“AI will not replace radiologists; radiologists that use AI will replace radiologists that don't.”
In the latest installment of SALT Talks: Pandemic Venture Investment Series, presented in partnership with OurCrowd, leading digital health start-ups provide a unique view into the revolution in artificial intelligence and how it’s shaping MedTech trends, and where investment opportunity lies. Zebra Medical provides automated, accurate AI imaging diagnosis, Diagnostic Robotics’ AI offers a triage and clinical-predictions platform, and BrainQ is developing an AI-powered electromagnetic field therapy to reduce neuro-disorder related disabilities.
AI technology can read scans and identify potential diagnoses before the radiologist reviews them. The introduction of AI to radiology will help address the problem of volume control in healthcare. This issue has been exacerbated during the pandemic where increased resources dedicated to managing COVID diverts time and attention away from other patients. “I kind of liken it to shining a flashlight in a dark room. We can see what's in the spotlight, but you're missing things around the edges.”
AI can process billions of data points to provide medical providers key insights to increase positive patient outcomes. This modeling is custom to each population and its history of chronic health problems. The pandemic has accelerated the adoption of digital health practices like telehealth, and has given faster rise to the need for AI population health technology.
LISTEN AND SUBSCRIBE
SPEAKERS
Ohad Arazi
Chief Executive Officer
Zebra Medical Vision
Yotam Drechsler
Chief Executive Officer
BrainQ
Maya Orlicky
Vice President, Strategy
Diagnostic Robotics
EPISODE TRANSCRIPT
John Darsie: (00:12)
Hello, everyone, and welcome back to SALT Talks. My name is John Darsie. I'm the Managing Director of SALT, which is a global thought leadership forum and networking platform at the intersection of finance, technology, and public policy. SALT Talks are a digital interview series with leading investors, creators, and thinkers. What we're trying to do during these SALT Talks is replicate the experience that we provide in our global conferences, the SALT Conference, which is to provide a window into the mind of subject matter experts as well as provide a platform for what we think are big ideas that are shaping the future.
John Darsie: (00:47)
We're very excited today to welcome you to the fifth installment of our Pandemic Venture Investment Series, where top entrepreneurs, investors, and business leaders dive deep into the challenges and opportunities arising from the pandemic crisis and discuss breakthrough technologies that are addressing issues ranging from coronavirus prevention and cure, to social distancing and food supply. This series is presented in partnership with OurCrowd, which is a leading global venture investment platform.
John Darsie: (01:17)
Today's episode is called Artificial Intelligence: Digital Health's Secret Weapon, and it features Ohad Arazi, the chief executive officer of Zebra Medical Vision, Maya Orlicky, the vice president of strategy for Diagnostic Robotics, and Yotam Drechsler, the chief executive officer for BrainQ. And the talk today will be moderated by OurCrowd Qure's managing partner Allen Kamer. If you have any questions for any of our panelists during today's talk, you can enter them in the Q&A box at the bottom of your video screen on Zoom. And with that, I'll turn it over to Allen for the interview.
Allen Kamer: (01:54)
Thank you very much John. This is Allen Kamer, managing partner of OurCrowd Qure, Israel's first digital health fund focused exclusively on digital health investments. I'm really excited to have this esteemed group with me today. As John mentioned, we're here to talk about artificial intelligence, digital health's secret weapon, and we have some really prominent companies and prominent people from those companies here to describe their weaponry, and how they're approaching the market, and how they're disrupting healthcare in a positive way.
Allen Kamer: (02:31)
And how we're going to start is I'm going to turn it over to each of the representatives on the panel to describe themselves, give a little bit of background about who they are, and then a little bit of background about their company. And then we'll go from there. So let's start with Yotam from BrainQ.
Yotam Drechsler: (02:51)
Hello everyone. Thank you for the intro. My name is Yotam, and I'm the cofounder and CEO of BrainQ. My background is in math, computer science, and cognitive science. I've actually worked for several years at the Boston Consulting Group here with Maya. So I'm very happy to see her with me again on this panel. And since 2016, I'm cofounder and CEO of BrainQ. BrainQ is developing a medical device that aims to reduce disability following stroke and other neuro-disorders. We do so with a wearable electromagnetic treatment, which is frequency tuned. And our novelty lies in the data science way that we retrieve the frequencies for the treatment. And this is where the AI comes in for us. That's in a very short.
Allen Kamer: (03:46)
Thanks, Yotam. Maya, tell us about yourself and Diagnostic Robotics.
Maya Orlicky: (03:51)
Sure, thank you so much Allen for the introduction. So hi everyone, very happy to be here. I'm Maya Orlicky, I'm the VP strategy and operations of Diagnostic Robotics. My background is biotechnology and biomedical engineering. I spent a few years, as Yotam mentioned, at BCG. Some time at fintech, travel tech, and then back to the digital health space. And what we're doing here at Diagnostic Robotics is we're developing medical grade AI products, for predicting both patient medical conditions, and guiding patients through their most appropriate medical journey. Our system was built using databases of tens of millions of EMRs, medical records, as well as tens of billions of claims data. And we support both healthcare providers, payers, and employers, by creating seamless data driven interactions, automate and optimize the patient journey through AI.
Allen Kamer: (04:50)
Super, I can't wait to get into this. Ohad from Zebra Medical, tell us about yourself and what Zebra Medical does.
Ohad Arazi: (04:58)
Hello everyone, my name is Ohad Arazi. I have the privilege of serving as the chief executive officer of Zebra Medical Vision. I've been in the medical imaging and health IT space for the past 14 years. I'm absolutely passionate about transforming healthcare with AI and medical imaging. Our company, Zebra Medical, was founded in 2014, and really since then we've been on a mission to transform patient care by teaching computers to automatically read and diagnose medical imaging studies. I'm sure you know medical imaging is already established as one of the most critical and influential domains in healthcare. Over 3.6 billion exams are taken across the globe every year, dealing with almost every type of medical condition. And the challenge with that is, with the continuous growth in medical imaging volumes and complexity, we're quickly reaching a human limit for effective interpretation of these images. So at Zebra Med, we're looking to empower healthcare providers to manage this ever increasing workload without compromising quality of care through the use of AI and machine learning.
Ohad Arazi: (05:57)
I think what's really unique about our approach is that we're the first company to use imaging AI to take on the challenge of population health, which is core to the triple aim framework and to value based healthcare at large. And our platform is proven to help assess and stratify risks for guiding clinical decisions, for interventions, for reimbursements, and most importantly, we're using images that have already been taken. You can imagine that dealing with population health has become even more critical since the outbreak of COVID, as the entire healthcare system has been really required to develop tunnel vision to deal with only what matters most. And as a result, many conditions that are non-acute are not getting diagnosed or treated. And I kind of liken it to shining a flashlight in a dark room. We can see what's in the spotlight, but you're missing things around the edges. And that's a very significant role that we'll talk about today, that I think AI can play.
Allen Kamer: (06:48)
Well thank you, all of you. I'm going to jump in and keep it with you, Ohad. We talk, the panel is titled AI: Digital Health's Secret Weapon. What's so special about your AI? What makes your AI really stand out, and how does it make an impact?
Ohad Arazi: (07:07)
First of all, we're looking to complement what radiologists do. When this industry broke out, there was a lot of conversation around, will AI replace radiologists? And I always like to say, AI will not replace radiologists, radiologists that use AI will replace radiologists that don't. And the reason for that is that the math is pretty simple, right? With this exploding volume in growth and complexity, if you just do very basic math, one radiologist on average reads 50 CT studies a day. Each study has at least 500 images. That's 25,000 images per day, which means that the radiologist spends roughly 1.2 seconds reviewing every image. So our AI solutions help radiologists by automating the interpretation of medical images for certain types of conditions.
Ohad Arazi: (07:53)
And the way that it works, very simply, is we receive a copy of all the scans that are taken at a particular hospital, and our software reviews the images and provides an initial diagnosis even before the radiologist has ever looked at the exam. And I think that complementary approach, the one that's really deeply embedded and truly seamless with a clinician's workflow, is the way to go. Because from my experience in health IT, it's as much a technology challenge as a change management challenge, right? And winning over the hearts and minds of our users, helping them bring AI and technology to standard of care, to the point of care, is often tied to really nailing the workflow and making it seamless for them to consume. And that's really one of the differentiators we put forward with Zebra Medical.
Allen Kamer: (08:38)
Thanks, Ohad. Now Maya, maybe you can tell us a little bit about your AI, and how is it used with physicians, how is it used in the field and in the real world?
Maya Orlicky: (08:51)
Sure, happy to. So I think our competitive advantage really relies on the fact that we have a very strong data sense capability, but it's also coupled with a very strong in house team of clinicians. So from a data science perspective, we really built our model based on tens of billions of data points in order to get the most accurate predictions, and improving our models at all times. In addition to that, we've also partnered with some of the world's leading medical institutions in order to make sure that we are really clinically validating our results. And we've started our work actually in the emergency department, where we think we were able to get the most clinical and rich data from a medical perspective.
Allen Kamer: (09:40)
That's great. And in terms of how that medical science, that medical practice, that knowledge base that you have in house comes into play in the product, can you tell us a little bit about what that entails?
Maya Orlicky: (09:56)
Definitely. So our work is essentially optimizing the user journey, and providing insights, both a decision support system for the physicians, and the medical setting navigation for the patients. So that's sort of one part where it really comes into play. And the other is all our work around population health management, where based on the data, we provide better risk stratification and interventions matching for different chronic populations.
Allen Kamer: (10:25)
Thank you. So Yotam, you talked about how your company is working with a specific patient subset and group. And tell us a little bit about, what is first of all electromagnetic field therapy, and how can AI be applied to that?
Yotam Drechsler: (10:48)
Right. So BrainQ is targeting the neuro-disorders world, an unmet need. Stroke, our leading indication, affects 15 million patients every single year. It's the number one cause of long-term disability. We are a therapy company, and maybe unlike Ohad and Maya, they use AI for diagnostics, which is I would say 95, 98% of usage of AI in the world is used for diagnostics. BrainQ is a therapeutic company. We use AI in order to distill biological insight that could not have been shown or revealed otherwise. It's often a matter of trust. So Ohad has mentioned before, it's a question about, would radiologists be replaced by AI yes or not? And I believe that we are getting more and more trust with AI. And BrainQ has taken it one step further. We have used AI in order to distill therapeutic insight from electrophysiology patterns, from a patient's brain waves compared to healthy individuals' brain waves, to find this kind of anomaly in patterns, and use it to direct the electromagnetic field toward the right systems.
Yotam Drechsler: (12:05)
To do so, we have accumulated a massive amount of electrophysiology, which we had to tailor ourselves, because this is virtually nonexistent data. And collecting this data in a very dedicated way allowed us then to use what's called explanatory machine learning tools. So without going into all the details, the first step is very similar to what typical machine learning does. We use supervised learning. We try to classify between different states. So it can be a patient that does a grip versus a patient that does a non-grip. And then we ask the algorithm, well why did you make something become a grip? Why do you think it is a grip? What features made you believe it is a grip? So we don't care about the classification, we care about the rationale behind the classification. And based on this rationale, we are able to inform then our treatment.
Allen Kamer: (12:59)
So that's very interesting. So it's a high touch point with patients. You work closely, you need to understand what's going on in each specific patient in order to make sure you can treat them appropriately. So how has your business been impacted during the pandemic? How has that really impacted your ability to connect with patients, to do research, to really progress as a business?
Yotam Drechsler: (13:33)
Is this addressed to me?
Allen Kamer: (13:35)
Yes, please. Please [crosstalk 00:13:36].
Yotam Drechsler: (13:38)
So I think that the first impact is not an AI impact, it's actually an interpersonal impact. As the CEO of a company, when the COVID-19 started, the world is shaken. And I needed to make sure that our speedboat is making it through this storm. It wasn't an easy one, to be honest. So I had clinical trials running in US and Asia. And the first thing that happened, and we realized is going to happen very soon, is a pause or a slowdown of all clinical trials. For a life science company, this is a very problematic situation. Second is you have your entire team under a lockdown here in Israel, and also abroad. And a lot of un-clarity from your stakeholders and so forth. So I think the first thing to realize, this is a lot about communication, a lot about making sure you are making the right judgment and the right calls in a relatively short time, and in very critical days.
Yotam Drechsler: (14:47)
Once we did get a good grasp of where we stand and how to take it forward, we also realized where the opportunities are, and they are. So in our case for example, BrainQ has developed a treatment for quite a while already, with the notion of what's called remote therapy. So the patients that we treat, actually have a very fragmented treatment pathway. A stroke patient in the US spends about four days in acute hospital, and then often he goes to a rehab center, and he spends another two weeks, and then he goes home, and so forth and so on. Several locations, several doctors, one patient. Every single treatment that's ever tried to kind of capture it by taking the patient, dragging him back and forth to hospital, always failed. We have had this notion for a while in our head, and we have been working on this IoT medical device. Again, a therapeutic medical device. Very easy to say when it comes to diagnostics, but not heard of when it comes to therapeutics.
Yotam Drechsler: (15:51)
And regulators and clinical and such were actually very cautious of making this move with us. They said, "Let's try it in clinic, and then let's realize if it works, and then we're going to go on." All of a sudden, now that the COVID is out, it's no longer a question of visibility of treatment. It's a matter of safety of the patient. Stroke patients are a risk group to COVID, as well as any other neuro-disorder other. They cannot be treated in hospitals. So having a remote therapy solution actually helps them engage into a treatment. So it's becoming an imminent need. Being able to leverage on this opportunity, accelerating the development of this product, and engaging with the right regulatory bodies is something that we have managed to do in a very short time.
Allen Kamer: (16:44)
That's amazing. That's amazing, and certainly good to hear that the remote capabilities are being accelerated by regulators. Ohad, over to you. The data about hospitals and some of their elective procedures, some of the other parts of their business that are non COVID-related are staggering in the sense of they're down significantly, as much as 60, 70% in terms of activities in some cases. How has that impacted you Ohad, and the Zebra team, in terms of the market's ability to work with you and to work with your products?
Ohad Arazi: (17:27)
Yeah, much like Yotam, we had to overcome initial challenges of kind of figuring out how to work in the new world, right? How to engage with hospitals, how to create mind share for them. They're very, very focused on one thing, and so a big learning exercise I think for our team culturally and professionally, but maybe to double click on what you talked on Allen, is the fact that especially as COVID broke out, many elective procedures were postponed or canceled. And we tend to think of elective procedures as something that's not needed. When I mean elective procedures, I mean mammography screening, I mean cancer screening that is deferred, right? So very, very substantial procedure types were deferred, and that really created a kind of a peak and valley environment, where we see now, especially across the US, where volume substantially dropped, massive pent up demand got built up and then that had to be dealt with in a very short timeframe.
Ohad Arazi: (18:20)
And to me, that really creates a very robust opportunity for AI to help normalize that curve, because of what I talked about upfront, of the increased volume and complexity that radiology has to deal with. Let's take the example of mammography. The outbreak of coronavirus, annual mammography tests during lockdown were postponed or canceled. I saw an amazing data point that every day from mid-March to mid-May, an average of 94,000 mammograms were deferred daily in the US alone. That means that over that period, almost six million patients experiencing increasing anxiety to be tested, and ran the risk of missing early detection as part of their annual screening during that period. And AI can solve that.
Ohad Arazi: (19:00)
And that same phenomenon is happening again now with the second wave, as the health systems need to refocus on COVID yet again. And so I think that the role of AI in being that aid inside the radiology cockpit, to help deal with peaks and valleys, with pent up demand that can build all of a sudden, and to allow the radiologist to focus at times on what matters most while AI can complete the picture and run routine mammo-scans, can do detection of chronic conditions and stratify the risk for them. I believe that's been a very important lesson, and I really encourage the health system, and we're trying to promote this with our vision, to make sure that the new normal is truly a different one, that the role of technology as an aid inside a clinical setting really does become different due to this compelling event that we're all experiencing.
Allen Kamer: (19:48)
And so does that mean that, Ohad, that the utilization of AI in radiology will likely build on this momentum, and that providers who historically may have been resistant or reluctant to change rapidly, now as they've gotten used to the new world order and gotten an understanding that the spike in demand may be coming in the short while post COVID, that it'll be more routine in their use of AI and the integration of tools into packs, or into other places in their workflow than they have been historically.
Ohad Arazi: (20:37)
I certainly believe so. I think it's causing health systems to retool, and they're needing to retool financially, because their volume, their inpatient volume has grown, which is a lower margin business. So their financial, the bottom lines are strapped, right? So that in and of itself is causing change. The workflow has changed, the peaks and valleys have changed. I think a good example of that is actually the role that AI imaging has played with COVID specifically. Because when COVID first broke out, most of the efforts to fight the pandemic were on testing, were on vaccinations or building immunity for it. There was very little to help the health system understand the potential for disease progression in patients, and really therefore help to create the best care plan.
Ohad Arazi: (21:17)
We know that one of the greatest drivers of coronavirus related mortality in the early days was the inability to detect early potential severe cases and provide critical care to those that needed it most. So healthcare providers need automated tools to support triage, and to determine how to best allocate ER capacity, ICU beds, ventilators, based on disease progression. And so the very first wave that challenged Italy or New York City truly served as a cautionary tale, where frontline workers had to make very difficult decisions on who they think had the greatest need to receive treatment. And then AI started to come in to play a role there, by analyzing CT scans. And our solution for this automatically detects and quantifies suspected COVID-19 findings, but more importantly, it provides a lung burden score, which really calculates the percentage of the affected lung volume, and enables better prediction of the trajectory of a patient with COVID as a decision support for the allocation of valuable ER or ICU resources.
Ohad Arazi: (22:21)
So to me, that's an example where a new need, which is understanding progression, not necessarily understanding classification, drove adoption behavior that we hadn't seen before. And to your point, I think that that new normal is going to look a little different in terms of the role that AI plays now, pre and post-pandemic.
Allen Kamer: (22:41)
Thanks, Ohad. Maya, talk to us about the emergency rooms, and talk to us about how, what you're seeing in the market, what you're hearing, and how that's impacted Diagnostic Robotics.
Maya Orlicky: (22:54)
Yeah, definitely. So I think we're seeing a lot of different changes, right? COVID had made a lot of impact on the entire healthcare space. I think on the one hand, it really was a huge accelerator for digital healthcare space, and really brought adoption of phenomenons that we thought would take decades in a matter of months, right? We're seeing telehealth is booming, and we're seeing really the need for remote monitoring. So I think that's sort of one thing that had a really huge impact for us, whereas the having the ability to really leverage AI in order to have the physicians focus only on what's most important, right? Only the life saving treatments, really the core of the physicians' work, I think that became very clear right? Whereas before, it seemed as maybe an added value, now it really became a necessity with the situation and everything that's been going on with the physicians. And we were seeing that both in the emergency department as well, right? In emergency department, optimizing the user's flow, making sure we identified those high risk patients and optimized the process, was really just, had a much bigger impact at this point.
Maya Orlicky: (24:16)
Another thing that we were seeing, so I think with the pandemic, there was an overall evolution. And we're seeing sort of the market react. So in the beginning, everyone was really interested in digital health with the focus on COVID, right? And we've collaborated initially with the Ministry of Healthcare in Israel, and with several states in the US and in India, to really try and provide risk assessments for COVID. However, as it progressed, we really understood that what's more important is thinking about the broader picture, looking at population health management, at the chronic populations, whereas you really need to find a solution to this new normal and make sure that they're monitored as well.
Allen Kamer: (25:01)
And so with these macro drivers you identified Maya, how does that change or alter your business plan moving forward? Do you have new-
Maya Orlicky: (25:12)
I think honestly... yeah.
Allen Kamer: (25:14)
Sorry, do you have new parameters of success, do you have new focus areas, do you expect things will return to normal post vaccine?
Maya Orlicky: (25:26)
First of all, I think that's the million dollar question, right? No one really knows what's going to be the new normal, but I think it will definitely not be as it was before. I think we don't see anywhere of this going back, the need for sort of digital elements and remote monitoring is going to increase. And I think we don't know if it's going to be exactly where we're at right now, but it's definitely going to increase dramatically from where it was pre-COVID, right? And so for us, what we're seeing is just everything is expedited in a sense, right? And some of the sort of chronic conditions, we're seeing an increase in those populations, right? So in our population health management, we're looking at behavioral health, which is really important right now. And [inaudible 00:26:18], and additional patient which are also in risk for COVID, so that has another sort of parameter which has an impact, and it just increase the need and the demand for that such monitoring.
Allen Kamer: (26:30)
Great, that's really interesting. And Yotam, on the clinical trials side, you mentioned that you had to pause some trials as a result of COVID. You mentioned some of the regulators' willingness to do some remote care. What changes in a post-COVID world for all of you?
Yotam Drechsler: (26:51)
Right. So COVID did change our plan, and honestly, I wouldn't expect this change in the plot originally. So we had too many indications we were running [inaudible 00:27:06]. One was in chronic tetraplegic spinal cord injury population, and the other one was on subacute ischemic stroke patients. And if you were to ask me a year ago, what is my first indication to market, I actually believed I'm going to do well on both. But my entire plan was develop based on the notion that spinal cord most likely going to make it faster to market. It's a niche market, it's a much kind of a slimmer penetration, and then go to the mass market of stroke. And this was my plan to the board, this was my presentation, the pitch that was there. And what happened is that spinal cord injury for patients is an elective procedure, as Ohad has mentioned before. Again, these are some of the most severe patients out there. They are paralyzed from their neck and down. Do they need help? They definitely need help.
Yotam Drechsler: (28:06)
But when it comes to hospitals' ability to treat them, even enroll them into a clinical trial right now, it just wasn't there. So things did slow down, and whenever there was kind of a slowdown in numbers of COVIDs in different states, okay, we had some more recruitment. And then it went down again. So things did slow down on that process. At the same time, we actually managed to complete our stroke study. And not just that, kind of get a two, three years jump, because our plan was to have a pivotal study in the US based on a clinic product just as a way to get to our ultimate product of one that could have a remote therapy. And all of a sudden, I don't need it anymore. So our plans have now changed to already engage with this remote therapeutics setting, and the doctors we work with, the sites, and also the regulators are much more adapted to it.
Yotam Drechsler: (29:10)
And we have already been experiencing this, kind of a move to remote therapeutics in the US, and it turns out highly successful. Again, things that were inconceivable just a few months ago, during the COVID we have taken our spinal cord injury patients and moved them home. And regardless of the efficacy side, which I cannot discuss that point, but patients were thrilled. They do not want to go back to clinic to receive the same treatment. Who wants to go in an ambulance back and forth three times a week if you can receive the same treatment at home? So these are, again, concepts that would've taken several years to mature, and now did mature.
Yotam Drechsler: (29:53)
I do want to make one more point. I think one of the challenges for a startup company in crisis times like COVID is also to make sure you don't de-focus. And you do, you adhere to your core values, and you make sure you stick to it. Because COVID will eventually, ultimately in a matter of months, maybe a year, maybe two, but we will overcome it most likely. And the startups, many of the startups I see around me and also us, we're established for a reason, and we develop often unmet needs, we develop therapies for unmet needs or diagnostics for unmet needs. Spreading around and placing too many times the COVID word in our value proposition also has the risk of losing the true value we're after.
Yotam Drechsler: (30:52)
So we do have to acknowledge it's COVID, we do have to take into account how do we speak to our stakeholders in times of COVID, but we also have to remember right, that whatever we target out there, most likely is still going to be a need. And not just that, it's most likely where we believe we have the best chances in order to I would say differentiate ourselves and maintain our competitive advantage. So for me, this was also an important lesson, not to go and develop my own COVID vaccine as well.
Allen Kamer: (31:26)
Right. Well there's certain some positive news on the vaccine side, so it's good that you've stayed focused on what's important, and it's really interesting to hear how you see an evolution in terms of the patients, in terms of what they want and how that's going to impact your strategy and your ability to progress post-COVID. Ohad, let me hear from you in terms of what you're projecting and forecasting as vaccinations start to spread hopefully in 2021, and the market begins to get back to potentially a new normal as Maya said. But tell me what you're thinking from a Zebra perspective.
Ohad Arazi: (32:21)
My view Allen is that this will drive a significant push, primarily in the US, from a fee for service model to a population health based model, or a value based model. Because one of the things that COVID really underscored is that kind of the cost and complexity to care for a patient needs to be viewed in its totality. And in the US historically, in a fee for service world, you're often looking at per procedure basis, right? And all of a sudden when you had to deal with very complex comorbidities that are tied to COVID or had to deal with a big influx of inpatients, it really stretched the finances of the health system, in particular because of the fact that we noted earlier, that the higher margin procedures were deferred. And all of a sudden, the hospitals lost a lot of their top line revenue. They were still doing the same number of procedures, even had more patients in house, but their revenue basis had substantially decreased.
Ohad Arazi: (33:13)
And so I think that's part of a broader trend that we're seeing in the US, of the move more towards population health. And I think it's also a theme in what Maya and Yotam talked about, because population health is often going to lead us towards a path where we can have early intervention, and that we can really link diagnostics and therapeutics, which I think is one of the key things that's emerging from each of the discussions we're having across our companies. And so I really think that this is a trend that we at Zebra Med have to understand how to capitalize on. It's been a big part of our strategy actually, moving from just dealing with acute care situations much more to population health.
Ohad Arazi: (33:50)
And I'll bring that to light for a moment of what that means. If I step off of a long international flight when I fly from Israel to the west coast of North America, which I often do, if I have chest pain and I've come off a long flight, I'll go to the ER and the physicians will want to rule out pulmonary embolism. So the radiologist will interpret a chest CT, and they'll look to either rule out or confirm PE. But what if it that same time that the radiologist is looking at the scan, an AI algorithm can run in the background, can identify additional findings that will have significant downstream impact on the patient's overall health like the fact that I might have subtle vertebral compression fractures, which are a leading indicator for osteoporosis, or I might have a buildup of plaque, of coronary calcium that is a risk factor for cardiovascular disease.
Ohad Arazi: (34:40)
And so the ability for AI to stratify the risks and to deal with my end to end health, even if it's not immediately, high degree of acuity, is I think where healthcare is going at large, again principally in the US, where it's been much more focused on treatment and less focused on prevention. And again, that's one of the main roles that I think AI can play, and maybe also Diagnostic Robotics and many other companies that are now looking at diagnostics with a broader lens of the complete picture, the complete comorbidities, the risks, that these patients bear, recognizing that even things that don't manifest right away can often have prophylactic treatments that we can deal with preventatively, and we can drive better outcomes as a health system in this post-COVID environment with greater adoption of technology and AI.
Allen Kamer: (35:31)
That's great. And do you see those population health trends emerging in other markets globally in short order? Are they advanced in certain markets before the US? What's the forecast there?
Ohad Arazi: (35:50)
Yeah, they're certainly more advanced in single payer markets. A really good example of that is [inaudible 00:35:55] Healthcare. [inaudible 00:35:56] Healthcare already works with us, with our both coronary calcium and our compression fracture product, because it's a payer provider. It holds both ends of the stick, right? It owns the provider and it owns the payer, the overall responsibility for the expense and the accountability for dealing with the outcome for that patient. We see that in many single payer systems. I'd say the UK and Australia for example have very progressive bone health programs. And bone health is just another disease. It's the most preventable cause of fractures, osteoporosis is. It impacts nearly half of women over the age of 50, and costs $52 billion to treat. And so it causes, just in the US, more than two million cases of broken bones annually.
Ohad Arazi: (36:35)
But it's also highly treatable if detected early, so if you can increase the detection rates for osteoporotic fractures, then you can get much better outcomes. And single payer systems have recognized that for years. The UK, the NHS is really outstanding in adopting this. They have a network of fracture liaison services, basically bone health clinics, that help support the older population with fracture prevention, and that's catching on rapidly in Australia. So certainly single payer jurisdictions I think are further along, but they're still working within the confines of cost saving.
Ohad Arazi: (37:08)
I think in the US, the shift from volume to value will also change some of the top line projections for the providers, because they realize that actually if they're intervening early and owning the end to end care for that patient, they can get not only better clinical outcomes, but also better financial outcomes. And those financial drivers I believe will drive adoption substantially of population health solutions.
Allen Kamer: (37:31)
That's great. So here's a question that came in from one of our participants, and it's for each of the companies, and Maya, we'll start with you. The participant wants to know, what are specific milestones that you're aiming to achieve, either in terms of revenue or clinical milestones, or even AI milestones, that need to be reached for a sustainable success over the next year or two?
Maya Orlicky: (38:00)
Sure. So I think I can respond to each of those sort of fields, right? So I think from a clinical perspective, that's where we've started, right? We've really started with validating our predictions clinically. However, there's always this sort of a specific milestone on a per-modal basis to really prove that we have better results, right? And I can share an example in the population health space, which is we started with the risk stratification score, and then from there we really saw that in order to make an impact, we need to also look at the type of intervention provided by the care management teams, right? In order to really impact the bottom lines. And so what we're doing now is we're also incorporating interventions matching for specific populations. So that's just one example, to show how we keep on adding sort of more specific detailed milestones, in order to show that we're improving clinically.
Maya Orlicky: (39:00)
With respect to the other components, I think there's always the matter of adoption which is really key I think in our field, right? And there we have sort of metrics in order to make sure that we're working both with payers and providers, with the care management teams, and with the users themselves, just to make sure that we're increasing adoption.
Allen Kamer: (39:22)
Great. Yotam?
Yotam Drechsler: (39:24)
Right. So if I have to summarize it to one key milestone in the coming year, then it is launching our [inaudible 00:39:33] study, the equivalent for phase three for stroke in the US. Most likely Europe as well with some of the top sites out there. This is a very unique study. So it comprises this, our therapy that has been already evaluated in previous studies, but it blends in the remote therapeutics, a very unique setup that I don't think anyone has ever tried before us on a large scale study for an unmet need. So making it right, making sure that we launch it with the right technology on our end, with the right design, with the right doctors, it's kind of a once in a lifetime opportunity to really make a dramatic change into this world. So kicking it off in the right way is the number one milestone for us in the coming year.
Allen Kamer: (40:32)
Right. Ohad?
Ohad Arazi: (40:38)
Allen, for us it's really scale. Our company was founded in 2014, we've been in this business for a while. We're I think really we're the trendsetters, and kind of pave the way for much of the imaging AI industry. And now as we continue to mature, our focus is scale. I think I shared with you upfront that we're on a mission to transform patient care by teaching computers to automatically read and diagnose medical imaging studies at scale. And to us that scale, today we have to the order of hundreds of thousands of patients that go through our system every month, and our goal is by 2022 to have millions of patients go through our systems per month. And so we're really looking to build that scale by first, continuing to sell directly to providers.
Ohad Arazi: (41:20)
We've also substantially increased our resale network, and have really been working on OEMs, meaning embedding our software into assets that are being used by radiologists or by imaging users, like the imaging modalities or the pack systems, the systems that are used to interpret the exams. Because we know that that will drive adoption, that will drive scale. So really for our mission to be meaningful, we need to touch millions of patient lives every month, and that's really what our team is working towards. We measure that metric, we talk about it all the time. We have a dashboard that shows how many lives we're touching on a daily basis, and so we're continuing to galvanize the team around that. And that's really our biggest goal in 2021, is to continue to really scale up this company, increase our presence in North America and globally, and touch more and more patient lives every day.
Allen Kamer: (42:06)
That's terrific. So last question for everyone as we're pushing up against time, it's a general question, but as we talked at the outset, this is about AI as being digital health's secret weapon. I'd like to hear from each of you what needs to happen overall for AI not to be so secret, for patients to feel comfortable, for physicians to feel comfortable, for overall AI to become widely adopted and understood as being beneficial in the delivery of healthcare. And we'll start with you, Maya.
Maya Orlicky: (42:46)
So I think one of the things is really around, and we've touched on it briefly before right? But around market education. So I think we're working with a very traditional market, and I think it's a gradual process to really understand first of all the impact, and sort of the goal of AI within [inaudible 00:43:05], I think Ohad said it as well, right? We're not trying to replace the physicians, we're trying to supplement them, optimize whatever we can in order to support them, so they can really focus on the essence of their job. So I think that's sort of one thing which, it's a gradual process to really get the entire healthcare community to really understand that.
Maya Orlicky: (43:29)
And then there is just the area of just AI explainability, right? Which is really sharing as much of the results as we can in order to have the physicians and the healthcare staff informed, and to be able to make the right decisions based on AI, but really have the data that they need in order to make the right treatment decision.
Allen Kamer: (43:54)
Great. Yotam?
Yotam Drechsler: (43:56)
Right. So as Maya said, and I think the overall notion is all about trust. And when there are [inaudible 00:44:07] partners, you don't want to go too fast. You want to take it kind of step by step, and build up this trust. For me, the partners are, just to make sure our technology companies that penetrate healthcare system, and there is the healthcare stakeholders, whether it is doctors, whether it is the medical centers and so forth. I do believe that this is where you build trust. AI, in essence it's also a buzzword. To some extent, I often believe that we kind of use it too much. I mean, AI is a tool to achieve a goal, and it's a way to use data science in order to interpret something that you maybe could not have been doing, could not done differently.
Yotam Drechsler: (44:57)
So when you go in AI and you use it for your pitch to investors, I think it's very important to realize the differentiations of what you have. And by the way, AI has been democratized by its tools, so almost everybody can do an AI today, so this is not really the differentiation today. It's really something that we use in order to help technology solve problems in a more traditional system.
Yotam Drechsler: (45:29)
So for us, I'm seeing two things. A, create this trust between startups, and I would say entrepreneurs with this medical system, and do it with communication and transparency in the most profound way. Second, maybe consider not to use too much of the word AI. Rather focus on the value you can bring and the kind of tools that you use, because some of it is just about... this is the downside of a buzzword. And thirdly, I think that time is often the most critical factor, and time, between when AI was introduced several years ago, I think a lot of time has passed. A lot of these values that we are talking about right now are almost taken for granted. I mean, our radiologists today know what AI is. There is no more... it's not scary anymore, okay? It used to be scary, a few years ago. And the boundaries can just extend and extend, so this is why we kind of use it these days for explanatory tools for therapeutic insight, but this is just the very beginning of the journey.
Allen Kamer: (46:48)
Thank you. Ohad?
Ohad Arazi: (46:52)
I'd really like to echo what Yotam and Maya talked about. To me, AI is still kind of going through its cycle of moving from hype to reality. And our role collectively as innovators in this space is to move from being experts to being partners. We have to be partners to the health system, AI is a partnership tool. And again, to me, the ultimate partner is about also changing our view on AI from the inside out to the outside in. I think historically when it started, it was driven by innovators that had a vision, and they were telling the market that vision, and it felt a little bit like a solution looking for a problem. And now when you hear about each of our companies, we're very, very focused on finding solutions that are valuable, and pain points that the health system has, and real needs that when users leverage AI to solve them, they can't live without them.
Ohad Arazi: (47:40)
And I think that's our role, is to continue to evolve as partners, build that trust like Yotam said, and really switch our focus from thinking about kind of the solution, the glitter of AI and what it can do, to really getting into the trenches and understanding, where is pain in the system? Where are needs that are unmet by existing technologies, existing workflows, and making sure that we tailor our AI solutions to solve those problems. That is really what creates stickiness, that's what creates adoption, and that's what will allow our broader sector I think to mature, and unlock the promise that it has, and the value that each of these companies can bring to the world.
Allen Kamer: (48:18)
Well thanks Ohad. So Ohad Arazi, CEO of Zebra Medical, Yotam Drechsler, CEO of BrainQ, and Maya Orlicky, VP of strategy and operations at Diagnostic Robotics. I'd like to thank all of you for your insights and comments, and educating us about the market, about AI, what you do and how you're impacting patients and patient care. I'd like to thank all of you for joining us, and remind everyone that the recording and more information is available at the SALT website, and hope to see you again sometime soon. Thank you for your time, and thank you for participating to our panelists.