Why Artificial Intelligence can Improve Eye Care for more than 40,000 Americans Yearly with Dr. Michael Abramoff, President and Founder, IDx LLC

Thanks for tuning in to the Outcomes Rocket podcast where we chat with today’s most successful and inspiring health leaders. I want to personally invite you to our first inaugural Healthcare Thinkathon. It’s a conference that the Outcomes Rocket and the IU Center for Health Innovation and Implementation Sciences has teamed up on. We’re going to put together silo crushing practices just like we do here on the podcast except it’s going to be live with inspiring keynotes and panelists to set the tone we’re conducting a meeting where you can be part of drafting the blueprint for the future of healthcare. That’s right. You could be a founding member of this group of talented industry and practitioner leaders. Join me and 200 other inspiring health leaders for the first Inaugural Healthcare Thinkathon. It’s an event that you’re not going to want to miss. And since there’s only 200 tickets available you’re gonna want to act soon. So how do you learn more. Just go to outcomesrocket.health/conference. For more details on how to attend that’s outcomesrocket.health/conference and you’ll be able to get all the info that you need on this amazing health care thinkathon that’s outcomesrocket.health/conference.

: Welcome back once again to the Outcomes Rocket podcast where we chat with today’s most successful and inspiring health leaders. I want to welcome you to the podcast again and have you go to outcomesrocket.health/reviews where you could rate and review. Today’s guest because he is an outstanding contributor to Health. His name is Dr. Michael D Abramoff he’s the president and founder of IDx LLC, a company focused on bringing forth technology for diagnosis and they’re doing it in a very unique and innovative way. And so I want to do is open up the microphone to our outstanding guest who is also a professor of ophthalmology and Visual Sciences and really want to welcome you to the podcast. Dr. Abramoff.

: Thanks so much. It’s very exciting to be here.

: Maybe I missed something in your introduction. I love to give you the opportunity to fill in any of the gaps and have the listeners get to know you a little bit better.

: Yes. I don’t want to make it long so I’ll be brief. As you can hear from my accent I was born abroad in the Netherlands I came here from Amsterdam to Iowa. Now 15 years ago it was a very good choice. I was probably one of the world’s leading centers for ophthalmology and visual sciences. So it’s very exciting to have the opportunity to work here and collaborate here in my background. I was always interested in computer science and medicine. I actually tried to combine them both during my medical school where I did what a computer science degree into backgrounds I left to do computer science to mimic the brain and understand better how the brain works in Japan. From there I moved to France where I worked for several years. Understanding brain scanning with computers essentially image analysis of brain scans and back to Holland did my residency and ophtalmology became a visual retinal surgeon as it’s called essentially a retinal specialist. And during that time I worked on a Ph.D. analyzing images was the eye. So I’ve always been trying to combine these two interests. People always told me it was very exciting that it was trying to combine these with never really materialized to anything useful until I started thinking about how you can really employ computers. Machine Learning mimicking of the brain to you know better the outcomes for people with retinal disease and what systemic diseases that manifest in the retina. And so that led me to think that maybe analyzing images of people with diabetes to diagnose diabetic retinopathy can really be a great benefit to patient outcomes.

: Absolutely. Dr. Abramoff and listeners Dr. Abramoff’s passion for the intersection of tech and medicine has led to the creation of IDx and they’re really just you know one of the few companies applying artificial intelligence in a very useful way in medicine rather than just talking about A.I. they’re actually putting it to good use. Michael one of the things that I always like to ask our guests is a hot topic and what’s a hot topic that medical leaders need to be focused on today.

: I think there’s a lot of buzz about artificial intelligence or AI. Right now it’s sort of a fashion saying words last year. Very exciting few years earlier you didn’t want to talk about it. Twenty years ago was really exciting shot comes and it goes. I have a right now I think there’s opportunities that were never there before because of recent increases in the power of hardware the hardware and also better use of new and better improved algorithms. These other things were around for a long time but the learning algorithms really found their place when the hardware matched the demands of fairly high computational demands.

: Yeah and Dr. Abramoff one of the things I find happens a lot. People talk a big game on artificial intelligence and they don’t really do much with it. Can you explain to the listeners how you guys are leveraging it to create better outcomes.

: Yeah so let me start with what I see. Maybe sometimes it gets a bit muddled and people talk about all the excitement about AI AI and healthcare and then Sardasht. I think 3 users for AI healthcare 1 is for research where we tried to do discovery right. And they’re great tools for associating outcomes with. Maybe early events or things in the medical records or in images and it’s really useful research to discover new relationships and try to better understand them so we better understand disease. So there’s one thing and eventually hopefully research leads to better outcomes. But it’s sort of more far away than what you also have is something called assistive AI where you have a specialist or a physician or a provider essentially assisted by an AI but it’s still at the same level as the specialists. Let’s say you have a radiologists or an ophthalmologist like me and are looking at images of their patients and the AI algorithm helps them maybe pay more attention to a specific abnormality that it otherwise would not have found. Defensively dental specialists still makes the call whether to call something disease or write to say something is normal or abnormal. So that’s another helpful use for AI and definitely that’s more comfortable for a lot of people and what we do is different in a way again because we use it in an autonomous fashion which is you know remember autonomous driving cars where it’s really the algorithm that takes the clinical decision to AI that takes the clinical decision and the specialist is not involved anymore. There are still physicians or providers involved by communicating with the patients. When they went the test to communicate the results and explained you know what is next and make sure that that the things that need to happen with the patient indeed happen. And so we have great expectations that it will improve outcomes because one of the big grants of that is that primary care front lines of care is really where where patients will first go. And any time you move to a specialist diagnostic expertise that specialists office into primary care I think will benefit outcomes. So in IDx that’s what we do we we take specialist knowledge like me as a retinal specialist being able to determine whether a patient is diabetic retinopathy or not. And we take that knowledge and we now encapsulated in an algorithm make a system around it and it can talk about later. What I mean by the system because it’s more than just an AI algorithm and put it into primary care where the patients already are and are being taken care of and there help that provider to primary care provider to better manage the patients and also have to refer them to the specialist.

: So you guys are focused on the automation portion of the three that you discussed.

: Yes I do a lot of research as you know I’m a physician scientist or do little research which I mentioned first and it’s all very exciting and very soon as a route to improving outcomes is a long one. Assistive or others are doing probably with the expectation that eventually they will move to autonomous. We decided to jump ahead and just go straight to autonomous which initially you know gave some challenges. I have a nickname called retinator and I think that’s seven years ago.

: I love that.

: There and it’s facetious but it’s funny of course it is also serious you know and these algorithms really do what the specialist can. Is it safe for our patients that we really want that. And so initially you run into some of that but that’s over now. People understand that these these are homes for specific role and can be very beneficial for everyone patients healthcare assist insurers tax payers primary care providers and even the specialists like me now.

: And you know the implications potentially to physicians like you I mean you get pushback from them.

: Well that’s why it’s called a resonator. Many years ago but now it’s not the case anymore again. So we’re filling an unmet need with this specific diabetic retinopathy section and also with other detection products that we are in preparation in the pipeline because it’s important to remember there are more than 30 million people with diabetes in the U.S. and less than 50 percent yet recommended eye exam every year. That is so important that preventing blindness and visual loss every year about 20 to 25 thousand people go blind or lose vision because of diabetes which is almost entirely preventable if you catch it early if you don’t have a retinal eye exam we don’t get caught early and then it gets missed and then you get symptoms but then usually it’s too late and irreversible damage to the rest of retina so the whole point is a we need to find these people that every decision in early stage so we can send them to them. Eye care provided to retinal specialists like me American treat them. So right now this is happening with 50 60 70 percent of people with diabetes. So rather than saying well it’s replacing what dies as a retinal especially as the row is indeed bringing patients who needed care to the places where Medicare can be giving. But on the other hand not sending me a lot of patients and providers like me. A lot of patients where the retina has nothing wrong with it. It’s fine. They can just come back next year I don’t need to see those. As far as perfect for primary care to have that happen over there.

: It’s super interesting Dr. Abramoff I had no idea the numbers you know 20 to 25 thousand people get blindness per year. That’s just an unacceptable number.

: Well you know sometimes there are diseases and there’s nothing we can do very little.

: Right. That’s true.

: But in this case we know it’s preventable.

: It’s preventable.

: What to do. It’s just that patients tend to show how this goes is that right now how it works is you go to your primary care provider and diabetes can get your congregations. Everyone with diabetes knows it needs to look at your extremities your kidneys your blood pressure all these things are checked for. And then when they see something wrong they will either manage it better or they will send you to specialists for care for your kidneys for example. Yes. For primary care who so comfortable looking at the retinas of people with diabetes so they are supposed to refer these patients to an eye care provider. The problem is that they tried to do that but many times people it’s an appointment maybe three months away. That is a long travel in some places in the country up to four to six hours of driving. And so many times there’s a reason not to go or the kids or the child gets sick and that’s another reason not to go. But then they come to someone like a retinal specialist like me they spent almost half a day in the waiting room after traveling a long time. And then I have about a few minutes time for them to tell them that nothing is wrong with the retina. Well they spent all this time having a fairly brief encounter with me because I see many patients today and then get sent back and say Come back next year well next time they are asked a year later to make this appointment they will just think well why am I doing this.

: Yeah I would feel the same way. Like when I went there last time I was fine. So you’re taking this approach and you get your empowering the primary care physician to do the test on the spot.

: Exactly. And so it’s empowering for primary care physicians. And you know I expect it to be a major improvement in outcomes because again we condemn prominent cases of blindness and visual loss.

: That is so fascinating, Dr. Abramoff and love where you’re going with this a little bit earlier you talked about you were going to dive into a little bit more it’s not just the algorithm but the system. Can you tell us a little bit more about that.

: Yes so I think it’s important to understand that you know artificial intelligence is great and it’s a necessary tool but it takes much more to move into primary care. And I love to give the example of the clinical trial side that I can talk about the clinical trial a little bit later but we did a clinical trial in inside’s primary care all over the U.S. including one in New Mexico where the closest ophthalmologist is for maybe six hours away I don’t remember. Saw these people I’ve never been able to get to people with diabetes in that clinic. The eye care they need so desperately. So we came there. We put the system in place which is a robotic camera a computer with the room essentially and we train people with a high school level education with never taking a retinal image before so we were mostly desk staff and we train them up for hours and then we left. And from then on they were able to do the imaging and operate the algorithm et cetera. And so that’s why I called the system because it requires an operator strange in a certain way. And leverate became sorry a robotic camera and Algorithm to be combined and only done its work. If you just sent images to an algorithm that’s not enough. So it’s very exciting because of the results of the clinical trial are very exciting. As we mentioned in the press release. And so I can tell you that you know it’s better I mean because FDA is the ultimate also argued that many in Europe are cleared and there definitely is better than me as a retinal specialist in terms of directing disease. So here you go. You have a clinic in the middle of no. Well those people don’t see it that way but you know for us relatively from Iowa it’s in the middle of nowhere and no diagnostic capability and we came in with this AI based system. There’s this clinic far away from any eye care provider and we come in with the AI based system which is again an AI AI robotic camera and an operator will be trained in a specific way for hours with no end any expertise in eye imaging ever. And we leave for hours later and there is a diagnostic gap that we are very excited about is really good.

: Now that’s super super fascinating Michael and you know one of the things that I find is troubling is that you know the reason why these things don’t get detected is because it’s trying to do early detection and a lot of times when you do early detection especially with things that take a lot of time and you know waiting and waiting rooms you may not find anything the first time around. It’s a disincentive for them to go back. This is just a wonderful way to catch it up front.

: Yeah essentially you make the eye exam as easily doable as a blood draw or a blood pressure cuff blood pressure measurements. Yeah. It’s really to make it almost a simple thing in the place where where the patients are.

: Is super exciting. And with the growing number of adults and children not just adults with diabetes it’s becoming more and more of a necessity. So pretty exciting that your two passions finally came together in a way that could really benefit millions of people.

: Really cool to finally took a long time and find it seems to be happening.

: That’s amazing. How do you feel about that.

: Well very excited. The funny thing is you know for 20 years I’ve worked towards you know having a really exciting result of you know the people to trial and so what that was done immediately a you start working on Oh yeah. What happens when clearance and how do we actually get to patients. And so there’s so much because we’re the first right. No one has done any type of a clinical trial for FDA. And so because we the first we we’re learning everything as we go and we’ve develop you know one thing is there is no guidelines. And so maybe a bit about guidelines for self driving cars is Level 1 and Level 2 that’s actually the Institute for Highway Safety has developed. And so I was asked recently to write guidelines simply for autonomous diagnostic systems. Made diabetic retinopathy first. So we developed you know 10 pages of guidelines which are currently not one there have review and actually ever meeting Thursday night to try to finalize them so that there’s so much work to be done. You know how do you pay for this. Who gets to do this. How do you train people. There’s a lot of things to be done because I expect once we get clearance that there will be many other groups and companies trying to do this also not only in diabetic retinopathy but in many other fields of medicine. And that’s you know I expect that to improve the outcomes for many many patients which is the whole point is why we’re doing this.

: That is so awesome. Dr. Abramoff you’re just breaking ground here.

: Well yeah.

: For the capabilities of these technologies I mean that’s that’s pretty exciting. Congratulations to you and your team on that.

: Yeah, thank you was the team right. So no I think it is awesome.

: Absolutely. Now within the company obviously this is pretty exciting and the things that you’re doing maybe what’s one of your proudest medical leadership experiences are moments that you’ve had to date.

: As a clinician. It’s always the patients where they came to you with a problem that seemed intractable. Maybe someone else that looks at it. You diagnosed them you do several tests and you were able to find a treatment that actually works. And these are very graceful patients that are always very rewarding if you say someone’s sides by thinking about carefully and then taking the appropriate steps to manage it. So that’s you know from the clinical sites on the company’s side. Like I said the description of where I wasn’t involved going to New Mexico into to side but our team when there were four people installed it likely that there are sites around the country and that really made it real for me that there is a difference between saying you want to do something with an algorithm and actually implement it in places like that. That was to me so far the most exciting events.

: That’s awesome. And listeners. This is why I highlighted the fact that Dr. Abramoff and his team at IDx are actually doing things with A.I. not just saying things with AI and so it’s pretty exciting what they’re up to over there really looking forward to seeing what you guys come up with. And when it finally launches so let’s pretend you and I. Dr. Abramoff are building a medical leadership course on what it takes to be successful. It’s the 101 course of Dr. Abramoff. I’ve got four questions for you. Lightning round style and then we’re going to conclude with your closing thoughts. You ready for that.

: I’m ready.

: What’s the best way to improve health outcomes.

: I think it’s so important that you make it accessible. So I focused on making it accessible to patients to make it more affordable meaning lower cost or at least higher productivity and at least equal or better quality. So I think it’s essential and there’s a lot of ways you can do that. I happen to focus on artificial intelligence. But I think if if you start with that if you bring that type of care that is more effective more accessible and more affordable. That is when outcomes start to improve. Very different pathway. Many of my colleagues are focusing their wishes inventing or trying to invent and develop new methods of treatments and new methods of diagnosis. And those are as important if not more important. But there is also a way of bringing the scientific knowledge we already have about what needs to happen with patients and make it more applicable and more widely available. I think that’s that’s what I tend to focus in at least in IDx. Now I have research interests and do other things but I think it’s essential that we do that.

: Fascinating. And what would you say one of the biggest mistakes or pitfalls to avoid. While creating in the medical spaces.

: I think many times people get excited about technology or new protein or new treatments and try to make it work scientifically without realizing how it can work in clinical practice and clinical practice can be very different from a research environment. And so it’s so important I think every interaction with clinicians. So you know on both sides. On the side of clinicians and also on the side of researchers we need to talk as much as we can to each other so we understand what is possible and enhance and what is achievable and what is actually going to work in clinical practice because otherwise I think it will benefit patients in the long term.

: Excellent. And this question is is an interesting one for you because you’re the one that is creating the change but how do you stay relevant in health despite constant change.

: Well I never said that to be relevant. OK so for me it’s a risk and b I want you know I’m a doctor. I want the benefits. I will still benefit patients. I want to make the world a better place with at least get Sidek or we store side or maintain sides and people are at risk of losing it. And so that’s my motivation drive interest which is why you know I do it this way but it has to ultimately benefit patients and then so affected them that a leader is really to be honest very important to me as long as it benefits patients I try to step out of the way if possible. So I’m glad you know the team knows much more about this than I could collectively do know much more about it than I ever can.

: Yeah and it’s really interesting. I gave my thanks for sharing that you know at the end that is what we’re here for and to benefit patients. What’s one area of focus that should drive everything in a health organization.

: I think that gets back to your first question. Again there’s a lot of reasons to develop new treatments new diagnostics and it’s really important. But at least as important is bringing the things we already know and the things we already have. So as many patients as we can and so you choose either one or the other I think it’s hard to do both. And if you want to benefit the maximum number of patients trying to find ways of translating the things we know to patients and that may many times it will require something like artificial intelligence it’s just a great tool I think to do that. There’s almost always an expert there how to deal with a problem that a patient has but it’s you know that that is not widely available that is the difference.

: It’s that scalability and that ability. Yeah totally I was interviewing one of the folks from dignity health and he said it really well he said that innovation and health care oftentimes is just implementing at scale and you’re doing just that. Dr. Abramoff So congratulations to you. This has been really really interesting. Before we conclude I’d love for you to just share a closing thought with the listeners and then share the best place that they could follow you.

: I will just repeat what I said before if you want to benefit patients and you are excited about technology and specifically about artificial intelligence finds ways to make it work where it benefits the maximum number of patients and many times it will mean it needs to be autonomous and their take that jump to get away from assisting a specialist and make it actually work. It’s ultimately just as exciting as self driving cars are potentially huge and will have a huge impact. The same with autonomous diagnostics

: Phenomenal. Well Dr. Abramoff if this is a fantastic way to conclude today’s episode just on behalf of myself and the listeners just want to give you a big thank you for carving out time for us and educating us on A.I. and and they’re really neat things that you guys are doing it at Idx really appreciate it.

: Also Me. Thanks so much.

Thanks for tuning into the outcomes rocket podcast if you want the show notes, inspiration, transcripts and everything that we talked about on this episode. Just go to outcomesrocket.health. And again don’t forget to check out the amazing Healthcare Thinkathon where we could get together took form the blueprint for the future of healthcare. You can find more information on that and how to get involved in our theme which is implementation is innovation. Just go to outcomesrocket.health/conference that’s outcomesrocket.health/conference be one of the 200 that will participate. Looking forward to seeing you there.

Automatically convert audio to text with Sonix