it's a probe that looks like an ultrasound probe to some degree that is placed alongside the GE probe on the outside of the subject, the human subject and just seeing how the sonographer uses the product, because it's a very tactile kind of medicine. It's not like you put someone in an x-ray, they stand there, you aim it, you push the button. Ultrasound is very manual, very tactile, people are applying different pressures, angles of the ultrasound probe. And so we wanted to observe very carefully how experienced sonographers and radiologists do it. And what we noticed led us to making some changes for the commercial product. As Michael pointed out, we've put in some indentations and grip lines, we've reduced the size of part of the probe by, I believe, about 15%. All of this benefits the commercial product that we will be launching mid-year. So I hope that's helpful. Like those things are not nice to haves. They’re so critical to the commercial success of the product. And we're really paying a lot of attention to that. In addition, I underscore in addition to the quantifying measurements that everyone is so interested in. But we want to emphasize, there's so much more to learn in this first study that is benefiting the commercial design of the product. And, we're leveraging all of it. I hope that's helpful.

[Operator Instructions] We do have another question from Michael Brcic with National Securities.

All right. First question is the initial study of 25 people. You said that, something about, they had like low fat levels. So you weren’t really able to pick up much on them?

Well, more than that, no, it wasn't quite that. It said the number of people we scan would not be representative of the normal distribution of fat in a population. And so while we might have people with a lot of low fat and get a successful signal, our interest is to correlate our measurements across a distribution of very low fat to average fat to high fat patients. And that's really the reason we, the core reason we expanded it. Is that helpful? And Michael certainly is on the line to explain it further if needed.

I'm happy to add some color to that. Yes. So just looking at the published statistics on the incidence of fat liver disease, we see publications quoting general population varying by region, of course, between 25% and 40% of the population, where 4% liver fat by MRI is considered a lower normal. So we had only seven people out of 25 above 4%, and three of those people were just barely above 4%, like 4.1, 4.2. So in our targeted range of sensitivity, which would be above 7%, we only have four subjects and for that reason, the investigators that we collaborate with decided to expand the study to get at least 10 subjects above 7% fat. We have plenty of very healthy, young, highly active study subjects and we'd like to get more to fill in the database points that would better mirror the population distribution of liver fat.

Got it. Okay, so it does work. I was just wondering, I was hoping that it just wouldn't limit the scope to a certain group of people. Because that takes me to the next question which is you mentioned in your prepared remarks, there was no current test to gauge live effect. No effective test, I guess.

No practical test. So, certainly, liver biopsy and MRI are actually quite accurate, but we can tend, for reasons of cost, access, surgical risks, that they're not practical, both on the front end for assessing fatty liver and doubly so Michael for monitoring the disease every 6 to 12 months going forward. So that's really our competition, if you will. We want to do something non-invasive cost effectively and safe.

So on that, what would it take for your test to become the standard of care for, if that's the right term, for that testing? Do you have to do something different?

Well, no, no, I think that's a great question. And I think first of all, clinical evidence, is key, both for the controlled launch of the product and as we gain traction, and I believe in time that as the clinical data builds and we share it accurately through our own means and through our collaborators, that we're able to influence advocacy groups, the American Liver Foundation and others that this is really a tool of importance and I think from there, there's an advocacy from these groups for new guidelines, which eventually we believe in the route to reimbursement is the path that I've used and others to secure positive coverage in terms of reimbursement. And we think that that process of data first, building a strong base, advocating the key clinical groups make it a standard of care in time, but it takes time. Michael Thornton, any additional thoughts on that?

Yeah, sure. So obviously, it has to work and has to be compared against the clinically accepted gold standards. That's number one. The other thing and this circles back to the idea of ergonomics is usability. So, we're providing a bigger probe than they're normally holding with ultrasound, because we're combining two different devices into one. And there's a bit of a learning curve and it's really incumbent on us to make it as easy to use, not induce use injuries or and keep it light so that people actually use it. It's one thing to make sure it works. It's another thing to make sure people use it. So we're very sensitive to that. And that's a key piece of the feasibility studies.

And Michael Brcic, another element which Michael Thornton just prompted is, yeah, the ease of use, the clinical data, but also the economic access to this product to make it a standard of care globally, it has to be deployable as ultrasound is and affordable. And our goal is still to bring this product to market as an end user price in the $40,000 to $50,000 range for the hardware element with software and we think and continue to hear back from clinicians that this is half to a third of the price of a new ultrasound system and is really quite attractive for the kind of increased utility that we could bring to market. So I think you've got the data, you've got the advocacy from the clinicians, you've got the work flow and usability of the product that Michael Thornton mentioned. And then you got the economic value proposition and deployability of the product. I think all of those things come together to eventually make it a standard of care. Is that helpful?

Yeah. Two more. This should be a quick one. It's a class, have you gotten a class 2 designation which is low to medium risk? What about it -- assuming class 1 is low risk, what about the product or what does it have to do with ultrasound would make it a class 2 instead of a class 1, if a class 1 is the lowest?

So maybe, I can take this question. Class 1 is very narrow scope. So these would be things like band aids, and also I think the limitation is you can't put any energy into the body that isn't gravity. So, an IV drip would be okay, but anything that puts any form of energy in the body immediately goes to class 2, even if it's not harmful. So in Europe, class 2 is split into two groups, A and B and one is low risk and the other one is moderate risk and we're pretty comfortable with our position as class 2. In the US, it's just class 2, if that’s helpful.

And I was just going to add that while we're very encouraged that our investigational device in Canada was determined to be a class 2 device for the study, that's not a guarantee that the product would be deemed such, but we do think that there's a natural sort of influence because that class 2 in Canada was derived from all the safety tests that we did, that we would naturally submit as part of the CE and FDA applications that we think it's a very good precedent. But I also want to be clear that it doesn't sort of automatically translate rubber stamp of class 2 everywhere.

The next question I have, because I've seen this happen with the new products that come out and some of the animals put in some pretty big numbers straight away on the sales side, but I'm assuming that with, right now, you're talking about a good handful of parallels and they publish, do you see a pretty slow ramp up at the beginning until the published material hits critical mass and then maybe more of a hockey stick type thing.

Yeah, I think that's a fair assumption. And so this controlled launch is, it's typically called in the industry is just that it's laying those critical seeds with those key opinion leaders and those influencers to really scale and influence the market more broadly. And so I don't, I think that it's critical that people understand having the CE mark and the ability to sell the product doesn't mean that we put the product in the hand of some unknown distributor or put it on Amazon and the thing sells itself. This is a clinical device, it's something that has to be published on and to your point, I think the balance of 2019 will be real, the product will be commercial, the hand product will have units on the ground and it will be, I think, a very careful introduction of the product with a much stronger ramp up in 2020, both because of the critical mass of data we will have accumulated through the balance of the year, Michael, but also because we will then have the US market come online on the heels of Europe. And I think that that's a natural process. I don't make any excuses. I want this product to be successful and I know that if we just put it in the hands of a no name distributor, that's not how we grow a business. We get it into the hands of key opinion leaders, influencers, they go and talk about it and then 10 people say, I'd like to have a demonstration of this product and then 100 people do that and then 1000 and I think the people that understand this space, the hardware, software, medical device space combined with people like myself and my team who have done this before understand that that's a natural ramp up to a successful product. I hope that's helpful and I think you stated it well, so thank you.

Couple of financial real quick ones. The convertible, what's that like and the interest rate and is it a pick or is that a cash?

Yeah. This is David. The convertible notes converted at $1.63. The interest was paid in cash and it's since been converted in November as a result of our raise -- of our common equity raise then.

Okay. So it's no longer outstanding?

No. It was issued in June and retired in November.

Finally, OpEx obviously went up a lot because of research and all that sort of stuff. Do you see that coming back down or because you're prepping for the launch and we do more studies, is that OpEx level going to stay around here or do you even see it going up.

Yeah, the OpEx went up in ‘18 over ‘17, simply because we did our IPO in May of ‘17, and therefore really didn't have a full year spending. So from a run rate standpoint, it was pretty close to for ’18, was pretty close to ‘17 on a monthly or quarterly basis. While the spending is changing relative to the lesser spend on development, we then are looking at spending, of course, you want to see us spending on the commercialization, of course, the continued development of an intellectual property. We don't make -- we haven't provided guidance, so I think I can kind of stop there and say, we're changing our spending. I don't think you'll see dramatic changes in the amount of spending.

I think that's a fair qualification, David. Yeah, it's really, I think, shifting gradually from engineering to commercialization in a natural way. But we continue to be extremely light and careful with our spending, and you're not going to see a huge increase in G&A outside of headcount and things of that sort, which we think are key.

And you're not -- we're not going to see a burgeoning or growing balance sheet either in terms of a lot of fixed assets, which is something else, is kind of defined as the asset light approach.

We will take our next question from Bill Morrison with National Securities.

So just following up on Michael's question about your go to market, once you get past these handful of key opinion leaders, what I say is to look like looking at a multiplier effect. Do you think you would get from that and over what rough time frame do you think?

Yeah, yeah, I think to give you the broad strokes, getting these human subjects and clinical data to support the initial key opinion leader adopters is also the best way to engage the GE salesforce. GE is our partner and they're committed to and have already introduced us to several GE radiology customers and they'll continue to do so. But, I worked at GE for almost eight years and GE sales reps have got a lot of products to sell. So what's the best way for me to get the attention of a GE rep, that's with these name brand KOLs in their markets, in their countries, using the product that they can reference, it's by having this increasingly dense and rich pool of human data. And I think that that along with the influence direct clinician to clinician from those original early KOLs is the other multiplier as you mentioned it, that the GE sales team will become more and more engaged. We're planning on hiring three to five specialists in Europe and in the US to drive that. I mentioned that in my prepared remarks and I think that I want to be very clear that we're not just handing it over to GE and hoping that they sell this for us. We're going to make darn sure that this thing gets sold with GE and with our own efforts, but I don't want to build out a direct sales force that would just kill my overhead. So I think that that measured approach of reaching critical mass of data, getting the key opinion leaders in Europe and then back here again in the US when we come back are the combination of factors that make a successful 2020 in terms of commercial mass. I don't know if that helps answer some of your question, Bill, but I hope so.

Yeah. I'm just trying to get a feel for how you're going to penetrate the market, are these key opinion leaders in slightly different segments or different regions or I mean, do you have somewhat of a knock-on effect from individuals or you’re all kind of in the same group?

Well, a couple of things. One, we've built up our own base of key opinion leaders and GE has their own targets and I would say there's a lot of overlap, but it's not 100%, which is good, it enlarges the pool. GE is king in radiology, and that's where most of the liver scans with ultrasound are done. So they're a natural partner and they’re a natural call point, they’re customers and a large installed base of customers that are using GE equipment that we could go and approach. They sell a lot of ultrasounds that we could associate our product with as a bolt-on accessory. But to answer your question, the KOLs are in both clinical disciplines, radiology and hepatology and I think that in time, as I mentioned on an earlier call, the radiologists need to learn a little bit more about liver disease and what they could scan for, which they can't do today because our product doesn't exist and the hematologist have a real interest in diagnosing this disease. But they don't all use ultrasound because you can't do what we're talking about yet until it brings it. So we think that there's a natural opportunity to go out to key opinion leaders of our own finding and of GE’s [indiscernible] in radiology and hepatology, collaborate through the GE sales reps and opportunistically with our own specialist sales teams, approach those hepatology clinics that aren't using ultrasound today and say, hey, we'd like to introduce you to this technology and potentially with our GE partner sell you a low priced ultrasound and an Endra device that quantify fat in the liver. We think that that combination of segments, channels and products is a great combination, but we're going to start with radiology and GE and we're going to work closely then into hepatology. I hope that gives you a little more color, Bill.

Great. Perfect. One last question for David. What's the share count all-in now of the warrants that were exercised.

The fully diluted, excuse me, the primary share outstanding is 7 million for 22642.

We do have one final question. It comes from Bruce Jackson with The Benchmark Company.

Wanted to get some color on the AI collaboration with Western University, especially if you could talk about the developmental timeline and what are the steps that you need to go through to get to a commercialized product?

Sure. Michael, if you're still on, I think that’s your –

Yes. Of course and thanks for the question. So we're really excited about AI opportunities because we develop a lot of data and developing a lot of data is what you need for training networks, for machine learning and developing AI analysis networks. The collaboration basically kicked off in January when we announced it and it's scheduled for a whole year. Now, we have a number of projects. One of the kind of low hanging fruit things is, we have some steps that the operator needs to do with the conventional ultrasound image in picking some layers of tissue from the conventional sound to guide our measurement. Those seem to be things that are very automatable by AI. And that's one of the things we're starting with. And then, of course, we just like to see if it does better than our system model of estimating fat, really two pieces. One is workflow and the other one is just improving accuracy and precision. And we've already started some of the data collection. We're starting with bench top studies because they're very easy to control and we know the right answers and that's a good way to train, just to see how capable or how applicable these AI methods are to what we're doing.

Okay, that's helpful. And then I know this is like looking pretty far into the future, but in the magazine, for example, we've got computer assisted diagnosis or CAD with the imaging. Is that something that you could potentially get into further down the road?

So quite a bit down the road when we look at PMA and being designated a diagnostic that might be applicable. But as Francois mentioned a few times, we're targeting class 2, which would not make it a diagnostic. It's actually just a tool to aid the physician. So we think that some of these AI methods would come in before we would make the shift to a class 2 and a PMA and billing code and all the kinds of things that we want ultimately.

There are no further questions in the queue. I’ll turn the call back over to you for any closing comments.

Thank you. Thank you, Tom. And thank you everyone who called in today. If we weren't able to address all your questions, please reach out to our partner MZ Group. Thank you all for your time. Many of you have been on this call now for over an hour and I hope you hear the energy and commitment from the management team. Look forward to having a good solid update later this month and appreciate everyone's support and continued interest and we wish you a good end to your day. Bye-bye.

Thank you. This does conclude today's teleconference. We thank you for your participation. You may disconnect your lines at this time and have a great day.