Engineering arrhythmia therapies with Jon Silva

YouTube video

On this episode of Engineering the Future, Jon Silva, professor of biomedical engineering and of computer science & engineering, discusses his work at the intersection of cutting-edge technology and cardiac care. Tune in to learn more about Silva’s interdisciplinary research in treating cardiac arrhythmia using augmented reality and his entrepreneurial journey from lab discovery to market application. Plus, get Silva’s insights on the future of biomedical engineering and a few sci-fi media recommendations!

Jon Silva: So the intersection of that technology, being able to connect to a computer with kind of a headset, along with AI that might inform decision-making during these procedures, I think it's going to be very interesting. I don't think we're going to have AI doctors where the AI takes over a patients' healthcare, but I think decision support will be very valuable.

[Music]

Shawn Ballard: Hello, and welcome to Engineering the Future, a show from the McKelvey School of Engineering at Washington University in St. Louis. This season our theme is Engineering Human Health. I'm your host, Shawn Ballard, science writer, engineering enthusiast and part-time podcast host. Today I am joined by Jon Silva, who is a professor in the Department of Biomedical Engineering as well as Computer Science & Engineering. Welcome, Jon!

JS: Thanks for having me.

SB: It's great to have you here. I want to dive right into your research. So, you're focused on creating better ways to treat cardiac arrhythmia, and I will confess I'm not totally sure what that means. Can you unpack that for me? What is cardiac arrhythmia, and how's it currently being treated?

JS: Yeah, so the heart's an electrical organ, and the electrical signal is what tells the myocytes in the heart to contract.

SB: What are myocytes?

JS: Little cells that squeeze together and cause the muscle contraction either in skeletal muscle or in cardiac muscle.

And so that electrical signal needs to be organized so that the myocytes can contract efficiently. So if it becomes disorganized, then the heart doesn't work well and the rest of the body doesn't get the blood it needs, and that can be deadly. 

SB: Okay. So that's what an arrhythmia would be when all that stuff is sort of not firing in the right way or it's unorganized. It's not firing at the same time, I could sort of imagine.

JS: Yeah, instead of being organized from the top of the heart and going down to the bottom like a tornado can start. If that tornado starts moving around, then the contraction of the heart becomes much less efficient, and it can't complete what it needs to do.

SB: Okay, so treatments for that… is that things like pacemakers or something else to fix that sort of electrical disorder?

JS: There's a few different treatments. Our lab a lot of time thinks about anti-arrhythmic drugs. Another treatment is a cardiac ablation, where a catheter is fed up into the heart and then pathways that the arrhythmia travels down are burned so that the arrhythmia can't travel those paths anymore, and that can stabilize the heart rhythm.

SB: Wow, okay. So you're working on finding better ways to fix these things. So you mentioned drug interventions as well as this kind of ablation. Can you tell me more about that? Like what is your research really focused on in terms of those interventions?

JS: Yeah, so we have we have a couple different directions. So one was improving ablation technology. So I worked with my wife Jennifer Silva. She's a physician who practices and who actually performs ablation procedures. And so we were talking about different ideas that could improve what she's doing in the cardiac cath lab. And one thing that came about was putting a hologram in front of her. 

So instead of right now, she kind of looks at two projections. And so it's kind of like trying to understand your hand in terms of looking at it two ways. And imagine you haven't seen a hand before and you're like, okay, it's very narrow and it's got five fingers. And that process of understanding where they're moving their tools in the heart based on two different pictures is very challenging. 

So we thought using new technology, specifically these augmented reality headsets – that’s  like the new Apple Vision Pro or the Microsoft HoloLens or the Magic Leap display – could provide her a 3D image of what she's looking at. And so that's that was the basis of our company.

Then in the lab, a lot of times what we're looking at is using engineering approaches to understand how changes at the molecular level cause functional changes at the whole heart level. So we use experimental tissue engineering approaches to make a model. And then we also use computational approaches. 

SB: Okay. And what are what are some of the things you're hearing from our collaborators over at, I can imagine, the med school, right, in terms of what computing and imaging techniques they're really excited about?

JS: Yes. So I think physicians are always interested in knowing where they are better, being able to target better, any improvement like that. I think there's been a couple of developments at WashU. One is using radiation to do these ablations. That's been something that physicians and researchers at the med school have been working on really hard. From our point of view, showing physicians where they are and helping direct their therapies more accurately is something that has been very beneficial.

SB: Okay. So you've mentioned a couple of times your company that you launched and that's SentiAR. Am I saying that right? Okay. SentiAR. Can you tell me more about that? Sort of, obviously a big part of this, you're developing these therapies. You're working with physicians to find out, you know, what interventions are going to be most impactful? What do they really need? But the step from research lab to actually on the market, that's quite a big jump for some people. But you've made that jump. Tell me more about that.

JS: There was a lot of learning involved with that. And I think we thought of this hologram project initially as a science project. So building the hologram, getting traditional funding for it. But we needed a partner. So we needed to collaborate with a big company to make it happen. So before we told the big company our idea, we talked to OTM at WashU and they provided a lot of help and connections. We also did the, it was called LEAP back then, but kind of a competition that they sponsor to get seed funding for companies. And so that process really helps make the introductions that made launching the company possible.

SB: All right. So that's the Office of Technology Management here at WashU. So, you know, they're, they're making introductions. They're setting up, you know, collaborations, getting seed funding. What other parts of the process were they, you know, you have sort of idea, then funding, collaborators, what, what next? Walk me all the way through.

JS: Yeah. So the first thing was to get funding, and then we used that funding to get the prototype a bit farther in the university. And that allowed us to hire a professional programmer who helped to co-found the company and also really got us a head start in the university. But then we ran into a big challenge in finding a CEO. 

So it's hard to find someone who will work for free and try and raise money to pay themselves. So if you don't have money, you can't, it's hard to find a good CEO. And then conversely, it's hard to raise money if you don't have a CEO. There's a big chicken and egg problem, kind of right when you're trying to start the company, where you need investment, but you need a CEO to lead the company and finding that connection is really tricky. And so we had help in BioGen, with BioGenerator and with the community in St. Louis and kind of making that happen.

SB: Okay. So that's another collaborative partner we have here. So I know the Office of Technology Management was involved in that. And then BioGen, they're over at the Cortex, is that right?

JS: Absolutely. Yes.

SB: And what's their role in sort of this company development or entrepreneurship?

JS: Yeah. So we had a lot of meetings with people at BioGenerator. They set us up to meet with different investors. We kind of gave some of the initial pitches there. We also met with an entrepreneur in residence regularly during that time to kind of crystallize the ideas and what the value proposition was for the company. What the market size was and everything else.

SB: Which seems like potentially large from what you've told me, right? It seems like any physician who's doing this kind of like tricky work inside the heart would want to be able to see better.

JS: That's what we're hoping.

SB: That seems right to me. I know if I were having that sort of work done on my heart, I would want the best picture of it, you know, short of like actually opening it up.

JS: Yeah. So that's something that the augmented reality helps with. Studies in the 70s found – that’s when these minimally invasive procedures were starting to be used – that open-heart surgeons or surgeons who opened up the body had less mental fatigue than the ones who were using these minimally invasive approaches. So it really is something that's pretty taxing for them.

SB: Well, that's an amazing journey from sort of lab to getting this out in the hands of physicians. How did you get interested in this question in the first place, right? You've put a lot of effort in this. What drove you to be, you know, really working on this problem?

JS: So I think my initial interest started when I was an undergraduate and the professors related the membrane of a neuron or a myocyte to resistors and capacitors. So I thought the analogy that you could think of the cell membrane as a capacitor, you can think of these proteins that let charges go across as resistors and you can make this this cool model to describe it. I found that very interesting. And that's what I did for my PhD. So making these resistor/capacitor-based models to think about the cardiac action potential, and it kind of built from there. Then I met my wife who's a physician and those cardiac procedures. So that really cemented that this is the direction that we're going to focus on.

SB: Okay, and when you were an undergrad, what were you studying? So I studied physics. So when you hear “it’s like resistors and capacitors,” you're sounding very physics oriented, but then you have the biology in there. Was that all in biomedical engineering or something else?

JS: It was. I actually started as a biophysics major. And then I switched to biomedical engineering once I heard about kind of this approach. I thought that was very interesting. 

SB: Okay, cool. So it seems like that kind of interdisciplinary focus has been there from the beginning for you, really thinking about what you can borrow from other disciplines, your other colleagues, collaborators to really bring all that to bear on a shared problem. Is that like a through line for you? Something that you find motivating even now?

JS: Yeah, absolutely. And I would say it's a little bit disconcerting how monumental the decisions you make when you're 18 are. How they take your life in a certain direction. But it was basically, I wanted to, I was pre-med at the time, but I didn't want to stop taking math classes. I thought biophysics would be a good way to do that. And then biomedical engineering was similar, but I really kind of liked these circuit analogies.

SB: All right, so you're sort of, you know, come out of undergrad, you've done these various, you know, multidisciplinary kinds of things. And then you're moving into this more cardiac-focused work. And I understand that some early work in your lab involved frog oocytes. Am I saying that right? What are those and what were you up to with frogs 

JS: Yeah, so that's a great question. So for my PhD, I worked on computational models. And one of my big frustrations was not being able to test kind of the hypotheses that came out of those models directly. So for my postdoc work, and early in the lab we used this, and we still use it, we use the oocyte model to watch the channels move. So there's these proteins that create the electricity in the heart. And we can watch those proteins move using a technique called voltage clamp fluorometry. And so you put a fluorescent molecule on the protein and when that molecule moves, then you can track its dependence on the potential across the membrane. So the trick is, is that you need a lot of channels. And so you need a protein making factory, which is exactly what an oocyte is.

SB: Oh, okay. Yeah, I was going to ask, what does that even mean? It just makes proteins.

JS: It makes a ton of protein.

SB: Okay.

JS: And so when we put the message in to make the protein or the instructions, it makes a ton of them. And then we can actually measure those very small changes in fluorescence that we need to watch the channels move. So that, a lot of time informs the basis of our models. It tells us at the molecular level what's going on with drug interactions, what's going on with genetic variants that predispose patients to arrhythmia. 

Our motivation is creating a quantitative model. So being able to say, okay, if this moves this much and it's interacting with all these different things, what's the emergent behavior that happens at the cell level, at the whole heart level? That determines whether the heart rhythm is stable or not.

That molecular data allows us to build a model that gives us the big picture. So, you know, you can only look at one thing at a time with a very powerful microscope. But the computational models allow us to put all that data together.

SB: Okay. Gotcha. And so as you're, you know, you've developed these various projects, building up to these larger, you know, complete models, what advice do you have for maybe aspiring biomedical engineers? 

JS: Yeah. So one thing I tell my advisees a lot is to get good at one type of engineering. And it doesn't matter necessarily which one it is, but if they get good at electrical engineering, and they connect that to biomedical engineering, then that's very powerful. Or computer science or chemical engineering. Really developing a skill set in a core engineering discipline and bringing that to biomedical engineering, I think is really helpful.

SB: Was that part of how you got to doing what you're doing now as well?

JS: Yeah, it's interesting because when I was an undergraduate, I really liked computers. I was good at programming. And my advisor told me that computer science wasn't real engineering. And I should do electrical engineering instead. And I said, Oh, all right. And I was no good at electrical engineering. And I don't do it anymore. I do computer science. So I think it's important to stick with what you like and what you're good at.

And I also think it's very destructive when people say that's not real engineering. So they say that about biomedical engineering, they used to say that about computer science. Engineering is engineering. I don't think we need to draw a line around it.

SB: Right.

JS: That's very limiting.

SB: I agree with that. Do you – I’m going to put you on the spot here – a quick definition for engineering that's broad and the way I think of it is, you know, what we do here in McKelvey is all about solving problems, right?

JS: Absolutely.

SB: We're working on solving problems that impact real people. Is that a fair sort of statement of what engineering is all about? Or would you add that? 

JS: Yeah, I absolutely agree with that. I also think you can kind of go two directions. So you can use engineering to build a device or to build a new algorithm. And I think that's how people usually think about it. But a lot of engineering is probing a system and understanding a system. So using quantitative approaches to describe a black box. So kind of reverse engineering. I think that's a lot of what we do in at least in the anti-arhythmic drug side of the lab is we're really trying to probe this back black box. We apply pulses. We apply pulses of different magnitude, different pulse trains. Then we create a detailed model and understand how the control system is working there.

SB: Could you dig into that more? So it's the black box, like how the heart works?

JS: Yeah, exactly. How the individual myocytes are functioning. So we know that these proteins move in responses to changes across the cell membrane. But we really, if we want to quantify it and build a model, we need to say, okay, how fast does it move when you apply a certain stimulus of a certain size? What if you apply repetitive pulses? What if you go up and then down? What kind of memory does the system have? And so it really is, you have this protein and it moves really fast. So how can you understand how it moves and how can you understand how changing that protein either through a drug therapy or a genetic variant might cause problems or stabilize the heart? 

SB: Okay, yeah. So when you're sort of talking about reverse engineering, understanding the black box that is the sort of heart system, seeing all those things and how they move lets you know like, okay, well, this is how this is built. This is how this is working. Now we have our full model and we can, I guess, then apply different things to that right – different drug therapies or different ablation techniques. Is that sort of what happens next? Have I followed that thread through properly?

JS: Yeah, I think it's interesting because it’s changes at the molecular level that often cause problems. But then we apply therapies at the whole heart level. But we don't necessarily understand how these molecular-level changes connect to the top. So I think a more sophisticated way of dealing with things. So I'm not going in there and burning everything.

SB: Right! As you're looking ahead to seeing this in use, what challenges are you looking at right now?

JS: So I think one thing that's underappreciated is proving the market. So first, I think when you come across an idea, you're excited by the technology, you're excited about the difference it could make with patients. But one thing that's super important for a company is being able to sell it and making sure that the effort to develop a certain thing was worth it. Because if it's, if it can't be commercialized, then you spent all this money and you develop something that nobody's going to ever use. So this commercial viability is something that we don't necessarily think a lot about in academia. But if we want to turn our technologies into something that's used, it's really important. So I think estimating what a market is early on and whether something can be impactful is really important.

And then I think one thing that's changed probably in the last 10 years is big companies aren't buying small companies based on the technology they developed anymore. Probably in the last five years, you have to actually start selling it and showing that people will buy it. So the kind of whole cycle of investment to creating a product, now you need to go one step further and show that people are willing to buy it.

SB: For you, being part of a startup company as well as a researcher, was that personally difficult? I can only imagine you're sort of really shifting gears between the lab work you're doing, even with the help and with the partnerships. How was that for you sort of deciding to take that leap?

JS: Yeah, so my family has been involved in quite a few companies. My grandfather had a construction company when he was younger. I have uncles who have web hosting companies and printing companies. My father started a company, but they all took on quite a bit of risk to do that. You know, mortgaging the house, and sometimes it didn't work out, which was very traumatic just in terms of the whole family. So being able to launch a company from a university, not having to quit my day job or mortgage the house, that's been something that's been pretty incredible. So I've always been interested in entrepreneurial work but I think the opportunity to do it from a university is pretty exciting.

SB: Okay, is this kind of entrepreneurial mentorship something that you incorporate as you're working with your students here at WashU?

JS: Absolutely. So I've had a couple students, so I have one postdoc who actually works in the company now, so he's commercializing the technology he built in the lab. And I also meet fairly regularly with students who are interested in starting companies, given my perspective. And a lot of times it's a reality check. I think it's sharing our experiences that made us understand that it was much harder than we initially thought. A lot of times convincing people that they need a professional CEO, they shouldn't just kind of try and run with this themselves, they need it. You know, it's possible, but it's very difficult, especially in the medical device space, to overcome the barriers of developing something with a quality system and doing it the way the FDA needs it done.

SB: Right, all of the steps, all the stages required are things that I feel like that's not really covered often in undergrad, right? You have to be working with someone like you.

JS: Yeah, I think so. I think we teach a lot of fundamental concepts, but there are some practical aspects that you can learn. It's almost like you need to learn outside the university because it's done different in different fields and different practices. 

SB: So you've been tackling all these kind of challenges that I wouldn't have expected, right? You maybe did as an experienced entrepreneur.

JS: No, we had to learn a lot.

SB: So you've been tackling all these things. How have you seen those efforts already making a difference in terms of positive impacts for human health?

JS: So one thing that we saw is there was a trial using our holograms at the Harvard system. So at Mass General and Beth Israel. And what we saw was that physicians were 40% less likely to miss badly with their catheter targeting. So that's like a miss of more than four millimeters, which is about how big the catheter is. And we saw that with the hologram, they ended up, they would miss by a little, like maybe one or two millimeters, but they would miss by a big margin a lot less. We're kind of excited about what the implications of that might be going forward. And that really showing the physician a 3D representation probably is helping.

SB: Okay. I want to get a little more sense of that four millimeters. So that's small. The size of the catheter, you said, what are the size of the like the catheters going up into the heart? What are the size of the channels that it's going into?

JS: Well, so it goes in through blood vessels. So, you know, four millimeters is a small amount, but also heart structures are pretty small. So if you move four millimeters, and it's on top of the structure that connects the atria to the ventricle, then now you're going to be pacemaker dependent for the rest of your life. 

SB: Oh my gosh. Okay, that's terrifying.

JS: Yeah. So they really want to know where they are. 

SB: So you're seeing already this making a difference in patient care. Looking ahead, what emerging trends, perhaps both in cardiac arrhythmia, but other areas of biomedical engineering, are you most excited about? Like, what are you watching for what's next?

JS: I think there's going to be quite a bit of development, both in AI and also in these augmented reality technologies right now. Those glasses are really bulky. But imagine if someone could make some AR glasses that were actually as big as the glasses that you're wearing, just a little bit bigger, then you could access your computer at any time looking around, overlay information that you're interested in, say prices at a grocery store, but for physicians that could be really interesting. 

So the intersection of that technology, being able to connect to a computer with kind of a headset, along with AI that might inform decision-making during these procedures, I think it's going to be very interesting. I don't think we're going to have AI doctors where the AI takes over a patients' healthcare, but I think decision support will be very valuable going forward, really using collective experience that informs very sophisticated models to suggest to physicians what should be done next. I think that's going to be really interesting.

SB: Yeah, and I guess, are you willing to speculate about how long we might be before we get to something like that?

JS: Yeah, I think it's going to be fairly quick. I kind of see these kind of big machine learning models evolving very quickly. I think it not that long ago, you couldn't ask a ChatGPT like question and get anything besides nonsense out of it, right? Or anything, it would be super trivial. And now I do a Google search and it kind of tells me what I was looking for and it summarizes the sources. It's pretty good. So I think using similar models that have data from the health records, which are enormous, there's an enormous amount of data in health record, and coming up with a way to do that safely so that people's privacy isn't violated, but also using that information to guide therapy, I think is going to be very powerful.

SB: Okay, it sounds promising. I'm glad you specified that it wouldn't be AI running the show, because I definitely, all the things you told me about like heart you know, ablation, I would prefer a person to a computer doing that still, I think, but definitely having access to the masses of data that are out there, that obviously AI is just better equipped to parse that quickly. 

JS: Absolutely, yeah, I think there's so much information out there, we're going to need computer help to really find patterns out of it. 

SB: I'm looking forward to that and I will definitely follow up with you in a few years about how this is going. All right, Jon, so in closing, I always love to ask people for a recommendation for either book, movie, TV, something. What we just talked about makes me hope there's some kind of like sci-fi angle to this, but do you have a favorite media representation of biomedical engineering or you know, something connected to what you work on that you've seen that and been like, oh, that's the thing, or that's terrible, that's not how it works at all. I'm equally open to these.

JS: There's a few terrible, it doesn't work at all, but I think the one I'd like to talk about is a movie, it's an older movie called Gattica, that I think I'm going to make required watching for my class because it talks about how basically a heart variant is supposed to cause an arrhythmia in this person, and he's being discriminated against compared to his edited counterparts. I think this was very forward-looking. I think that those variants are the exact things that we study in the lab, and I think humans are going to start editing themselves very soon. The therapies to edit out genetic variants that predisposed to some pretty bad pathologies are already approved. And so I think this is going to accelerate in the next 10 years and having a media representation of what can go wrong, I think is very valuable.

SB: Yeah, it's been a while since I saw Gattica, but I feel like that's not like a feel-good story, right?

JS: No, it's not.

SB: Okay, so if you want to look in terms of maybe what we should not be doing as we're moving forward with this.

JS: Yeah, my wife says as long as it's post-apocalyptic, I'll love it, and that's true. Mad Max, I thought was great. And Silo, I think, is another show that I'm enjoying quite a bit.

SB: Amazing. I'm also watching Silo right now, so these are great recommendations. Thank you so much, Jon, for sharing those and for being on the podcast today.

JS: Sure, thanks for having me.

[Music]

Click on the topics below for more stories in those areas

• Research
• Podcast: Engineering human health (2025)
• Biomedical Engineering
• Computer Science & Engineering

Back to News