Latest / Powering the Future Within / Episode 18: "The Courage to Go All In" with Bill Hunter of Canary Medical
Transcript
- speaker-0: Welcome to Powering the Future Within, the podcast where we explore the breakthroughs shaping the future of MedTech. I'm Omari Bauknight, CEO of ResidentLake Medical. Each episode, we speak with a leader who is driving the next wave of medical innovation, advancing technologies that are smarter, less invasive, and more connected than ever before. This is where the future of health takes shape. Let's dive in. Today's guest is Dr. Bill Hunter, president, CEO, and Chief Medical Officer of Canary Medical, where he is pioneering a new generation of intelligent implantable medical devices that combine advanced sensors, data, and artificial intelligence to transform patient care. Dr. Hunter is one of MedTech's most accomplished physician entrepreneurs. Over the course of his career, he has founded and led companies. Responsible for some of the industry's most influential innovations while helping shape how breakthrough technologies are developed, commercialized, and adopted around the world. Today, through Canary Medical, he is helping redefine what an implant can do by enabling devices to continuously generate meaningful clinical insights long after they leave the operating room. Bill, welcome to Powering the Future Within. speaker-1: Thanks for having me. speaker-0: Really really, really excited to to have you here. ⁓ very much a a topic I think that our our audience will enjoy hearing. I'm curious if you were to think about when you finished medical training back in the nineteen nineties, would the people around you, your colleagues at that time, would they have predicted a career for you that centered more on innovation rather than traditional clinical practice? speaker-1: No, not at all. I I I d I really been in love with medicine since I can remember. and as hokey as it sounds, I I I don't think I ever wanted to do anything else. ⁓ and so I was inc incredibly excited to, you know, be a doc and travel the traditional medical course. and ⁓ ended up in this line of work, you know, by accident, not by design. So I I think ⁓ most people would be pretty surprised that I ended up, you know In a completely different direction than where I started. speaker-0: Was it a particular occasion or something that ⁓ triggered your your interest in moving more towards innovation? speaker-1: No, I was a hundred percent accidental entrepreneur. so I I was a a lesson stellar undergrad. ⁓ and so I went to graduate school as a way to make up for my ⁓ the ills of my youth and and try and get into medical school. And and so I I went to a graduate program at the University of British Columbia and anyways, long story short, the the work that I had done in graduate school led directly ⁓ to what became the the taxes stent and the drug coded stent and and what ultimately became Angiotech. And ⁓ so while I was in medical school, I, you know, because I went to medical school also at the University of British Columbia. So I I continued my research during medical school and we started a company. we we really only did that because it was way easier to get money through matching kind of corporate government. grants than it was to, you know, to go to a major granting agency and and get a grant. And you know, my I was I I didn't have a great academic lineage and I was never going to compete for NIH grants and stuff like that. So we we honestly we we set this company up as a way to fund the research that we were doing and and to make these drug coded scents. and the plan had been to just continue on and and go into medicine and I woke up one day and the darn thing actually worked. I mean the the the animal studies that we had done were successful and we'd done a couple of of human patients and they were successful and you know, I took a year off, which became two years off and and and and and so one day I just woke up and it had consumed my medical career. But it it was completely accidental. It was completely not by design. and I I just kinda went with the flow. I I I lacked career planning, I guess, and that's how I ended up in this business. speaker-0: How did you manage to balance medical school while building a company, while conducting research? It seems like three things that occupy someone's full day alone. speaker-1: Yeah, I I was lucky. I I was the the graduate program I was in was the Department of Anatomy at at UBC. And as a result of that I had taken a bunch of the the kind of heavy courses that you take in medical school as well, like gross anatomy and neuroanatomy and histology. I'd already taken them. And so I actually was was able to kind of TA during the first part of my medical school because I I had a lower case load when I when I went into medical school and so I was able to kind of balance both. And ⁓ you know, I I've always loved research. I I I'm totally into it. I had planned my whole life to have been an academic clinician. Like I I I never doubted that I want to be involved in research. I just never thought that I would solely end up on the other side of of it. And so I I was able to balance it as I went through. ⁓ when I interned and did other things it got a little bit more complicated as the hours got longer. But ⁓ It was ⁓ yeah, you know, it it just kind of all worked out. And I had I had great support. I had ⁓ I had co founders who were exceptional scientists and and ⁓ it it was actually kind of ironic that it was the medical student who ended up being the CEO. I was definitely the least qualified of my of my founding group. ⁓ but I was also the only one who didn't have a day job. So that's how I I think I ended up in the in the ⁓ position because you know, my other founders were university professors and an interventional radiologist and And they had they had real jobs and were making a real living, so so I was the one who had the spare time, quotation marks, to to pursue this this business idea. speaker-0: Different type of a short straw, I suppose. speaker-1: Yeah, it really was. Yeah. It was it was more the, you know, everybody takes one step back when they were looking for volunteers. I think that's kind of how it turned out. speaker-0: So you're you're conducting this and we often see in MedTech a gap between what's technology, what's impressive from a technology perspective, versus what actually fits clinical reality. Did you find it helpful to be going through medical school as you evaluated drug alluding stents? speaker-1: Yeah, I I I truly did. and you know, now I'm at the other end of my career. I have a lot of people who are who ask advice about stuff like that. And the the one thing I will say about medicine is it it's a wonderful grounding. Like i if you're going to have a background, it it it's really helpful. you know, the the PhD programs are outstanding and I've I've worked with great PhDs. the I won't say limitation, but the the issue with a a PhD program is that you tend to be incredibly focused in in one specific area. and obviously my graduate training was was like that as as well. ⁓ but in medicine you have this incredible breadth. You know, you you don't know a ton about anything, but you know a lot about everything. And so it's really helpful. And as you get into, you know, development And you, you know, have to learn about toxicity. Well, you know how the liver works and you you know how, you know, the the organ systems work. And so it it really is this wonderful diverse background that allows you to to have an idea of what's going on. And and secondly, you know, particularly if you spend time in clinical medicine like I did, I I did practice for a while, you you realize what is practical and what isn't. You know, people come forward with ideas and they're like, ⁓ I've got this Great treatment and you know, we just have to deliver it intrathecally, you know, five times a week, and you're like, Yeah, that's just never gonna work. Doesn't matter how successful the outcomes are, you know, that's just not practical from a clinical perspective. So I also think it gives you a grounding as to, you know, it kind of reality bases you as to what what will actually be practical in a hospital or clinical setting that, you know, may not be apparent if if you haven't, you know, spent time in in that environment. speaker-0: Right. Right. The the clinical realities of practic the practicalities of applying some of these technologies are are really important. So if you if you go back to that that animal lab, you come out of the lab, you're seeing that the drug eluding stent seems to be working in in your animal evaluations. And of course we're talking about using drug eluding stents for the indication of coronary artery disease. Drug eluding stents would go on to become one of those very rare med tech innovations that achieved massive adoption almost immediately ⁓ after approval. Coming out of those animal labs, did you have a sense at that time of how profound the impact could be? speaker-1: No, ⁓ there definitely was an aha moment, but it came much later. It it came in after the the big clinical studies. but kind of a fun fact going back to at that time and place, ⁓ coronary stents, when we started working on on drug code stents, stents weren't actually approved yet for coronary use. And they had been approved for biliary stenting. And ⁓ the the size of of the biliary tract was roughly the same size as the kind of main branch coronary arteries, the the so-called benestent lesions were about the same size as the common bile duct. And so people were using biliary stents in the coronary system ⁓ to try and and do that. And so we were working on drug-coded stents a couple of years before Coronary stents were actually approved, much less drug-coded stents. ⁓ and so early on, it it seemed quite futuristic. It seemed a long way away because, you know, the platform hadn't been improved yet, much less the new and improved platform. Having said that, though, we we were hardly the only people in the world who were focused on that. You know, there were other groups who were doing really great work as well, but you know, it it seemed very far off. But the Nice thing about is that the pig models in coronary stenting are fairly predictive of of the human models, at least with respect to recinosis, not always with respect to coronary artery disease, but with respect to the scarring and re-sinosis ⁓ you know aspect of it. That if you can get it to work in a pig, you have a reasonable probability that it will work in man, although, you know, not a hundred percent for sure. But, you know, once we had those, that data was interesting enough to the big players in the game, which were, you know, Metronic and Guidant and Jane J and Boston Scientific, for them to to take an interest because, you know, the the cost step up in coronary scenting going from an animal model to, you know, large scale human studies was hundreds of millions of dollars, right? And well beyond our capability. So that was an important stepping stone. But so many things had failed to transition in man. That I don't think we really knew what we had until we had the first human studies. speaker-0: Well, and I I'm also curious, it's a it's a really important insight that at that time coronary stents, bare metal stents had not been approved. So a as you kind of look into the future and you see the difference in performance between drug alluting stents and bare metal stents, it's really about reducing the amount of rstenosis. So again, we're talking about coronary artery disease with bare metal stents. One of the the challenges really was there's a high proportion of patients that would see stenosis or restenosis through the stent of the coronary arteries. And eventually what we saw over time is that drug eluding stents perform better ⁓ because of the the the drugs, essentially reducing the amount of restenosis, so therefore reducing the amount of recurrent coronary artery disease. So I'm curious if you if you didn't have ⁓ coronary if you didn't have bare metal stents in the market for coronary artery disease, what were what was the thinking around how drug eluding stents might be better because we didn't really have definitive data at that point, right? speaker-1: Yeah, a cardiologist once said to me that the interesting thing about the evolution of interventional cardiology is that we kept creating new problems that had to be solved. So every every generation of device created a new clinical problem that required a new clinical solution. So ⁓ you know, obviously angioplasty balloons came first. and and they were great. They could, you know, open up an artery that that was was closed. But they created a new problem, which was called acute closure. So when I was early in my training, you know, code blue to the calf lab was a really common occurrence because when you'd you'd crack the artery open with the balloon, when you'd withdraw the balloon, sometimes you'd you'd structurally damage the artery so much that it would just collapse. And and so you'd have this acute closure and and they'd literally have to whisk people off from the calf lab to open heart surgery to, you know. repair the artery because there's just really nothing else you could do at that point in time. So the stent was there to solve that problem. So the stent was a metal scaffold that would keep the artery open. So if you crack the artery with the balloon and it collapsed, there was a a scaffold there to keep the artery open. Well then that was great. That took the failure of angioplasty from, you know, I don't know, maybe thirty five, forty percent down to, you know, twenty, twenty five percent. But that created a new problem. The body's like, well, this stainless steel's not supposed to be there. And a scar grows over top of the stent. And lo and behold, the stent closes at thirty days or sixty days because of scarring and re-sinosis. So then drug coated stents were designed to solve that problem. And then when drug coated stents solved that problem, we had a problem with clotting and it it just was this, you know, we kept going down from, you know, forty percent to you know twenty percent, drug coded sense got to one or two percent, and then you know, and so on and so forth. But every time we solved the problem, there seemed to be a new one that emerged. ⁓ although, you know, the the success rate kept going up and up and up, which is why they were so successful. But it's been this never ending evolution of of, you know, trying to to create a new solution for the new problem you created with your last invention, right? speaker-0: And to that end, I'm I'm curious, ⁓ given your perspective and and all the experience, there had been some evaluation and investigation and desire to see whether or not bioabsorbable stents could therefore then be the the next step in terms of addressing potential issues down the road. Do you have a perspective of whether or not we'll we'll see we'll get back to that for treating coronary artery disease? speaker-1: Yeah, it's and and that creates a whole new so we we experimented with with those polymers as well and and there's pros and cons and like everything in biology, I think there'll be an area that it it's it works for and an area that that perhaps it's less for, but you know and I'm my my knowledge is antiquated now 'cause I haven't worked in the field for a while, but you know the issue was that yes, it was great. If you could get rid of the stent, then you got rid of the foreign body, which meant that you got rid of the stimulus for restenosis. So perfectly logical. And in some cases that appears to work. In other cases, what would happen though is that remember the stent was still there forming a scaffold function. And if the artery hadn't completely healed and remodeled itself in an open configuration and you took away the structure, you then were faced with this, you know, potential problem of of what was was going on. And and second, you know, if the the scent stayed on the surface, those polymers, if you if you think most of them are what are called surface eroding polymers. And if if you think about when you're a kid and you have those suckers that you you would suck on the the kind of classic things, they get thinner and thinner and thinner and thinner as you surface erode them, if you will, but eventually they get so fragile they break. and of course an artery is pulsating, you know, 60, 80 times a minute. And so, you know, there was always the potential that if they weren't buried in the wall, then these polymer flakes could, you know, break off and and become embolic agents. And so again, every time you solve a problem, you created the potential for a new one. although, you know, what I've seen and I ⁓ advise for a company that that I think is doing some really cool work in ⁓ neurological stenting with a a ⁓ degradable stent. I think there's going to be some clear places where it totally makes sense and and is is very, very useful. But it it's always been this yin and yang of of the law of unintended consequences when you work in these areas. speaker-0: Yeah. And and most of the time it's on unpredictable, unintended consequences. speaker-1: Yes. Yes, absolutely. speaker-0: So I'm I'm curious, looking back, ⁓ drug eluding stents would go on to be one of the most rapidly adopted technologies ever ⁓ commercialized in in the medical device med tech industry. What did working on that transition teach you about what separates incremental innovation from true platform level changing medicine? speaker-1: Yeah, I I you know, I I take no credit for for that product launch. It was it was Boston Scientific and and ⁓ and Jim Tobin who who really drove that. But if if I'm not mistaken, I think we took 70% of the world market in 70 days. It was it was absolutely astounding. And what I learned about that is the need to have the courage to go all in. ⁓ Boston Psy, who's who's obviously the 800 pound gorilla now, ⁓ was the number four stent company when when drug coded stents started out. I mean, Guiden and Medtronic were ⁓ you know, the the winners, if you will, of the bare mouth stent market. and Boston Psy really had that kind of underdog mentality, right? ⁓ I mean There's a there's a always a very big difference with working with a market leader versus working with somebody who wants to become the market leader. And they were willing to change their platform as well. Like their bare metal platform had not been overly competitive, which is why they were number four. And they were working towards changing the platform. At the same time, they were adding drug coding to it, and they went all in and and built manufacturing capability to actually supply the market. if they were that successful, which was a a huge risk and a you know, a huge investment. And so, you know, there's no point taking 70% of the market if you can only manufacture to, you know, ten or fifteen percent of the market. So they took so many risks to be there and that they were able to meet that need just kind of shows that, you know, if you're really serious about truly disrupting a market, you you have to invest ahead of time you have to you have to dare to be right. ⁓ because if you miss it, you know, and you you catch the tiger by the tail, but you can't actually scale to meet it, you know, you may lose a lot of customers or you may lose your lead, you know, in the year or two it might take you to catch back up. So takes a lot of courage, takes a lot of you know, you you've got to be willing to to not just be right, but potentially be horribly wrong. ⁓ if if you're going to to do that. And I think that's what I learned. It doesn't matter how big your company is, when you get to these transformational type technologies, it's still a bet the company type situation. Or at least, you know, for the people in charge, it's a bet your career type ⁓ piece and and, you know, healthcare is is ⁓ I'm a I love baseball, so I always say that healthcare is a a home run hitters game, right? I mean If if you're going for it, you you have to be willing to walk back to the dugout having struck out a bunch of times. But, you know, when you make contact, it's a beautiful thing and and you know, you really have to have that mentality. I I think even if you're running a big healthcare company, you still have to be worrying willing to take a significant amount of chance and risk if if you wanna, you know, be number one. speaker-0: Well and and you should certainly have demonstrated courage in terms of leaning into smart implants long before it was even a broader industry conversation you were building towards this future. W was there an original insight that convinced you that this category would eventually matter? speaker-1: Yeah, and and I've walked back to the dugout struck out a few times too. Believe me. you know. ⁓ now I I look at the the probably the two things that that I've I've invested most of my life into, and they're relatively simple and straightforward insights, right? ⁓ you know, the the premise behind Angiotech was was not rocket science. We didn't invent some new molecular pathway or, you know, some Nobel Prize winning piece of biology. You know, the insight was medical devices get infected, they get inflamed, right? They they get overgrown by tissue, and pharmaceutical companies make things called antiinfectives and anti-inflammatories and, you know, antiproliferative agents. So why don't we put them together? ⁓ and why don't we make a bioactive implant? Now, at the time that sounded, you know, pretty crazy. ⁓ now that sounds incredibly logical, but it wasn't a huge scientific insight. Now there was a lot of technology that went into that. You know, you had to do polymeric drug delivery, you had to do a whole bunch of things you you couldn't possibly imagine. But I don't think the insight was, you know, book my flight to Stockholm kind of thing. It it's a you know, it's a pretty straight straightforward. When I think about canary, it it's also pretty straightforward, right? I mean, this thing knows where it is on the planet, it can orient itself, it can, you know, take all kinds of readings. How come, you know, this thing can't? You know, it and the technology certainly existed. Again, we didn't have to invent the drugs that we put onto the devices, we didn't have to invent the sensors, we didn't have to invent, you know, the the transmission technology or the power technology or anything like that. And so I think the insight itself was fairly obvious. And, you know, my sense is that ten years from now it'll be kind of a given that most medical devices will provide some sort of feedback to the the patients and the clinicians. ⁓ so it it's really just, I guess, getting to the obvious conclusion before everybody else gets to the obvious conclusion. speaker-0: Right. And and the applicability for sensing and understanding is wide open throughout the entire human body and coming from the the cardiac world, you you landed on orthopedics as the first application. How did how did that unfold? speaker-1: ⁓ that again was ⁓ was not by design. ⁓ you know, I always felt, having spent, you know, twenty years working on stents, that if you could build a stent that could tell you whether it was open or closed, you'd take 70% share in 70 days as well. ⁓ you know, an angiogram is an expensive and potentially hazardous way to figure out whether an artery is open or not. And and if you were able to do that, you know, by Holding your phone up like this, you know, that that would be a a blockbuster product. The problem was the technology wasn't there yet. ⁓ the the power sources weren't, you know, available to do that. The the sensor capability was not small enough, et cetera, et cetera. I mean, the technology will get there, but when we started doing this, you know, in twenty twelve, the technology was nowhere near that. And so the only thing we could make was, you know, the size of a triple A battery. to have power and sensing and transmission capability. And the only device big enough to take it was a a total knee. so it was, you know, ⁓ out of necessity that we we went into orthopedics first. You know, we now have a cardiac program that's, you know, fairly advanced in in pivotal trials trying to do non invasive ⁓ pulmonary artery pressure monitoring. But that was only after the technology had moved forward for a full decade. And and I suspect, you know, were we to have this conversation ten years from now, you know, the capabilities of of all of those would would allow us to get into smaller and smaller and more and more you know, intricate device designs. But in the beginning, you know, we we went with the only thing we were actually capable of building. speaker-0: Right. And and as you evaluated orthopaedic medicine, was there anything that surprised you about, you know, the practices or the workflow? speaker-1: It's really different. really different. ⁓ to the point where I'd say that nothing I learned in cardiology has helped me in orthopedics. And and I've I've spent a lot of time thinking about that. And I I I have a theory, which ⁓ is strictly a theory, and I have no idea if this is even true, but this this is my theory. When I think back over interventional cardiology, so you know, I graduate medical school in nineteen ninety two. you know. the cardiac surgeons are still the king of the world. ⁓ you know, it's all bypass and open heart valves and and all our s and and I think of the evolution of interventional cardiology in the space of my career. It's it's unbelievable. I mean, it's not just stents. It's like if you had told me you were going to do valve replacements off of a catheter back then and you were going to, you know, you were going to close septal defects and like you know, you were going to tore out the artery, you're gonna use sound waves to to break plaque. I mean, the the pure in the pure innovation in cardiology has been breathtaking. And what that means is that the cardiologist has been kind of conditioned to accept innovation, right? Because a lot of this stuff started off kind of clunky, but it got better and better and better and better. You know, classic disruptive business theory, Clay Christensen stuff, you know, start good enough and just get better and better and better. And so cardiologists have had this nonstop series of innovation, which means they're early adopters of technology. They are receptive to new technology. ⁓ If you go to a cardiologist and you do a study that says, you know, this treatment's better than that treatment, they will adopt it quickly. So the flip side of that 70% in 70 days is that cardiologists had to be willing to change their practice relatively quickly as well. And so I think that was because they were in a very technology driven field. I look at orthopedics and orthopedics is very interested in technology, the you know, robotics and ARVR and and all this stuff. They're very very interested in data. It's not like, you know, they get a l a bad rap for, you know, hammers and and saws, but they they are actually very technology forward. But I'm not sure technology has had the same impact in there. business. If if you look at implants from 25, 30 years ago, their success rates aren't all that different than the ones that were put in yesterday. ⁓ if you look at robotics, it you know it has an impact, but not a huge impact. It's not like, you know, total knees went from, you know, having a 20% patient dissatisfaction rate to a 1%. It was kind of like 20% to 19 to 18. Anyways, moral of the story is I think the innovation in orthopedics has been more incremental. It's not like if I missed this, there was a dramatic change in in patient outcome. And I've I've heard a lot of orthopedic surgeons say, don't be first, but don't be last, which is, you know, these innovations haven't had transformational effects. They've had g incremental effects. And what that means is the orthopedic surgeons have not been huge early adopters of technology. And so what I've seen in orthopedics is that It's very much a more graduated and and ⁓ thoughtful ⁓ introduction of technology. There's there's not a huge rush to the new new thing, ⁓ because in the past they haven't seen massive benefits from the new new thing. speaker-0: Was that something you flagged as you did your assessment of the market in terms of bringing new technology to the space, or is that something that you just discovered as you went about your pro speaker-1: That's the s that's the school of hard knocks. you know, we we went out and and we said, Hey, you know, we think data's gonna make a really significant difference. We think, you know, we can do X, Y, and Z. And ⁓ people were like, Fantastic, you know, come talk to me when you have the proof. ⁓ in interventional cardiology, like, well, how how can I how can I participate in this study? Very, very different ⁓ type of approach. speaker-0: Yeah. And and so now you have the proof your initial application has been in total knee replacement, as you mentioned, you've accumulated a significant amount of data. What has been the clinical power you've observed in having smart implants? speaker-1: It it's it's really made me think about how these technologies evolve. you know, when you first start you know, you first start collecting data from any type of of implant or wearable or it doesn't even matter what it is, but when you first start collecting longitudinal data, ⁓ you really become initially a monitoring company. Because the the only thing you can do is collect, you know. and so you you start out at as a monitoring company. And if you look at the evolution of digital health and you you kind of go back to the first monitoring devices, it was all about the RPM codes, it was all about remote patient monitoring, and and that's where people got started. Then if you're fortunate enough to get enough patients where you get thousands or tens of thousands of patients. then you become a bit of a diagnostic company. ⁓ so you have enough data now to interpret it. Once you know what normal is, you start to be able to identify abnormal and you start to be able to see changes. ⁓ you know, I I'm fond of using the ECG analogy. Like I'm imagine if you developed ECG, you know, on day one, the first thing you'd you'd do is you'd just be able to monitor people, but you wouldn't know what a normal ECG was. But but once you had enough that you knew what a normal ECG was, then you could see abnormal ECGs and you could start to say, ⁓ when I see this particular pattern, that's an infection. When I see this particular pattern, that's heart failure or whatever. And so you become a diagnostic company because you you're able to tell the difference between normal and abnormal. And then, you know, something I'm sure we're going to spend a lot of time talking about. The next step is can I become a predictive company? So do I have enough data? So back to the ECG analogy, can I find silent ischemia before the MI? So can I see the those sentinel changes before the problem? So I have the opportunity to intervene, you know, you know, a priori. And I I think that's what we have started to find. You know, our our first things were kind of, hey, we got all this data, not sure what it means, but you know, we can tell you. how they're doing. Then we were like, okay, we now know what a normal recovery looks like, right? Until I know how what the range of motion for a 55 year old male on day 17 is, I have no idea if the next patient is ahead or behind ⁓ where their expectation is. That gets me into the, hey, red light, green light, or diagnostic phase. And then eventually, can I know enough about this pattern of recovery to say this person's at risk for developing this problem? Doing well, or conversely, this patient's heading on a normal trajectory and we can, you know, leave them alone. And and that's been kind of the evolution of of the data pattern. And and you you kind of grow up with your data. You can't get out in front of your data. You you you have to follow your data and and you the the state of your company and and what you're able to do with it is dictated by the size of your and and intricacy of your data asset. speaker-0: Yeah. I love how you talk about those three those three phases, you know, monitoring, diagnostic, and then eventually you want to get to predictive and preventative. I think our healthcare system as a whole, we do a pretty good job in the first two and not great in the third. Wha why is it so difficult for us to make that transition into predictive and preventative? speaker-1: I'm you know, with the caveat that I'm obviously horribly biased because I've I've spent like 12 or 14 years now working in this field, ⁓ you can't you can't measure what you can't monitor, and you can't predict what you don't have. And if you look at at medicine as a whole, we you know, as we evolved even from kind of clinical diagnosis to blood tests and x-rays and then to MRI and you know ⁓ and C T and and all the rest. All of these were kind of post hoc phenomenon, right? So the patient presents, the patient already has the problem. And and so our entire system was really diagnostically based, right? So the idea was to, you know, narrow down the infinite to the probable and then institute the right treatment. We didn't really have that kind of linear data that you needed. I think the reason I always use ECG or in specifically actually halter monitoring is that that was really kind of the first time we had this ability to send somebody home and and follow their patterns and start to understand what those patterns meant and and identify kind of of rare situations before they became complex situations. So I think Medicine as a whole is is has been following that same pattern I just talked about, right? I mean, the first we could kind of monitor people, we could look at them, we could put a stethoscope on their chest, we could see how they were doing, and then we got tools that allowed us to diagnose them and and I think now we're starting not just with monitoring, but with AI and and other things, we're starting to be able to predict. And I think medicine as a whole is going through that. three phase evolution and and we just happen to be fortunate enough to be part of it right now. ⁓ but it it's it's really kind of a technological shift that's that's happening within medicine. speaker-0: And so with with the canary implants, what are some of the things that you feel as though we're starting to be able to predict or can predict or in the f short future will be able to predict? speaker-1: Yeah, the the ⁓ I'd I'd put those into ⁓ probably three buckets. ⁓ you know, initially we were obviously very orthopedic centered. ⁓ orthopedic surgeons are concerned about orthopedic complications. So they're concerned about MUAs and infections and stiffness and and pain and limping and falling and things like that. And so our device was, you know, heavily weighted towards looking at things like range of motion or cadence or stride length and and using kind of gate parameters to address gait abnormalities. And I would say 90% of our effort went into that very early on. Then kind of as we evolved through that, we looked at why people go back to hospital and we looked at why people were went to the emergency room. And it turns out that only half the time it was for orthopedic problems. The other half of the time it was for medical problems. So You know, obviously many of these patients have comorbidities, they have diabetes, they have COPD, they have, you know, heart failure. And you give them an anesthetic and you put them through a a you know, a complex procedure, and to nobody's surprise, you know, they come back to the hospital because they have atalectasis or pneumonia or cardiac decompensation. And so then we realize that, hey, you know We collect things like step count and walking speed and and things that can be used as global parameters of overall health. Right. And this isn't of academic interest. And, you know, I I'm I'm sure we can talk about this later as well, but with this whole CJRX where the the docs are going to be paid for the 90-day episode of care, they're going to be responsible for those readmissions. And not all of them are orthopedic. So being able to identify the medical patients suddenly becomes really important. And the exciting thing, which we've only just scratched the surface, is something we didn't anticipate. Again, one of these law of unintended consequences. But, you know, when we went to school, they taught us that there's a Parkinsonian gate and an Alzheimer's gate and a CHF gate and a, you know, and I worked with some really great emergency docs who could watch somebody walk from the waiting room in and go, Acute asthma attack, congestive heart failure, right? Exactly like that. And, you know, it turns out that there are gate changes associated with all those abnormalities. And when you get a hundred thousand patients, then you can see what a Parkinsonian gate looks like or an evolving Alzheimer's gate looks like, or and so on and so forth, right? And and so that I'm excited about as a law of unintended consequences where gait can be a biomarker, not just For the acute recovery from TKA, but is this a biomarker we can use in disease states going forward? And I think all that speaks to is that if you can collect a physiologic data stream that's meaningful, you can do an awful lot of things with it. And, you know, clearly not to jump into the hype, but if you can apply AI to that, you know, you can do some some things that, you know, would have been very hard to do in the absence of AI. speaker-0: Yeah. Yeah, that's fascinating. ⁓ I think Malcolm Gladwell in one of his books once ta once talked about the referee at a a tennis match and their b ability to invariably predict if someone was gonna have a foot fault, which is sort of a rare event in in a tennis match, especially at the professional level. And you would have these long term refs be able to say, ⁓ that that player's gonna have a foot fault, and sure enough they would have it. ⁓ but they wouldn't be able to explain what they were observing or seeing that allowed them to make that prediction. Sounds very similar to what you're talking about in terms of gait and being able to say, Okay, that that patient likely has asthma or CHF or what have you. speaker-1: Yes. And you know, to extend that analogy, if you if you look at professional tennis players, they're already moving ⁓ to return of serve before the racket even makes contact because they've recognized the pattern as to whether or not the ball's gonna end up forehand or backhand on them. And in fact, if they wait for the ball to actually be hit, they they can't get to it. Right. ⁓ and and that's what you you see. ⁓ the the pre Gallon awards are ⁓ or kind of industries, Nobels, if you will, that that ⁓ go through these and and I I've I've gone to them the last couple of years. And one of the winners I thought was was quite fascinating where they had looked at hundreds of thousands of ECGs and they were able to use AI to diagnose the cardiomyopathies that are not diagnosable by ECG. And you know, because the AI was able to see a pattern in those ECGs that was beyond sitting there with calipers and and measuring and and and doing like, you know, the original AI that figured out ECG, which was the human brain. And so, you know, having these patterns and uncovering these patterns and and the rest, I think, are ⁓ you know, really fascinating next steps. ⁓ you know, there was there was a big ⁓ study in Europe where they looked at ⁓ patients' health records and they just by looking at ⁓ it was a very influential paper in nature and just by looking at, you know, if you saw the doctor for this and then you had that diagnostic test and that and you just followed that, that was predictive of you twenty years later ending up with disease X or Y, right? And and so these early pattern recognitions before the fact. Yeah, I I think that's going to be Nobel winning science, actually. But it very fascinating, just these these linear data streams. And back to your initial question, I think that's what was missing before. We had these snapshots. We had x-rays, we had blood tests, we had these single moments in time. But we lacked these kind of linear data sets. And so the more linear data sets we collect, the higher the probability that we're going to be able to predict. speaker-0: ⁓ But you have to go through those steps of monitor, diagnose, and then predict and that Yeah. That's fascinating. And I would imagine with your assessment of gait, you know, part of the thinking was ⁓ maybe we could predict the the potential that the patient is going to have a fall but then you've been able to expand the universe. Is that kind of where it started? speaker-1: Predict. That's right. Yeah, you know, I mean when I first talked to orthopedic surgeons, they said, you know, if you can tell me if a patient is infected or is about to get infected, or you can give me advanced notice, I wouldn't use any other device. So I think we were like, all right, let's focus on infection. And maybe someday we'll succeed in doing that. ⁓ we're we're starting to see some patterns associated with that. But it's also the hardest thing to do because infection is, you know, a one percent incidence, and mathematicians will tell you you need, you know, ten thousand patient records before you're even gonna have anywhere near enough numbers to take a sniff at at at doing that type of work. And since patients are lost to follow up and blah. Anyways, you do the math, you get into like a hundred thousand patients by the time you can do that. And so what I think happens, again, ⁓ you you follow your data, not the other way around, is that Common things occur commonly and you see those things first. so stiffness is the most common complication of TK. Not everybody has a a manipulation, an MUA as a result of it, but you know, it's it's a common problem. So you see stiffness first, and then you see, you know, things like risk of ED readmission or risk of a thing, because you know, 15% of people go to the ED, so it happens fairly commonly. So you start to see the brand. And then, you know, you find the stuff that happens five percent of the time, one percent of the time. And then you find the rarer and rarer events. And so while if I was doing classic product design, my record research would have said, if I can find infection, I'm gonna have a home run product. That just that wasn't gonna happen right away. So we've we've had to kind of crawl, walk, run our way towards bigger and more impactful insights. ⁓ just by by getting the necessary data in place to do it. So it is a bit of a leap of faith when you get started. You have to trust that the data's gonna lead you somewhere. and you take the risk that it might not. speaker-0: Yeah. And and how many patients now have some form of a canary implant? speaker-1: We're well over twenty thousand now. So we are you know, we we just submitted an algorithm for D V T. ⁓ so we're now getting into the rarer events. you know, we we're submitted an algorithm for stiffness, we've submitted an algorithm for ⁓ risk of of hospital readmission. ⁓ and you know, the goal here is is not perfection, the goal is is ⁓ a flag that says, you know, this patient is at risk of, right? And the way I explain it to clinicians is like, look, I you know, if I see an elevated white count, ⁓ that doesn't tell me whether the patient has a, you know, a pylonephritis or a pneumonia. It tells me the patient has an infection. And and clinical judgment, what what it does is it tells me which questions to ask so I can get to the correct answer, right? So, you know, if we see these changes, that doesn't mean the patient's going to have a DBT or this patient is going to be readmitted, but it flags them as not progressing according to normal, which should trigger you to ask the questions, you know, do you have a fever? Does it, does it, you know, hurt when you pee? ⁓ that will help you tease out what the data means, right? There's there's still an interpretive part of all of this. ⁓ but it is quite fascinating. You can you can start to see as as my head data scientist says, when you you have your knee replaced, everybody starts by not walking very much. Everybody's stiff, everybody limps. ⁓ but those who are gonna get better go this way and those who don't go flat. And mathematically these people start to separate as early as two or three weeks. And so even though they might not be back in the ER until day forty five or day fifty, right? They're You're starting to see the separation of those curves fairly early. And the higher the numbers, the better the math, the earlier you can push back that, you know, early warning signal. speaker-0: Yeah. A and so you're generating more and more data. We're getting closer to becoming more and more predictive. How do the how do you think about data ownership? I'm sure that's a conversation that's happening inside the company quite a bit. speaker-1: Yeah, and we we take two approaches. ⁓ you know, legally in the United States there's probably I think about four or five states now that say that it's the patient that owns the data. I wouldn't be surprised if that becomes somewhat universal over the next, you know, few years. We've always taken the approach that it's the patient that owns the data. and I I just I think medically that's the right way to do it. ⁓ we have the good fortune that we can be fairly transparent with our data. You know, you can't get our data unless you put this thing in. and so, you know, we can be magnanimous because, you know, the person who put it in, you know, there's no data without our implants. So we have been taken care of by by collecting the data in the first place. And that has meant that we, you know, share our data with everybody. So we we will share it with the hospitals, we share it with academic investigators, we let people probe our database for insights. ⁓ we actually submitted 49 abstracts to AOS and AUKUS last month. and many of those are are not really smart knee related. They're just related to if I give this drug which outcome do I get? Like does it affect range of motion or activity positively or negatively? And and so we've been able to address some kind of basic science questions. So because of our our business model where it's not so much the data but the device that's proprietary that allows us to get the data has meant that we can, you know, A, that we can make sure that it belongs to the patient. We don't have to hoard it. And and two, you know, we can share it so people can gain insights from it because You know, our business model allows us to do that. speaker-0: Yes, yes. And and so you're building this intelligence layer. It goes across multiple implant categories. That means that partner selection is a critical aspect of of what you're doing. How how do you evalue partners? Because in a lot of cases you're the intelligence layer that goes inside of another technology or inside of another device, right? speaker-1: Yeah, I mean, for those who haven't seen our stuff before, ⁓ where where we've evolved to is is basically building little cartridges so that they can slide into an existing medical device to convert it from a a regular device to a to a smart implant. And that gives us the flexibility in terms of, you know, do we work with one party? Do we work with everybody? And as I've thought about that from a business perspective, I've come out in in a way that probably isn't superficially obvious, but is when you think about it. So I categorize medical devices into proprietary and commoditized. And I don't mean that in the sense that everybody else does, but whether it's patented or not patented, I'm what I mean is there are definitely proprietary devices. And by that, yes, sometimes they're patented, but but The commercialization and development of that is a very, very sophisticated endeavor, right? I mean, I I wouldn't start a knee company. You you need all kinds of tools, you need lots of sales reps, you need all kinds of, you know, ancillary equipment and robots and all kinds of other things to be a knee company. ⁓ that that's it's just too big. I feel the same way about coronary sense. You'd need catheters and guide wires and balloons and imaging and ivas and Who knows what else, right? And so if the device is very complicated or it's part of a very complicated ecosystem, then partnership's probably the way for us to go. But then there are devices where, you know, that matters less. ⁓ I don't think orthopedic surgeons care much about their intermellary nails or their screws or their hardware. And, you know, in in those situations, the best device is the one that the the doc would prefer to use and and the the label on the package doesn't matter all that much to them. And in those situations, I feel like we can develop them ourselves because we don't need a sales force to teach people how to use an intermellary nail or how to do use an interbody cage. And so if we had a smart cage, they could integrate it into whatever it is they're doing. And that allows us to go it alone because the serial partnership model is a very, very difficult model to run. ⁓ and so you do need to have a couple of things that you keep yourself where you control your own destiny because, you know, the the your partner, even with the best intentions, may have a different objective in their business at some point in your partnership than you do. And and so I want to make sure that yes, we have a partnership portfolio in these kind of high tech, high touch type interventions. But I also want to make sure that we have some standalone devices that we can develop. And ourselves and control our own future. speaker-0: Yeah. It's it's nearly impossible to keep incentives fully aligned all the way through over multiple years ⁓ and priorities change over time, et cetera. ⁓ yeah. So so one of the places you've been in ⁓ advancing quite a bit is in spine and and you've received a FDA breakthrough designation in in spine. Beyond that that headline of receiving breakthrough, what does this change for for canary and and what you're examining in the spine area? speaker-1: Yeah, so we have have thought a lot about what are the right markets and and one thing I've learned about market research is is that it's very product specific. I d I don't think it's a a formula ⁓ as much as some of the consultants in the world might lead you to believe. ⁓ and what we've learned about our technology is that You need a device that has a reasonably high failure rate. and part of that is mathematical. And if something fails 50% of the time, it's gonna take far less patients to start to figure out why than if it fails one percent of the time. The second, of course, is clinical need. Like, you know, people are gonna pay more and be more interested in in procedures that have high patient dissatisfaction rates and And you know, there's there's more pricing pressure there because complications cost money and that gives you the ability to prevent complications. I mean, the reason we sold so many stents is we took the failure rate from you know 25% down to 1% and and the cost of failure was very, very high. And so we could capture some of that value in the pricing of our device. So spine is attractive because you know 40, 50 percent of people are unhappy with their outcomes. secondly, you know. as a spinal surgeon said to me, you could send the same patient to eleven docs and get eleven different diagnoses ⁓ on and and twenty different surgical approaches. And and so it's also interesting because having objective data to help standardize procedure or approach or things like that are are quite valuable. You'll see in in some of the abstracts that that we just submitted in in TKA where we're starting to look at if I do the surgery this way versus that way, what what happens, you know. so those are are kind of where you want to go. Spine really ticks all of those boxes. ⁓ we have breakthrough for hip, knee, shoulder, spine, and revision, but we went into knee first because knees fail a lot more than hips. ⁓ and so, you know, most hip patients are actually really quite satisfied. So going after, you know, that group would be a lot more difficult. And to me, spine is kind of one of the holy grails of orthopedics because, you know, it it really is a procedure that's obviously very, very necessary, but there's still a lot of question marks associated with it. speaker-0: Yeah. So these domains and spaces where maybe it's primarily driven by art and finding a way to make it science, taking the subjective and making it objective. speaker-1: Yeah, it you know, once we got to drug coded sense, it was hard to go higher. I mean, once we got to kind of one percent failure rates, like you you're talking pretty big studies to go to half a percent or or a quarter, right? Whereas, you know, when we went from come bare male stents, we had a it was a a much bigger target ⁓ to go after. And so as I look at where we can go with these, because quite honestly, almost every medical device there's probably an engineering way to attach sensors to it one way or another. it ⁓ you know, you need to look for those areas where there's a a significant dissatisfaction or unmet medical need. speaker-0: So after everything, Angiotech, Canary, more than two hundred patents and applications, what's a problem in medicine that you're still eager to solve? speaker-1: You know, I I think this will be my last rodeo, although I've been wrong every time I've I've said that before. I I'm really enjoying this. It it's been really, really hard. We talked about this when I first met you. ⁓ funding in med tech has been incredibly difficult ⁓ the last little while. you know, the med tech companies ⁓ when I first started out, there was a whole group of middle class companies, right? So there were there were companies, you know, you had the multinationals and the huge companies, but you had 20, 30 companies that were kind of mid-sized. And and what that meant is that there was a lot of customers for new technologies and innovation. And, you know, you you could partner with kind of a mid tier company, you know, that was ambitious and and they might partner with you fairly early because they wanted to get to the technology before the big boys got it. And And the ecosystem was was much more evolved. And now we really just have a small number of very big companies where as a medtech entrepreneur, you you know, you now have to not only get approval, but you usually have to build revenues and you have to have a significant amount of of value before you've reached the threshold where you're interesting to a Boston Psy or a Medtronic or a Zimmer or a Striker. Anyways. With that in mind, having just been through in that through that again in my career and seeing how hard it is to go from idea to whatever, I I'm pretty excited about this concept of linear data and and so I'm pretty excited about the concept of of hopefully being entrepreneurial within Canary and and going into different indications and and going into different fields as opposed to you know, coming up with a a big idea and starting from scratch again. I I think I think the wounds of the last twelve years are too deep. and, you know, I I I do feel like this whole thing has some run in it in in terms of day. I know you feel the same way and and that that's that's kind of a a next logical frontier in medicine. I think I'm gonna hang on for dear life and see how far I can go with it. speaker-0: Yeah, no, I absolutely agree. And I I think the applicability of linear data over so many different indications can be so so powerful. So if we were to fast forward fifteen years from now, is there an aspect of medicine that you think will be commonplace that maybe sounds a little futuristic today? speaker-1: Yeah, you know, one of the things and this may or may not sound crazy, but I and you know, it's it's it's not Schwarzenegger and Terminator, but I I I do have this kind of concept as as we we think about hardware becoming sensor enabled and communicative. And you know, as you get older you accumulate more and more hardware. ⁓ and I'm I'm quite fascinated by this this concept of almost like a central node, a central computing node that communicates with multiple aspects of of the body. ⁓ you know, one of my professors always said to me that the problem with patient compliance is the patient and and these wearable technologies are all very, very good and and and very, very interesting. The problem is anytime you ask somebody to do something, you you get a drop off. But I think monitoring will become less invasive and less involved. And I I have this this concept, or no, I don't it's not my concept, but I have this feeling that, you know, continuous blood pressure monitoring and respiratory and all these things are going to become commonplace and relatively small subcutaneous implants will be able to communicate with each other across the body and and and and provide kind of more holistic information going forward. And and so yes, we're gonna reach a s sense of data overload for sure, but it's also gonna be correspond to data processing capabilities that, you know, are probably beyond our imagination. And I do see this this kind of monitoring continuously of multiple things across multiple parts getting more and more predictive. I remember reading a book years ago at by Eric Schmidt ⁓ at Google and and he near the end of the book asks the question like what will happen to your business when computing power becomes almost infinite and almost free? And we're getting close to that now. Yeah. And so I I do think ten years from now, you will see an awful lot of data gathering. in a many different forms for well patients, not just you know unhealthy patients. And you will see a lot of computing power and a lot of insights being generated you know, in the field. ⁓ you know, when I was a kid, I I learned how to fix my old TR6 so that every time it broke down I could repair it on the road and continue to to drive. Now I take my car in and it gets plugged into a laptop and the diagnoses are done the same way. I I think medicine will evolve a little bit along that that timeline where, you know, there'll be a a kind of a data source that is used routinely when a patient walks into the emergency room. ⁓ you know, that was the vision behind EMR was that we'd we'd have this, you know, this continuous data set. Well, I think we'll have a matching data set of kind of continuous monitoring and continuous physiologic insights that will plug in when you walk into the ER that will will help fill in the gaps diagnostically and prognostically. speaker-0: And I think then you have a a data interpretation layer on top of that is incredibly powerful. speaker-1: Yeah. And, you know, ten years ago, you know, I I mean, what did it how how many billions of dollars did it take to sequence the human genome? You know, in in two thousand, right? Now it's a couple hundred bucks. ⁓ so so ten years from now, all that genomic information, all that, you know, hereditary information, all that predictive information, I I think will be carried around with us. and for those of us in MedTech, I I do believe that direct source feedback will become very, very important. I mean, the the one thing about med tech is that it is very much a site specific solution. Like when you make a pill, it it goes through the entire body, but but when you put in a coronary scent, you're only concerned about that much of artery. That's it. That's the only thing you care about is what's going on in that part of of the artery. And so I think in med tech, the expectation that there will be data sources, accurate data sources, diagnostic data sources from pretty much everything that we s put into human beings, I I think that has to become a reality in the next ten to fifteen years. speaker-0: Absolutely. So we started the conversation and you gave a lot of great insight about how important it was that you were kind of almost bedside as you started to evaluate drug looting stents. We have a lot of clinicians who are considering being entrepreneurs who who listen to this podcast. Any advice for other clinicians who are considering moving into a career focused on building innovations? speaker-1: Yeah, I mean I mean I my oldest child, my my daughter's an R2 in in emergency medicine right now. And ⁓ she's been very, very interested in looking using AI to look at at patterns of when people come into the emergency room and you know, they're they follow very specific economic patterns, socioeconomic patterns and and different things as well. And You know, the advice I I give to her and and to clinicians is that continuing medical education was always important, ⁓ clinically. But if you have an entrepreneurial bent, ⁓ stay apprised of of what's happening in all the other things, whether they be engineering or computer science or whatever. And and what I think has changed a lot since when I went through medical school is I'm I'm seeing a lot of people who finish their medical degrees, then do a graduate degree and then do a residency and and they're doing things like AI and computer science and all kinds of other things. And so long winded way of saying if you have an area of interest that you think can attach to medicine, pursue it. get involved with it. ⁓ the way that it it's been done, at least for the foreseeable future, I think it's going to continue to be that way, which is that, you know, if you can affiliate with solid researchers outside your field and you happen to be able to link into an academic institution that gives you access to, you know, great minds in the area of expertise that you need, you know, collaborate, get involved. I think it's going to be a lot harder to be an idea doc who just goes in the garage and and ⁓ you know, comes up with a a new catheter design or something, that will still happen, but I think that's getting a lot harder ⁓ than when I started. And I think it's now becoming more of these integrated teams. And as a physician, you bring a very important piece to that puzzle. You you bring that that clinical view, that what works that what doesn't. Workflow is incredibly important. I don't think if you didn't practice medicine, I don't think you could ever understand the importance of workflow or how to deal with it. And so, you know, I think these calling kind of multidisciplinary teams, even though we've been talking about those for 25 years, and I I I think I think though they are becoming more and more important. And I think you should honor your medical training and and understand the value that it brings into that ⁓ and and you know use it to collaborate and get involved with these these programs as they come along, because I think the next 10 years will be absolutely explosive for medical development. More than more than I've seen in the thirty years before. speaker-0: Absolutely. Well great perspective and and great advice. It's been a a tremendous conversation. Really appreciate you coming on and sharing your perspectives and your experiences. And I know everyone in the audience really appreciates it as well. speaker-1: My pleasure. It was a lot of fun. It was great meeting you at ⁓ L S I and and and all that came out of it. So appreciate you for having me. speaker-0: Absolutely. And we'll be following the the canary story here very closely. And thanks to all of you for joining today on Powering the Future Within. If this conversation inspired you, please share it with your med tech community. Stay inspired and keep innovating.