This week I am talking to John Martin, MD, Chief Medical Officer, at Butterfly Network, Inc.(@ButterflyNetInc). John shares his personal interactions with the world of medicine both as a vascular surgeon but also as a patient.
The Magic Wand of Medicine
We dive into the world of healthcare innovation and technology, focusing on ultrasound imaging and its potential to revolutionize medical practice. Butterfly innovation in ultrasound technology, particularly the development of a handheld device with ultrasound on a chip, has been a game-changer, making ultrasound more affordable and accessible, putting it in the hands of individual clinicians, nurses, and potentially even patients.
We discuss the role of artificial intelligence (AI) in enhancing diagnostic accuracy and consistency in healthcare. AI tools, when integrated with ultrasound, have the potential to reduce costs and improve patient care. And as John points out ultrasound could replace traditional methods like chest X-rays in certain clinical scenarios, ultimately improving diagnostic accuracy and patient outcomes.
Listen in to hear the overarching vision to democratize access to medical imaging and bring ultrasound to the forefront of patient care, from medical schools to clinics and even the home. By leveraging technology, training, and AI, the aim is to create a future where healthcare is more equitable and efficient, ultimately benefiting patients worldwide.
Listen live at 4:00 AM, 12:00 Noon, or 8:00 PM ET, Monday through Friday for the next week at HealthcareNOW Radio. After that, you can listen on demand (See podcast information below.) Join the conversation on Twitter at #TheIncrementalist.
Listen along on HealthcareNowRadio or on SoundCloud
Raw Transcript
Nick van Terheyden
And today I’m delighted to welcome back to the show, Dr. John Martin. He is the chief medical officer for butterfly network. John, thanks for joining me.
John Martin
It’s a pleasure being back. Thank you.
Nick van Terheyden
So you and I met at one of the shows you shared a lot of the excitement. It was in person, obviously, we saw a little bit of the technology. And I’ve certainly been covering this before we dive into butterflying. And what it is that that technology is, tell us a little bit about your background. And you know, what brought you to this point in your career, if you would,
John Martin
yeah, and I think it’s very relevant to the discussion. A lot when you deal with physicians, you want to understand where does the credibility come in what you’re actually talking about. And so, first of all, I’m a vascular surgeon, by trade, and I still practice. So I’m clinically active and understand the nuances of the challenges of practicing and healthcare today, I also served in my previous capacity before butterfly as the head of heart and vascular services at a regional health center. So I was not only responsible for the vascular side of the house, but also the cardiology side of the house, which will be relevant to our conversation, I was the president of the largest cardiology group and metropolitan Washington for a time quite unusual for a vascular surgeon, and then helped lead its acquisition into a large health system and then became the vice president of physician operations for that health systems Medical Group, where he had responsibility over a number of these different disciplines. It also gave me kind of a window into in both those roles as president of the group service line leader. And in charge of operations rail system, the importance of economics of health care, how you actually impact care delivery the lives of physicians, and how that impacts care delivery. And at the end of the day, I guess the last thing is I’ve been a patient as well in health care pretty significantly. And I think that’s another perspective that I bring to the table. And so I like to think that my background in health care from from all those different vantage points, incredibly valuable to butterfly, and now I’ve sat for many positions, maybe the dark side inside a company, I think it’s not so dark over here and love the opportunity to straddling both worlds in the way I practice and work now.
Nick van Terheyden
Yeah, so thanks for that. Now, I’ll just, I always say that I’d say I’ve moved to the dark side, but I, I just I don’t feel it’s quite as bad. And in fact, it’s the combination of industry and technology and all of these other people that contribute to advancements. And, you know, in this particular case, let’s get into the detail of butterfly, I think it’s important to understand. So historically, you know, we’ve got imaging capability in the healthcare setting. And, you know, over the course of time, it sort of expanded from laying of hands on patients, you know, we had the advance of the stethoscope, which, you know, move from actually placing your ear against a patient’s chest, which always struck me as Wow, that must have been quite the thing. You know, and we moved to all of these additional modalities, I’m not sure where in the sequencing of imaging ultrasound came in, but it was a revolution when it arrived. Tell us a little bit about that. And, you know, bring us up to speed with where we are now.
John Martin
Well, I think a lot of, you know, ultrasound imaging has been around for, you know, I think you could look at about 5050 years now, on how it’s been around. It’s in its evolution over time has really been focused on the fact that it doesn’t use ionizing radiation, which I think it’s the real advantage that it served initially. And I think where that’s expanded now is as the as the device is moved from being these big, bulky carts the same size as a portable X ray machine these days, all the way down to the world in which butterfly now lives and that is a handheld device, expanded the use of imaging into all those different care venues that made it so critically valuable in decision making. And I think handheld ultrasound point of care ultrasound, your lovely POCUS word has been around for 10 years, but the device is still works relatively expensive. It required multiple probes to scan different parts of the body. So it never broke through crashed through that ceiling of really becoming a personal device. It was an asset that was owned by the hospital by the department by the practice, not an individual physician walking around with their own. And I think what butterfly really did that was special was became the first kind of commercialized ultrasound on a chip. And obviously following the path of where computers went where cameras went, when you put something on a chip, you dramatically change it. And so by doing that butterfly made an entire system be in one single probe so you can scan the entire body. It made it orders of magnitude affordable, down to the level that An individual physician could own one. And what I think the other novel part of this was we recognize it’s not just about getting a picture, it’s fitting, getting that picture in the midst of the workflow, integrating into the health system. And ultimately another learning for butterfly, candidly, a lot since since we’ve talked the importance of education and training into that. So butterfly has now evolved into, if you will, a complete system that we bring to bear so that someone can go from not leveraging the power of ultrasound at the bedside, and knowing nothing about it to becoming not only competent, but confident that you can use this modality. Because I will, I will say this, and I’ll say this to the end, if you know how to do this, you’re a better doctor.
Nick van Terheyden
Yeah, so let’s just a little bit, dive into some of the detail because I think it’s important this, you know, single chip sensor is quite the move that was really the innovation that I think changed things, it used to be that there were at least two elements to the device, they were in not enormous, they were large, comparatively speaking, you know, machines that were wheeled around, I can’t even remember a portable version of this. But what you managed to do was to essentially put everything, so the actual creation of the ultrasound waves and the sense of it. So you you put the nose of a dolphin into a chip, is that a fair summary?
John Martin
Yes. And we leverage the power, the processing power of the chip itself to process those signals. And then you take on top of that, you know, what you do with the smartphone, leveraging the power of that massive computer, that’s the side that fits in our pocket there as well, combined those two things. And then we added the third element, and that is the cloud, where images could go. And that cloud serves that that place where security and data transfer into health systems came about. So it’s those three elements that were key, but the secret sauce, you know, the the key element of innovation was putting ultrasound on a chip that changed the game for everything.
Nick van Terheyden
Right. And that really sort of brought this down, as you said, the individual device. And, you know, I’ve had guests on here, in fact, I had a lady that was working out in Africa, she was using one of the devices, you know, completely revolutionary in this space, because it was almost, you know, the rarity to have access, and then she was walking around the facility. Now it did mean that she had to be because she was the action of this, this. So it tied her to the whole process of actually carrying it out, which I want to say is not traditional in the healthcare setting. And you know, this is also one of the changes that I think you’re introducing, and you made a comment earlier, which is, we all get better as a result of this. We devolve this out to experts, ultrasound technicians, and I know that I definitely wasn’t one, because I’ve watched them doing this. And you’re saying now pulling it back in potentially with the clinician.
John Martin
Yeah, and I think it comes in a combination of two different things that I think are really important. And that is, we’re building out those educational tools to make it easier. So we have a whole learning management system, we’re about to introduce an application that allows you to work independent of being in the clinical environment to train something we’re gonna you know, we’re going to introduce here shortly, that I think is going to be important, but we also add artificial intelligence tools that help guide you to get the right image help you interpret those images, how do we leverage the best of technology and then bring that to the bedside to accelerate your journey to competency and I think that’s the key is that, okay, I don’t want to be intimidated by ultrasound. That’s something a tech went to school for three years to learn, or it was where fellows in ultrasound, learn how to do it. We’re accelerating that timeline, with the combination of great educational tools, artificial intelligence. And there’s a third element that’s really important. When we get into medicine, and I became a vascular surgeon I kind of was engulfed in I need to learn everything about vascular surgery. And so I went through a long fellowship with all these different things. When we train people on ultrasound, now, we tell them to focus, learn this skill, learn how to use Wong, ultrasound, just get good at this, it’s a start, you don’t have to know everything about everything. Learn this one, it’s important enough just to have that skill, that it’s worth adding to your clinical armamentarium. And then you can expand on that as you begin to gain comfort with how you scan and where you get the images and how you move your hand. And that’s really another big change that I think we’ve introduced to the importance of being very focused in your education.
Nick van Terheyden
Yeah, it reminds me a lot of my commentary around chat GPT, which, you know, there’s all this resistance for all, you know, some very good reasons, but in this particular case, I think it’s all about experimentation as much as anything, trying these things out to see how they apply if you can get them into the hands. So let’s talk a little bit about the economics because I know you said it It It’s democratizing access, because it is significantly less. But this is not a, an insignificant investment for an individual physician to make. How, how are we going? Can we get it lower? Can we improve this so that it really becomes? You know, and I’m, I’m going to, you know, bristle, Eric, on I think he was in favor of this. So, you know, can we get rid of the stethoscope? And I don’t want to because it’s, I like carrying that thing around. But, you know, can we really replace that with an alternative technology.
John Martin
So I love to share with you a quote, that I think you may find that it’s interesting. And this was this was written and I’ll share with you when and what it’s about, but I think your you and your, your listeners will find interesting. And this is a quote about a particular device that it will ever come into general practice, I am extremely doubtful, because it’s beneficial application requires much time, and gives a good deal of trouble both to the patient and the practitioner. Now, I think I’ve heard people say that a lot about learning ultrasound. But you know what that that was that was by a doctor Forbes written about the status, I don’t see any three. And so we’ve literally come full circle here. And yes, it’s, it’s it’s a couple 1000 A little over a couple $1,000. But the price of a computer, you know, not that far off from the phone that we all carry?
Nick van Terheyden
Yeah, that’s true.
John Martin
And if you look at the value that it delivers, it far outweighs you know, the cost of of what it makes, I think one of the great commitments butterfly made at the beginning. Maybe we could debate it, we made it so inexpensive, that as a company, you know, you got to work hard to make it work. Because the margins are not huge, obviously, in something like this. But it was because we felt so strongly that everybody should have one. And so I think the value is there, it’s not so much how much something costs, but how much value it brings to me. And then when I talk to CFOs of health systems or practices, I’ll show you where your money can come back. Don’t worry about the cost of this. It’s it’s worth it. And for me as a physician, to be a better doctor is priceless. You can’t put a number on that for me.
Nick van Terheyden
But fair comment for for those of you just joining, I’m Dr. Nick the incrementalist today, I’m talking to Dr. John Martin, he’s the Chief Medical Officer with butterfly network, we were just talking about the innovation and the democratization of access to ultrasound, you know, the real opportunity of sort of putting tools in the hands of physicians for specific tasks. Let’s talk about something that, you know, I gotta say has constantly troubled me is the incidental Omers that this is inevitably going to pick up? Because the reality is, if you do put this in every pocket, I you know, you can’t resist it all. Let’s just have a look, are we going to start finding things in because to your point, you should be using it for a specific purpose. Is that part of a responsibility? Or is that something that you’ve sort of thought? I’m curious about your thoughts?
John Martin
Well, you know, I’m kind of one of those interesting people to have this conversation with, because, and you know, this, and it’s been out there, so I’ll share it. The incident Aloma and using the device in me, was my cancer, and I bound a metastatic cancer with with my device. Now, that being said, it could easily have been a benign thing that I got all worried about, and maybe it generated an operation, you can certainly think of that in many different scenarios. I think at the end of the day, it’s the power of the physician knowledge and expertise combined with technology that makes the good clinician, I don’t see these tools taking over clinical judgment. I think as we get better and better with artificial intelligence tools, they’ll add value and the incidence of times we’re incidental, almost will be able to be aided by an artificial intelligence tool and say, Nope, don’t need to worry about this. You do need to worry about that. I think that’s coming. And that’ll be additional, really powerful information. But I think there will be always incidental aromas. And it’s differentiating that and using my good clinical experience and judgment to know the difference, I think is important. And I know that everybody around the world is hot on saying, how do we leverage now all this information to create smarter systems, so the chances of having an incident alone will become smaller and smaller and smaller, that we know the difference between what I got to worry about and what I don’t?
Nick van Terheyden
Yeah, absolutely. So I think fair comment. I mean, this is existed for time immemorial. And you know, we do and it’s the application of knowledge on top of all of the technology and the experience that is now lifting all boats to sort of elevate that discussion. Obviously, you can’t sort of go through life without stress and without the sort of discovery of various things. So, you know, it’s, it’s going to be inevitable. Let’s talk a little bit about AI. I mean, I think we’re in you know, definite hype of The value proposition of how this, you know, delivers value. You’ve partnered with some folks we’ve talked about that I’ve certainly, you know, discussed it in terms of bringing additional value with the capabilities. Where do you see this going? I mean, it’s, it clearly needs, I think it’s necessary because of the complexity of some of this knowledge that we’re trying to impart to the people using these devices. What are your thoughts about how best to sort of enable this so that you get the maximum value at the clinical point of care?
John Martin
I think it’s really picking those applications in which the AI is particularly suited and understanding the true value that it brings to the table. So for instance, let’s use some things that are out there that are pretty consistent ejection fraction. You know, we often talk about ejection fraction, like it’s some real magical, precise number, was such a fun conversation with our machine learning people trying to have this explained to them, it’s really kind of this, you know, we cope with me, what does the image look like? What did the stuff go to? And their response is always what must be a real accurate number? Well, no, there’s beat to beat variability. What artificial intelligence tools bring to the calculation of ejection fraction is the consistency, right? Same methodology being used every single time, and so that I can make decisions, knowing what was calculated was done exactly the same way. And if it went up, or if it went down, and by how much I can No, reliably, we measured it the same way. Something relevant to what we’re going to talk about later is B lines, when you’re talking about congestion B lines appear as an artifact when there’s congestion. But when you’re counting them manually, there’s a little bit of subjectivity to that. And if I’m going to make a decision, are they wetter or drier? Based on a subjective thing between interobserver variability? How accurate are we, if they’re all using the same tool, and we’re looking at from time to time, that’s much better medicine. And that’s where I see these artificial intelligence tools being incredibly helpful. Not to mention, you know, differentiating a benign from malignant tumor and how accurate that can be. But I really see them being incredibly helpful to create some consistency in the way in which we manage patients, which is so desperately needed.
Nick van Terheyden
Yeah, I like that sort of trajectory of consistency, that’s always bothered me. And I’ve got to say, it makes me smile every time I see people saying, you know, if your ejection fraction is less than x, and I’m going, Okay, well, as long as you know what that is, because it’s not quite as easy to, to determine. And I think that’s true with a number of things. Obviously, you’re talking about point of care, you know, specifically emergency room visits for difficulty and breathing. We’ve obviously been through this big pandemic, huge issue around that, you know, the traditional method is, well, you know, it’s a clinical assessment, there’s a little bit of that, you know, is it this? Am I hearing that here? You’re saying you’re adding in an AI tool? Is it going to replace something, or we’re going to reduce costs as a result of bringing in an additional AI? And, you know, let’s be fair, there’s an additional cost with this, is this going to bring in? Or get rid of something? And if so, what?
John Martin
Yeah, and, and I think I’ll give you a couple examples. One I think it’s going to get rid of, and I will put myself as a crusader. If f is the, the Knights of the roundtable, I will because I want to get rid of the chest X ray, at least the chest X ray, when it’s used as routine, if you come into the emergency room in short of breath there. And there’s a suspicion that you have heart failure, there’s the chest X ray BMP reflex, those two tests are automatically done. But if you look at the accuracy of a chest X ray for congestive heart failure is terrible. I mean, it really is terrible. Their studies range from 50 to 70%. Really good ones. I might as well guess my clinical judgments is good, is that accurate, but it’s gotten every single time. You look at the sensitivity and specificity of lung ultrasound, all different kettle of fish. Now you’re up to 90 plus percent of the time, it’s much better. And oh, by the way, if you’ve ever been in the emergency room and waited to go get a chest X ray, or waiting for them to do it, what is the agonizing moments that you’re short of breath or as a physician, what’s the diagnosis as opposed to triage, you know, a nurse brings it out, scans real quickly spits out beelines tells the doctor here’s the ultrasound, he says give you lasix. I’m on my way home by the time the X ray technician is rolling in for the other patient. And so I’d like to see it there. I’ll give you a second quick example. And there’s a study just published that compare the use of ultrasound for kidney stones and hydronephrosis. And the claim that the sensitivity of that was so good that you could probably eliminate the CT scan and incredibly expensive test that takes time and just do point of care ultrasound. So I think there’s some great opportunities to reduce cost.
Nick van Terheyden
Yeah, so you talk about that. And there was an interesting point you made there. That I think is right given to this, because I hadn’t heard this. And I think it’s important to understand you said, this is in the hands of somebody else, not necessarily the physician here. You said nurse scan. So I’m assuming that they’re, they’re both going through the training that you’ve talked about, but also some support in doing it, or is it? Is it sufficiently facile that they’re able to do that producing a value, which, as you importantly said, is consistent, that allows you to then make that assessment? Is that what’s going on here?
John Martin
Yes. And we’re doing clinical studies right now. And we’ll take it from we’re doing them with nurses actually doing the scans? And obviously, you have physician oversight? And what’s the quality of the image? And what did we get? We’re doing studies with a patient scanning themselves, can you get the image good enough to make clinical decision, and that’s where we’re going. At the end of the day long. Ultrasound is a great example. And I don’t wanna say it’s easy. It’s easier than many of the other complex tests. And it’s clearly been shown something that you can learn and learn quickly. So in the span of a few hours, we can train you how to do lung ultrasound, we did a great study where they watched a five minute video, you still need oversight, you still to make sure you get good quality. But yes, this is something that you can learn that is transformational in the way in which we deliver care to patients.
Nick van Terheyden
Right. So I’ll look for the package with my butterfly in the mail, so that I can sample this out and try it for myself.
John Martin
Well as as a physician, sure. We’re not quite there yet with the FDA for you as a patient, but we’ll get there.
Nick van Terheyden
You know, I’ve been on this trail for a while I keep I keep hoping I keep you know, I go to the mailbox every day looking for that opportunity. Because for me, it’s that it is it’s it’s excitement around there. So, as we sort of close this out, tell me where you think this is going, we’ve done some amazing things. We’re excited about it, you know, putting it in the hands of other clinical specialists? Where is this going with? Obviously, the combination of AI as well, I think,
John Martin
yeah, I guess I would say where it’s ultimately going, our journey is dealing with two thirds of the world that has no access to medical imaging, it’s going to go there, we’re going to improve health quality, and we’re going to try to eliminate health inequality. By having this device available to everyone, it’s going to move to the left to the care continuum. So it doesn’t have to just be in high end hospitals. It’s moving into clinics done by someone other than a physician. If ambulances and transport were exists today and ultimately into the home, it’s going to move into rural America into FQHCs. Around the country, how do we make the power of imaging available to everybody, we’re gonna leverage training, artificial intelligence tools, and a truly remarkable innovation, ultrasound on a chip to make that possible. Healthcare will be better doctors will be better medical school education is getting better because of it. I think that’s the world that we see. And I don’t think it’s going to take the Hubble telescope to see out into the future to know that it’s really just around the corner, and evolving in a bright in front of our very eyes.
Nick van Terheyden
Oh, hold on, I think the Hubble telescope looks backwards, not forwards. I think it’s looking back in time. Well, we
John Martin
looked backwards with the Hubble telescope to the quote about the stethoscope and said, You know what, let’s take that same kind of mentality and understand that that’s really where ultrasound is going. And it will be in the hands of every clinician one day.
Nick van Terheyden
So I think, you know, one of the things that I would expect to see how to see in the near future is putting this into the hands of the educational areas where we’re bringing on new cohorts of physicians, para clinical folks. Emergency services, is that starting to happen?
John Martin
Yeah, we’re in over half the medical schools already in this country with butterfly and we’re and the number of them that are giving it to every single student grows every single year, in this in this is not an if any longer in the world. When I came to butterfly, it was if we’re going to be able to get there. The only question now is, is when and that’s really a matter of how fast is everybody going to pivot? The more data that comes out that shows this, this makes you a better doctor, and we can deliver better health care, the faster all of us, hopefully will, will adapt and adopt this kind of way of managing patients. We’re not good at changing in healthcare, but the time has come to change faster, because candidly, our patients deserve it.
Nick van Terheyden
Yeah, and I think it’s just it’s not going to be just terrestrial. I think it’s going to be extra terrestrial. And we’ll be putting this in space as well by the sounds of it. So exciting times as usual. Unfortunately, we’ve run out of time, so just remains for me to thank you for joining me on the show. Always a pleasure, John, thanks for joining me on the show.
John Martin
It’s been a pleasure discussing this with you. You always have such great insight into where we’re going in health care.
