From Missed Diagnoses to Life-Saving Alerts: How AI is Helping Doctors Detect Structural Heart Disease Before It Turns Deadly

Good afternoon everybody. Kevin Stevenson here. Thanks for joining me on today's episode of I Don't Care with me, Doctor. Kevin Stevenson. You know, I always have to have some sort of a Baylor tie to almost all my podcasts. You guys who've listened to me now for, can you believe six years I've been doing this? That's amazing. But you know, I love the green and gold. We always have to get Baylor integrated in this. So today's guest, we didn't start that way but Don's got I got a little Baylor ties so we're gonna go with that in the beginning. Hey, today's guest, Don Fowler. Don is the president of Echo IQ USA. Don, welcome to I don't care. Kevin, thank you so much for the kind introduction and it's a pleasure to be here. You know, I gotta give Don, you know, Don just moved out of the great state. You know, I totally understand Don moved to be closer to kids and forget the kids, it's the grandkids, right, Don? I mean, that's the main thing. Absolutely. So blessings for you up in the frozen tundra of outside of Philly, but know you're always welcome back here in Texas. Well, thank you, Kevin. And I do have a tie, as you had mentioned. I'm still one of the, what we call ambassadors for Baylor, University. I had the good fortune and opportunity to be asked to be part of their mentor program for their, school of professional selling, which for me was something that I never knew existed. My son transferred from Syracuse University down to Baylor. And during that transition, I had the opportunity to come out to the university and met with a lot of the business leaders and found out that they had a program called Professional Selling. I, at that time, was the CEO of Toshiba, was very interested in finding out what this all meant. And before I knew it, was one of the ambassadors to the program. So I've been doing that for about six years now, and it's the most rewarding, fulfilling thing I have the opportunity to do. It really is a great opportunity to be in front of these young students and really help and coach them as they start their professional careers and just an absolute delight. So while my address may be Philadelphia, my heart will still be in Texas. Always appreciate that. And hey, thanks for your work with Baylor. Okay. I got to admit this. I'm so old that I was at Baylor when the professional selling program started back in the eighties. So yeah, I just dated myself, you know, but, but anyway, what will it make you feel better? Somebody introduced me the other day as having over four decades of experience in health care. Went, oh, I turned another decade. That's wrong. I told people there was no child labor laws when I started, but nobody believes that anymore. Yeah. When they start throwing decades around, you know, Dom, we've been doing this for a long time. For sure. Hey, you mentioned Toshiba. Yeah. Talk to my audience a little bit about, you've got an amazing background. Share with them a little bit about your time at Toshiba and Siemens and some of the other things before you got to Echo IQ. Absolutely. Kevin, I was blessed. I had aspirations being a doctor, like I think so many folks do. Medical school had a different option for me. They said no. But but I still had a passion to be close to the industry. And I wanted to figure out how to do that. And through through a lot of good luck and good fortune, the folks at Siemens gave me an opportunity that, you know, I never thought I'd get coming right out of college. And I was able to have some incredible mentors at Siemens, was put in positions, not necessarily promotions all the time, but positions to get me ready for the next challenge, the next job. And and I spent twenty seven years there. It's still the core of my being in the health care space and just a wonderful company. But within that transition of of businesses, sometimes we get reached out by by recruiters. And a recruiter called me up one day, twenty seven years into my career, and said, we've got an opportunity that you're on a short list for Toshiba to become their CEO in the United States. And it was a hard decision because I had a great career and still have lots of friends at Siemens. So I didn't leave with animosity. I left because there was another opportunity. And I was given, I was humbled to be given the opportunity to lead an organization in the US. People always said to me, you know, what's it like to be a CEO? And I told them, I said, you know, the responsibility is staggering. And it's not because of what we have to do. It's the lives that we're entrusted with. Right. And I think as a leader, you wake up every day and say, boy, don't screw this up because there's eleven hundred people counting on you. And I think that was the pressure that that the job comes with. You know, what we get to do and the people that we get to interface like yourself and some incredible doctors and so forth. That's just that's just gravy. That's the wonderful part. It's that responsibility of making sure that people are entrusting their lives, their careers and the stuff that you do and the decisions that you make. So I had a blessed career that for thirty plus years, I was with some very, very large organizations that were doing some incredible work. And I had the opportunity while was at Siemens to live overseas. I lived in Germany for three years and ran a global sales and marketing arm, the magnetic resonance business. So it was wonderful. But after thirty plus years, sometimes it's time to step back and say, what's next? What are you good at? And what do you like to do? And through a networking opportunity, I had somebody reach out to me that I had met in Houston, Texas, and that he had a colleague of his that was living in Australia, that was a chief commercial officer for a company called Echo IQ. And they were interested in commercializing a product here in the United States and when I'd be interested in helping them grow this business. Taking a look at that, it was a pretty quick decision on my part. Okay, so now that you're here, let's talk about Echo IQ. I mean, because you and I have talked about it before professionally, some months, maybe almost a little over a year ago, but really want to hear kind of that initial product and where Echo IQ has gone since then. Sure, that's great. Thank you. So Echo IQ very simply put, we are an AI enabled decision support tool for cardiologists to help them in the diagnosis of structural heart disease. So that's what we do at a high level. How we go about doing that becomes the secret sauce and the magic of all these things. But when we first took a look at structural heart disease, we said, well, there's a lot to do with structural heart. So let's start first with severe aortic stenosis. And we started there because it's probably the most common of the valve diseases. One in two people don't even know they have the disease. It's asymptomatic for a very long period of time before it becomes symptomatic. Unfortunately, the outcome for untreated severe aortic stenosis is very poor. It's about fifty percent at two years. So, it's worse than a lot of metastatic cancers like prostate, breast, ovarian cancer, colon cancers. And it's a tremendous expense on the health care systems as well. There's about eleven and a half extra days for patients that go untreated at a cost of over four billion dollars So there was a need. In addition to that, it's a disease that sometimes goes underdiagnosed. And it goes underdiagnosed not because doctors are bad. They're far from it. They're incredible. But these are hard cases to identify sometimes. They don't present themselves like textbook cases of severe aortic stenosis. There's a condition called low flow, low gradient or paradoxal low flow gradient where you have these cases of aortic stenosis that that have some discordant information. And it causes a doctor to sit there and pause for a minute and say, is this severe? Is it moderate? I don't know. Right. Because there's too many conflicting things. So we were able to do is build an A. I. Tool that actually assists the doctor in those what I call fringe cases, those difficult cases that there and says, look, we look at this the same way every time. And we compare this to a database that was built to be the largest database of echocardiographic measurement data tied to mortality in the world. So we've got a whole bunch of science and we've got a lot of mathematics in our favor, and we do it all in three seconds. And we get that information to the doctors. So that's really what we do that's different than, some of the other AI tools that are out there. And by the way, there's a lot of good ones. Ones. So it's not disparaging comment, it's just the differentiation. No, you're right. And when you and I talked about this some time ago, AI was so foreign to us. And so I think I know you were a little bit ahead of us. And so, that's why things just, we just couldn't move forward with you guys. But now looking back, we've added some AI capabilities to MRI and a few other things. And I finally broke down and said, you know what, just because you're a little bit seasoned in your career, find out more about AI. And so I started doing it and really have started using it for other things. But let's talk a little bit, there's a lot of seasoned guys like us in the healthcare industry that have a hard time really kind of getting their arms around So what are some of the challenges in adopting new technology, particularly in health care, but just in and around the AI space? I think there's a few things. First, there's a lot of AI fatigue. As you said, everybody that's walking through your door now is talking about AI. So as you're trying to ingest this, you have to figure out where does this fit into my current workflows? How do I use this? How do I enable this? Even if you're gung ho and say, yeah, I think this is the right tool for us, there's still those questions that need to be answered. So you have to do your job. You have to do your homework at understanding your client's workflow. You have to understand how they diagnose today, how they treat patients, how they triage patients. And then how do you fit into that solution so that you're not upsetting their workflow? Because I can have the greatest tool in the world. But if I tell you, have to change everything you do in your hospital, you're going to throw me out as quickly as I I pay. Right. So that's the first challenge. The second challenge and a rightful one is cardiologists in particular want to understand how I do this. Right. They don't want to know about a black box that information goes in, information comes out, but they want to understand what takes place while that's happening. So for us at Echo IQ, what we try to do, and it's hard, and I will tell you, we'll be the first people to tell you it's hard to describe neural networks to somebody, But to understand the process of how we look at things, we talk in terms of phenotypes, for example. So we don't sit there and say this patient has severe aortic stenosis. What we say is this patient has the phenotype probability for severe aortic stenosis as high, medium or low. That's how doctors think. Doctors think in terms of phenotypes, they don't just sit there and go, Oh, Kevin, you have. They start to talk to you. They start to diagnose you. And then as you're talking, they're building this phenotype in their mind about what you may or may not have. We do the same thing. But what we do, which is different, is instead of just looking at the valve, we look at the entire heart. And the other thing I should have mentioned early on is we don't look at the images themselves. Images sometimes vary in their quality, patient body habitus, sonographer skills, all of those things combined for sometimes not the best images. It happens. So what we do is we look at the measurements that come from those images and then build a multidimensional view of the heart so that we can build this phenotype of a disease. So not only are we looking at what's happening at the atrium, but we're looking at the ventricles, what's happening in the septum, what's happening at systole and diastole, pulmonary circulation. So we look at one hundred and twenty measurements of the heart and then make a determination of does this patient have the phenotype probability? The other thing that I said in our passing comments early was that we had the world's largest database of echo measurements tied to mortality. And one of the things that we found when we began to map the phenotype probabilities, we compared them to the all cause five year mortality for patients that are in guidelines. And what we found is when we say somebody has the phenotype probability is high but don't meet guidelines, patients have almost the same exact outcomes over a five year survival curve, those that are in guidelines. So that tells you, you know, we can't keep these patients too far away from the structural heart teams, right? We can't wait for a year or two years because sadly some won't come back. So, and then the other thing that it says is those that are in medium also don't have a great outcome. So this becomes a wonderful triage tool, right? So you think, how do you use this? Well, we could certainly help you with those fringe cases, those difficult low flow, low gradient cases, great. But we can also help you identify those patients that probably need to be seen on a more frequent basis, those patients that may need a follow-up CT scan, those patients that, you know, maybe that low flow and you want to do a debuted study. Right. So there's a lot of things that we can help offer. And I think that's part of the challenge that health care has today is you're seeing more and more patients. Patients are getting older. The disease is getting more prevalent. And what makes it hard and as somebody who has elderly parents still, thank God, they think shortness of breath is a condition of old age. You know what today, dad? Well, when I walked up the stairs, I have shortness of breath. Well, that's not good. Well, I'm eighty nine. Of course it's not. It's normal. It's not normal, but that's what people think. So we have to do what we can to help from a triage standpoint as well, identify these patients as early as we possibly can. Sure, absolutely. Yeah, as somebody who just went through the dreaded heart cath about four or five months ago, you know, I'm a lot more attuned to my body now. You know, surprised I can't do what I did in my 20s. Well, but, and so it's really interesting that there are other mechanisms for me to stay on top of my healthcare. And I think, you know, in my learnings about AI and seeing the just the, you know, kind of the basic capabilities that I'm looking at, yeah, it's fascinating to see, how it's being used in particularly the diagnostic space. And so we're talking a little bit about some of the challenges in adopting this. One of the things that a lot of healthcare workers are very fearful of with AI is, you know, is AI going to take my job? And, you know, and it's interesting, particularly the radiologists, you know, they're worried that, you know, hey, I'm not gonna be needed because, you know, AI can come in here and develop a model that, you know, is really reading these scans, you know, much more efficiently and much more effectively than I can, again, a myth. So let's talk a little bit about, can you dispel that myth a little bit for the radiologists, but for really a number of healthcare workers? Absolutely. My belief is that we were always going to need doctors, no question about it. And I've seen again, in my forty years of being in this industry, I've seen the job of the radiologist evolve, right? Radiologists used to hide in reading rooms with x rays, you know, flipping up on their viewers and so forth. This is way before digital. And you had, you know, text loading up more films for them and so forth. So their life has changed, right? They've become more interventionalist for sure. They're getting more in tuned with patients. Some people went into radiology because they didn't want to talk to patients. Some of my best friends are radiologists and said, boy, it's great. I don't have to talk to anybody. I just get to do my job. But that has evolved clearly. And it's the same thing for cardiologists. AI is not going to replace a cardiologist. What AI is going to do is provide them additional information that's going to make them better at what they do, right? It's going to take the tedium away. When we do a guidelines check, right? They could do the same thing, but who wants to follow a flow chart that says if greater than fifty, go to this box, if less than this, go to that box, when somebody can do that for you in two seconds. It presents you with the data. Now you bill that as part of the information that you've got. Now, we don't look at images, but they still do. So now they'll compare the information that the AI has said with what they're looking at. They'll compare that with other information. They'll compare that with the conversation they have with the patient. So this isn't a tool to replace somebody. It's a tool to make them better. And it's a tool candidly, to democratize how we treat patients across the spectrum. You come to a big rural community facility and that's great. You'll have one tier of healthcare. But if you're out in a rural area and you don't have a cardiologist that's used to seeing these types of cases day in and day out, if you give them the opportunity to say, you've got the same toolset as the doctor that's in a large university setting, and we can make sure that you can make that same decision based on this best information, then why not? And that's candidly the biggest challenge we have is how do we get our solution further down, right? One of your previous guests was a primary care doctor. And if we could start to be closer to that patient at the primary care doc, when they're saying, you know what, I've got some shortness of breath, well, maybe you need to go get an echo. And right there and then before they get to a university, before they get to a structural heart doc, have solutions like ours in place that can sit there and say, you know what, you need a referral to a cardiologist or you need to be referred back to your primary care doc, you're okay. And if we can do that, I think we could start to address those challenges in rural communities, right, for example. That's really good news because, you know, I think recently rural healthcare is really up in arms. And the long term viability for a number of our hospitals. I mean, here in Texas, the number of hospitals that we closed here over the last decade is staggering because you know, didn't have enough doctors, didn't have enough capital to keep up. It just didn't have enough nursing or you name it. They just didn't have enough of. And so, you know, the limited providers that they have out there, if they have access to, you know, state of the art tools like this, you know, you're right. The people out there are going to get equivalent care that they would be able to get in a large setting. Well, and the other thing too, you know, speaking of the limitations of staffing out there by the adoption of AI, you know, these physicians aren't spending all their time trying to create their own guideline, you know, and so they're much more efficient. They can see more patients and, know, frankly, you know, back to the scarcity of income in the rural setting, they can make some more This makes it more advantageous for them to stay out there. And so that's a really good thing. When we go back and take a look at those cases that are underdiagnosed, as I said, the majority of them tend to be those low flow, low gradient cases. But here's the alarming statistic. The majority of those tend to be female. Females are typically underdiagnosed in this space because they present differently. And I have a few hypotheses that are unscientific as well. But but I think structurally, too, right? Their hearts are different. Their annulus is different size. How they how they present is different. And they typically present with low flow, low gradient type symptoms versus the traditional male. I also think the unscientific pieces, I also believe the females think that they have to take care of their family first. So when a doctor says, how are you? The answer is, I'm fine. I have to get back and take care of my family. So, you know, they're not being disclosing of what their symptoms are. So if we could have tools that help uncover that for them, then I think we could also aid the the diagnostic cardiologist and the rest of health care in making sure that everybody, whether it's a socioeconomic disparity or whether it's gender disparity, that everybody is looked at the same way. And because we just look at numbers, that's all we know is, you know, so when somebody says, well, do you have any bias? Well, no, because we're just looking at their numbers. So there is no bias, right? So we address that box pretty quick. So you guys are focusing on the structural heart right now. Any ideas about, you know, what's next for Echo IQ? Sure. Well, I think, you know, we just entered into agreement with Mayo to do some clinical research on a heart failure product. So it's not going to stop at aortic stenosis. We'll continue looking at other heart diseases. But I think you bring up a good question. And question is, where else can you take this technology, right? Even if you stay inside of this, how do we get it out into the rural communities? How do we get it how do we start to partner with some of the handheld devices as an example? Places that and some of that technology needs to change a little bit from what they're delivering today. But how do we work alongside of them for that? The other interesting conversations we've had is with some government officials that talk about from a from a DOD perspective, how do we get technologies like this in the field? Right. We're not putting civilians in harm's way because we've got, you know, we've we've got soldiers, you know, in theater that are that that are needing attention, needing care. So I think this A. I. Technology space is going to continue to evolve. It's going to continue to take off and we're going to find new ways that we can solve problems, which is exciting. But we're going to do so hand in hand with the physicians, not in exclusion of the physicians. And that's the most important part. I've heard these trite sayings that say, you know, AI won't replace the physician, but it will replace the physicians that don't use AI. Yeah, maybe, maybe not. I do believe that the physicians that embrace technologies like this will be aided and will, like I said, have that tedium reduced. We know from blinded studies that folks that have AI assistance actually read studies twenty four percent faster than those that don't. It's not any other reason than it's like having the answer key to a test you're about to take. We give you the answer upfront. We say this patient has the phenotype probability. They meet guidelines. They don't meet guidelines. Now you go read the study and you go, yep, you know what, I agree with that. And if there's a decision point of, geez, is this really the case or not? You kind of go back and go, well, AI looked at it and they've looked at this thing millions of times and they said it was, so it must be. So my gut, my intuition was right. And I move on to the next case. Yeah, well, it allows them to focus on, like you said, they have an answer key. So why not go straight to let's take a look at that first. And then if it doesn't fall, okay, it opens up to other possibilities. It's decisions. Yeah. So obviously you and I talk a lot about healthcare here, but is Echo IQ stepping out of the healthcare space? Not right now. To be honest, I think right now we see health care as a sweet spot for us. I think for us, the advantages, where else can we take this technology? You know, and part of this is also building the right databases. You know, some of your previous guests talking about AI really talks about the integrity of the information that you're dealing with. And for us, that was one of the big first steps we had to make was in creating this large database. And there was two founders, Doctor. David Playford and Doctor. Jeffrey Strange, Professor Strange, that set out to build this large database called NEDA, the National Echo Database of Australia. And it's through that that we have this robust data set that we could train the algorithms on and so forth. So as we begin to look at other opportunities outside of maybe cardiovascular or outside of healthcare, I think first and foremost, you have to start with where can we build the right database and where can we train these things on accurate information. So right now, I think we'll stay in our lane of health care, but I think it's going to continue to grow throughout that space for sure. Very good. Okay. We got a couple of minutes left. Don, any final words from our audience? You know, this has been, you know, we're geeking out about this because you and I love this stuff. Have to tell you, you know, as I said, I was blessed with forty years of standing side by side with some of the greatest folks in the world that are doing incredible things. I've told the success that I've had is only because I've been surrounding myself with really smart people. And that continues today with Echo IQ. I would say that to the audience, don't be afraid of AI. Understand it. Challenge it for sure. And I think we have an obligation as those that provide the technology to be able to stand behind it, to be able to prove to you why you should trust us because it is a trust, right? We get that. I understand that. I can't sit here and just say, just because I have to prove how we got to that answer. But once we do that, be open minded about it. And then also, know, guide us a little bit on how you want to implement and then let us challenge to figure out how to fit into that space because that's the other challenge is you can't continue to buy all these workstations, right? And we don't sell workstations, right? We sit on where your cardiologist is. For us, the last thing a cardiologist needs is swivel fatigue, having to log into someplace else, right? Don't need that. So we are where they are and we'll continue to do that from an offering. So we are agnostic when it comes to those things. But I would just say challenge us, but don't say no just because you hear AI. It actually is something that will help your audience. And if they're interested in finding out more, they can call me directly at don. Fowler echoiq dot ai or go right to our webpage at echoiq dot ai. Well, that's great, Don. I really appreciate it. It's always great to talk to you. Great to see you. It's a pleasure and I look to coming back to Waco and, we still gotta get to shorties. We promised the last time we did that. Exactly. We we've gotta go. We've gotta go get wings and thank you for throwing that out there. Maybe I go to shorties for some advertising dollars. But there you go. You know, I appreciate you. You're thinking ahead there, but but, Don, always a pleasure to see you, folks. Don Fowler, president of Echo IQ USA. It's been really interesting as always, particularly as I said, I've had a lot of guests lately talking about AI and it's been a lot of fun getting the chance to learn more about that. How not only does it integrate within the healthcare space but certainly in your everyday life. How can I utilize AI in my business? So again, like Don said, don't be afraid of AI. So with that, I'm Kevin Stevenson. Thanks for joining me today on I Don't Care With Me, Doctor. Kevin Stevenson, and we'll see you next time. Thanks.