The heart’s electrical system keeps all its muscle cells beating in sync. A hard whack to the chest at the wrong moment, however, can set up unruly waves of abnormal electrical excitation that are potentially deadly. The resulting kind of arrhythmia may be what caused the football player Damar Hamlin of the Buffalo Bills to collapse on the field after he took a powerful hit during a 2023 National Football League game. Today, powerful defibrillators are usually used to help resynchronize hearts in distress. But Flavio Fenton, who studies the electrical dynamics of the heart, tells Steve Strogatz about a new method under development for treating arrhythmias by stimulating the heart with mild, precisely timed shocks — or possibly even with light.

Listen on Apple Podcasts, Spotify, Google Podcasts, Stitcher, TuneIn or your favorite podcasting app, or you can stream it from Quanta.

Transcript

Steven Strogatz (00:03): I’m Steve Strogatz, and this is The Joy of Why, a podcast from Quanta Magazine that takes you into some of the biggest unanswered questions in math and science today. In this episode, we’re going to ask how can we use math and physics to stop deadly heart arrhythmia.

(00:21) You may remember the horrifying scene that occurred during a recent pro football game, when Buffalo Bills safety Damar Hamlin collapsed on the field after taking a thunderous hit. One theory is that the slam he took to the ribcage disrupted the rhythm of his heart, causing its normal electrical waves to go haywire. The resulting condition known as ventricular fibrillation can kill someone in a matter of minutes because it stops a heart from pumping blood effectively to the body and brain. And as Damar Hamlin’s stunned teammates and millions of TV viewers looked on for what seemed like an eternity, medical personnel struggled to revive him.

(01:01) The instant Flavio Fenton saw the footage of the hit, he knew what had happened. Fenton is a professor in the School of Physics at Georgia Tech, and cardiac arrhythmias are his specialty. Fenton studies mathematical and computational models of arrhythmias and the strange spiral waves that underlie them. And he also conducts experiments on animal hearts and on donated human hearts. He’s hoping to find a way to stop arrhythmias without having to use traditional defibrillator paddles that send a huge blast of electricity through the patient’s entire body. Instead, Fenton is trying to fight waves with waves. He’s making waves of his own to snuff out the pernicious spiral waves that can send a heart into disarray. The goal is to find a gentler, less damaging way to treat arrhythmias.

(01:50) Flavio, thanks for joining us today and tell us about the amazing work you’ve been doing.

Flavio Fenton (01:55): Oh Steve, thank you so much for having me. It’s a pleasure to be here.

Strogatz (01:58): How does the heart work when it’s working properly?

Fenton (02:01): The heart is an amazing system. And one of the things I would like to say, that’s one of the things I think separates us a little bit: how we investigate the arrhythmias is, I’m trying to do it from the point of view of a physicist. Most people who investigate cardiac arrhythmias are biomedical engineers or cardiologists. So we tried to do it from the point of view of how physics works in the modeling of the heart.

(02:21) The evolution of the heart in different animal species have different ways of how the heart work, but the main point of them is to contract. So they try to contract so they can eject blood and circulate oxygenated blood to the body. Mammalian hearts, we have four chambers. We have two atrias and two ventricles. They’re connected, but they’re electrically disconnected.

(02:40) So the way it works is you have some cells that are auto-oscillatory, they’re called sinoatrial node cells. They start the beating of the heart. And then by diffusion propagates through the atria, and then from the atria goes to the AV node which is the only part that connects the ventricles from the atria. And then goes into the ventricles, and then the ventricles contract. The atrias are basically receiving chambers. So the blood receives into the atria and then the atrias send it to the ventricles, which sends the blood to either the lungs or to the body. So the one that sends the blood to the body is the left ventricle. So that’s the most — thicker part of the heart, it’s the powerhouse of the heart.

(03:16) The main thing is the contraction, which is a fluid dynamics problem, but it originates from an electrical signal that makes the cells contract. Because cardiac cells, they have a membrane that separates the inside from the outside. So there’s different concentrations of ions between the inside and the outside. So at rest, they are depolarized in general at about minus 80 millivolts or minus 75 millivolts. When they get excited, the voltage goes above threshold, like about 10 millivolts. So there’s an amplitude about 10 millivolts, that changes the voltage. And when the voltage is above threshold for about 200 milliseconds, calcium gets released into the cells. And calcium is what produces then the contraction. So the contraction is driven actually by an electrical signal.

(04:00) So when you try to investigate how the heart fails, there are multiple ways in which hearts fail. There are some mechanical and some electrical. So we always joke that when we have people who study the dynamics of the heart and the arrhythmias of the heart, you can separate those into electricians and plumbers. So I am mostly an electrician. So I’m most interested in how the electrical disturbances initiate arrhythmias, but it can be also mechanical. And we try to work together to combine together, but the studies that I investigate are the ones driven by these disarrangements in electrical propagation.

Strogatz (04:35): Good. I’m glad you make that distinction between the electrical and the fluid mechanical or plumbing aspects of the heart, because I find that when I’m listening on TV, or just hear people in conversation, maybe they have a relative or a friend, and they’ll say, “This person had a massive heart attack.” Or they might say “that person had heart failure.” Or then you hear the phrase “cardiac arrest.” So I think in the public’s mind, all three sound like something that you don’t want to have happen to you. But they’re not the same thing. But let’s hear about like, what do you mean by a heart attack versus heart failure versus cardiac arrest?

Fenton (05:18): So, when you have, for example, a heart attack, what happens is that when the heart contracts, it sends blood to the body, but it also sends blood to itself. So at the base of the aorta, where the blood goes to the body, there are two arteries that start and go down through the whole heart. And when the heart pumps blood to itself, it oxygenates itself. That’s how the heart keeps itself alive. So what happens is that when one of those vessels gets blocked — by clogging, by when you have high cholesterol and then a vessel blocks — then blood doesn’t go to that section of the heart. So that section of the heart will not be oxygenated. It loses excitability, and then it can initiate actually an arrhythmia driven by the electrical conduction system (which I’ll tell you in a second). Depending on where the block occurs — if it occurs very low in the branches, then only a small section of the heart gets affected. If it happens very high up, then a large section of the heart gets affected and that section of the heart can die, stop contracting.

(05:18) And there are two causes that can happen when you have a heart attack. Either the whole heart stops contracting, or it initiates an arrhythmia, which is ventricular fibrillation. This arrhythmia can happen because there’s a section of the heart that is not contracting, doesn’t allow propagation of the waves. So the waves will start forming these complicated patterns that can form. That’s what happens when you have, basically, a heart attack.

(06:38) Heart failure is when the heart eventually starts to change in time, morphing such that it can get thicker, for example. There are many different types of heart failure, but the heart becomes thicker, and the contraction diminishes. So you cannot contract as well. So the ejection fraction diminishes, and then you cannot oxygenate your body well. And that requires different treatments, medicine, and in the worst-case scenario, then you have to have a heart transplant. When you have sudden cardiac death, that’s an arrhythmia that happens when you get these disturbances in electrical signal and initiate complicated arrhythmia.

(07:17) So basically, what happens is you have electrical waves that propagate. But these waves can be disturbed and produce spiral waves. You can actually have a spiral wave of electrical activity rotating around the ventricles or the atria. And those will make the heart contract faster. Because it turns out that these spiral waves, when they form in the heart, they rotate faster than the natural pacemaker. So they take control over the heart to a faster rhythm. And that’s what is called tachycardia. You can have tachycardia in the ventricles or in the atria, depending up and down on the chambers. The spiral waves, in general, they can destabilize relatively easy. There are many mechanisms that can produce that.

(07:57) And those are some of the things we investigate. So they don’t remain for too long stable, and they break into multiple spiral waves. When you have multiple spiral waves, every section of the heart will have little spiral waves that rotate really fast. But then they can be out of phase. So what happens to the whole heart is that now it’s not pumping, even faster. It’s just that every section of the heart is beating at their own phase. So the heart is just shivering. It’s not even pumping, so it’s just trembling. And it cannot pump any blood. So when there’s no blood being pumped, then you die within seconds.

(08:25) So the only way when you have a case like that, you have to come with a defibrillator and stimulate all the cells with a very large electric field that defibrillates the tissue. This external defibrillator, it starts with 150 joules, and it can go up to 300 joules. That’s a lot of energy to defibrillate. Because when you require this, to excite all these cardiac cells at once, so then you terminate the spiral waves. These electric fields, yeah, they’re really huge. And then they can excite all the body around yourself, all the muscles, so they can be quite painful.

(08:57) Just to give you a comparison, the energy that it requires to move a muscle is about 0.001 joules. That’s why it takes forever to lose weight when you’re on the treadmill. It requires a lot of motion of the muscles to lose a little bit of energy, to lose weight. That tells you how strong these electric fields are for defibrillation.

Strogatz (09:15): Thank you. That was a very nice tutorial. So I guess it’s clear from what you’ve said, we’re not talking about heart failure in this episode. And we’re not really talking about heart attacks, except in so far as they can — by killing a piece of the heart — they can set up the circumstances for rotating spiral waves or other electrical problems. I mean, that’s really what we want to talk about.

(09:36) So you’ve mentioned tachycardia, where the waves are making the heart beat so fast that it doesn’t pump as effectively as it would normally. Or in the worst case, fibrillation, which I have to tell you when I was a graduate student. I had a professor of biomedical engineering who took us to a medical school to actually feel a fibrillating heart in our own hands. And it’s a pretty unforgettable experience. And it’s very weird and slippery. Like you said, shivering or trembling. It feels like worms, right? It sort of feels like there’s all these worms wriggling around in your hand, as you put your hand on a heart.

Fenton (10:13): In the literature, a lot of people including Art Winfree, he used to mention that when you visualize the heart fibrillating, it’s like it was worms moving around in the substrate, right? And when you see a heart fibrillating, that’s how it looks: like worms underneath driving the structure of the contraction.

Strogatz (10:29): As you say, it’s extremely dangerous. You’ll die in a matter of seconds or minutes because the blood is not getting effectively pumped to the brain or body. But if we go back now to this case of Damar Hamlin, what do you guess happened to him, when he was hit?

Fenton (10:44): So we’ll never be sure of what happens, right? But most likely what could have happened — that’s what I thought at the very beginning when it happened — is that the way that you initiate the spiral waves in the heart, you break the symmetry of the waves.

(10:57) Let me start with a property of these excitable systems like the heart. Another excitable system similar to the heart is, for example, the fire. Fire gets excited and it propagates. It produces a wave that propagates. But you see, you can never burn that fire behind a wave of fire that passes, right? Because there’s no grass to burn. So when you have two waves of excitable system that crashes, like in the case of the heart, electrical waves in the heart, or two fire fronts, when they collide, they annihilate each other. It’s not like water waves, when they pass each other. These waves, when they collide, they annihilate each other. It’s easy to see in the case of the fire, because behind the wave of fire, there’s nothing more to burn. And that’s how the firefighters always say, “the way you fight fire is with fire.” It’s because in order to terminate a wave of fire, you use another wave in opposite direction that collides and terminates them. This gives you these — what is called a refractory period, that behind a wave, there’s a little time before you can excite again another wave. In the case of the forest fire, it’s a long time waiting because you have to wait until the grass grows again, so you can burn grass again.

(12:04) Another example of excitable system is a toilet. The toilet is a perfect example for an excitable system. You require a threshold of excitation. So when you move the handle of the toilet, you move it a little bit, and nothing happens. But if you flush, well, to put enough force on a threshold, you pass that threshold of moving the handle, then there’s a release of water. And then you cannot release water again because you have to wait until it fills again. So that’s a release in excitation after that. And you have to wait a little bit of time before you can flush again. So the same thing for the cardiac cell. Once the cell excites, you have to wait a little bit of time before it excites again.

So what happens — imagine you have a wave propagating, and behind the wave you want to excite. So if it’s just really refractory, you cannot excite there because the cell will not respond. But if you waited some time after the wave has passed, then you can excite, and you can produce a wave that propagates, right? Now imagine in between those times — between the time that you’re too close to the wave, or too far of the wave all the way back so you can excite an activation — in between there’s a region where part of the tissue will be refractory so you cannot propagate but part of the tissue can get excited. So that actually breaks the symmetry of the propagation of the excitation. And it can produce basically it propagates in one direction but fails propagation in another direction. And that’s how spiral waves will form.

Strogatz (13:26): Maybe you should give us a little visual. I mean, because I hear the word “spiral,” and everyone knows what to picture when they think of a spiral. But what is it that makes a spiral wave? Can you sort of walk us through it? I mentioned rotating spiral waves, tell me about how to picture what’s happening.

Fenton (13:41): Imagine you have a wave, just a wave of something, right? Like in the stadium, like you have the Mexican wave and you excite everybody. So now, imagine there’s a wave. So you have a front of people that are standing up, and then the rest of the people behind the wave, they stand up and then finally they sit down.

Strogatz: Yes.

Fenton (13:57): Right. So you have a wave with certain width. So now, just think about the front and the back, right? You have the front of the wave and the back of the wave, and think about it’s just propagating through the stadium. Now, imagine that you just break the wave from the bottom of the stadium to the top of the stadium, right? So you have a wave there. But imagine I only excite half of the stadium from the bottom to the top. So you have a wave front and a wave back. But it’s a continuum. So if you continue where the front and the back have to join, there’s going to be a point where we call the phase, right? The phase of the front and the phase of the back wave, they’re going to meet. And at that point where they meet, there is what is called a phase singularity, and there the phase is not defined. That’s exactly where the front of the wave matches with the back of the wave. So that’s when you create a spiral wave. When you break a front, and then you create that the back and the front meet, and then it’s going to start rotating around that point of singularity.

(14:52) And actually, we did that here at Georgia Tech just to show it a little easier. We collected 600 students and put them on a grid. And then we gave them instructions similar to the activations in the stadiums, that if your neighbor is excited with the hands up, you get your hands up. So if you start with one corner, you get a wave of propagating like in the stadiums. But what we did is, starting with a symmetry-breaking. In the case that, in the center of the square of students, we tell at the beginning there’s one line of students that are going to get excited, but not all the way to the top, just half of the domain. And then we tell them that the very first time, if you were on one side, you are going to be activated, but if you’re on the other side of these students that have their hands up, you keep your hands down. The first time. So that breaks the symmetry. So the wave is only going to propagate in only one direction. But the wave, as I said, it only goes from the beginning of the square of the students to the middle. This wave will start, it will actually produce a spiral wave with the students moving their arms.

Strogatz (15:51): [Laughs] Do you have a film of this? Is there some video we can watch on YouTube or something?

Fenton (15:54): Yes, there are a couple of YouTube videos on that. I can give you a link to those ones.

Strogatz (15:55): Send us the link, because I think we’ll link that in the program notes. So people can take a look at this. It sounds pretty dramatic.

Fenton (16:04): And it looks better when you accelerate, so we accelerate the video a little bit so it’s faster, so you can see the spiral wave. And this spiral wave will continue there rotating as long as the students have energy, right? As long as they can… That’s the important thing about these spiral waves: once they form, take over the system. And then, actually, one interesting thing that we observed there is that because the students are not always paying attention to — they’re a little bit sometimes not paying complete attention — so as the wave passes, sometimes they may get excited a little bit before or a little later. They say, “Oh, the wave passed” so then they activate later. So then that destabilizes enough that actually can break the spiral wave into multiple spiral waves. So we actually show how fibrillation can happen so easy by the destabilization of these activations between the cells, which in this case were the students.

Strogatz (16:50): You’re going on about this because of Damar Hamlin. What’s the connection?

Fenton (16:53): That’s right. So first is how the spiral waves form an arrhythmia, right? So the question is how a spiral wave formed in the case of Damar Hamlin. What we think happens is called Commotio Cordis, which is when the cardiac cells not only get excited by the neighbors, but they have ion channels that are stretch-activated channels. So that means that if I touch a heart, and I press on a heart, I can produce an activation. So you can stimulate a heart with an electric field or an electric shock. But if I press — in the case where you were talking about when you were touching a heart, if you had squeezed the heart, you could actually activate many, many, many cells in the heart, that could actually defibrillate. So that’s sometimes when they do in cases when they have open chest, and before they had electric shocks for defibrillation, sometimes massaging the heart can actually help initiate a wave or terminate an arrhythmia. But they had to be touched in the heart directly. But basically, anytime you stretch the cells, they can produce an activation.

(17:51) So when he got hit in the chest, it was such a strong shock that it actually deformed a little bit the front of his chest but also it was enough to perturb the heart and press the heart. And not only that but it happened in the worst possible time. In order to get the initiation of the spiral waves, as I said, it has to be exactly on the wave back at a particular vulnerable window when the wave is passing by and you get excited.

(18:17) So when you see in the hospital movies, when you have the ECG, the electrocardiogram, which is what shows you the electrical signal of the heart. You see a little signal that is called the QRS and then T wave, or the electrical signal of the ventricle. So what that measures is the whole electrical signal from all the cells as they propagate and get excited, is the global measurement of all the cells in the heart. So the killer is the big spike that you see first is activation of the heart. The wave that’s propagating through the heart that starts the wave. And the end of the wave is that T wave that you see on the ECG, the little, smaller hump at the end of the signal. That’s the end of the wave.

(18:57) So if you excite, if you perturb the heart exactly during the end of that wave, during the end of T wave, that’s when you can actually initiate an arrhythmia. So what happened is that he got hit hard enough, some of the cells get activated. And he got hit exactly when his heart was finishing activating during the T wave. And that initiated the spiral waves that then initiated fibrillation. If he had been hit just a couple of milliseconds later, 20 milliseconds later, or 20 milliseconds earlier, he may not have gone into fibrillation.

Strogatz (19:30): Yeah, yeah, yeah. Which I mean, we need some explanation like that because people in football and other contact sports are being hit all the time. And it raises the question, why don’t you see more of these events with people collapsing and having fibrillation. So you’re saying, you have to get very unlucky. You have to be hit during your vulnerable phase.

Fenton (19:51): And very hard, right?

Strogatz: And hit very hard.

Fenton (19:53): What happens actually during Commotio Cordis is that there’s statistics that 50% of the cases happen from people playing baseball. In baseball you have a ball that is hard and fast that can actually excite — hit you in the chest often. So 50% of the cases of Commotio Cordis that come to the hospital, they come from baseball. And very often happen to younger people because they are not developed enough that when you get hit in the chest, the pressure can go into the heart. So it tends to be more people who play sports where you can get hit with a small ball. Like that it happened also, for example, with hockey. Chris Pronger in the 1990s, in 1998, in the playoffs he got hit with a puck. And he went also down. And his case is very interesting because he didn’t go right away, like Damar Hamlin, he went right away, he went right away down to the ground. In the case of Chris, he took a couple of few more seconds. My guess is that the hit, when it started, it produced just a single spiral wave that took time to break. And he got VT before going to VF. And in the case of Damar, it probably went to VF very quickly. And that’s why he lost consciousness right away.

Strogatz (21:00): So, VT — ventricular tachycardia. VF — ventricular fibrillation. That’s even deadlier. Let’s go back to this question of defibrillating for a second, because you mentioned the astounding — did I hear you right? You said something like hundreds of joules required? Or used nowadays in defibrillators?

Fenton (21:18): Right, so if it’s external it goes between 120 to 360. If it’s internal, it can be as low as 20. Well, “low.” Twenty joules, but still it’s painful.

Strogatz (21:27): If a patient talks about what it feels like to be defibrillated, how do they describe it?

Fenton (21:32): Very often when you have fibrillation you pass out. So you will not feel it very often. But in the case of atrial fibrillation — sometimes you have AF, atrial fibrillation — you have to go to the doctors, and then they’re gonna do a defibrillation, they have to do a shock. In the case of a student of mine, he tells me that he got AF and that he went to the hospital and they cardiovert him, and in order to cardiovert him, they gave him some sedative, so he was sedated. And then he says he remembers hearing someone screaming. And then later they told him no, it was him screaming from the shock, but he didn’t remember that it was him. So the shock is big enough that it is painful. That’s why they sedate you. And that’s why actually it is very important when they do a defibrillation, they have to hook it up to the ECG. Because as I mentioned, when they go in to defibrillating the AF, you hook up to the ECG, so you know, when you do the shock, you don’t do it during the T wave or at the end of the T wave. Because then you can initiate fibrillation in the ventricles. They always, every time they defibrillate you, they hook up to the ECG, and then they do the shock at the safe time.

Strogatz (22:33): So you’ve mentioned now — and I don’t think we’ve emphasized this distinction so far, so we should, probably should — atrial versus ventricular fibrillation. I remember some years ago, we had a president — I think it was President George Bush, the elder, the father of George W. Bush — who, if I remember, had atrial fibrillation as a kind of chronic … like he lived with it. If I’m remembering right.

Fenton (22:33): Yeah, I think so.

Strogatz (22:40): Ventricular fibrillation, if untreated, will be deadly.

Fenton: Right.

Strogatz (23:01): Because you’re not pumping any blood. But atrial fibrillation is something you can live with?

Fenton (23:06): Yes. So this is the nice thing with this separation, as you were saying. If fibrillation happens in the ventricles, you have to defibrillate within seconds, minutes, right? That the longer you take to defibrillate, the harder it is to defibrillate — because the tissue becomes less excitable, because there’s less oxygen, and then you have less oxygen to the brain. And the chances to recover are very low. So you have to really defibrillate very quickly in the ventricles. In the atria — the atria and the ventricle are physically connected but electrically are disconnected. So when you have fibrillation in the atria, still the ventricles can contract — not completely or regularly, but can contract and send blood to the body. So atrial fibrillation, you can live but you always have… You feel tired. You cannot really move because the ventricles are not contracting as good as they could. And there’s also, because the atria is not pumping blood continuously, some of the blood can remain there, and it’s easier to produce clots of blood. The clots can go into the body and then it can give you a stroke.

(24:05) So when you have AF, it increases your chances to have a stroke. And AF, it happens to most people when they get older. Well, not most people, but as you get older there’s much higher chances to have AF. About 2.2 million people in the U.S. have AF. Like 70% of people with AF are between 65 and 85 years old. And one of the interesting things about AF is that it starts slowly. The wave starts breaking and producing spiral waves, but then they disappear. They go away. So self terminates. But as they keep appearing, they bring out more, the longer it appears, the longer they remain. So the more often you get AF, the longer the episodes and the harder it is to terminate.

(24:46) So if you started developing AF you want to try to go to the doctor’s and get either medications or other methods that are called, like, ablation. They can go inside it with a catheter and then burn sections of the atria. So then these waves don’t have space enough to rotate and then they self-terminate. So there are methods to try to terminate the arrhythmias. And they work better the sooner you find out you have AF because as longer — it’s very interesting — as it happens, it also remodels the tissue. So the tissue becomes a little bit larger and also the electrophysiology remodels it as well. So every time that you have more fibrillation, it’s easier to continue the fibrillation for longer until it becomes sustained. And once you sustain, the only way to do it is basically these kinds of ablations that they have to go there.

Strogatz (25:31): So let us go into the final section of our discussion here, which is to really focus on work that you and your students and postdocs and colleagues have been doing about fighting waves with waves. So why don’t we begin with, what is it that you and your research team have come up with as an alternative to the classic defibrillation that we’ve been talking about so far?

Fenton (25:52): So one of the nice things about electrophysiology of the heart is that it really matches well with what we call in physics or in math, applied math, an “excitable system.” An excitable system has a lot of mathematics behind that can be used for nonlinear systems or chaotic systems to investigate the dynamics of these activations that can happen in space, in time and space. So the nice thing is that actually, when you have fibrillation, which are multiple spiral waves, the dynamics is not random. You can write equations of motion to describe how it happens. And we have shown — and other people as well have shown — that it can be chaotic. So the dynamics of fibrillation is chaotic. And because it’s chaotic, it’s not random, there are ways you can control. You can actually investigate how the arrhythmias will behave, so that you can actually perturb and in a particular way with small perturbations and control.

(26:42) The nice thing about chaotic systems, as we know, is that there are periodic orbits that you can form in time. And you can find ways to perturb in a particular time with a particular strength that can be very small and control the system. So one of the things that we can do is know when to perturb with small shocks instead of one big shock. So we develop a couple of methods, and other people have also been working on this area, that we tried to figure out how to use nonlinear dynamics and chaotic approaches that are using chaotic systems to minimize the perturbations that actually can work on terminate or control a system.

(27:17) So imagine — I don’t know if this is a good analogy — but when you have a box with a lot of coins, and you want to put all the coins into one edge, you can maybe make a big shuffle, and then all the coins will go to one side, right? But instead, you can do small little shuffles, and little by little, you can move the coins into an edge. So that’s the main idea, that if you can perturb in a particular time in particular places with small energy, then you can actually synchronize the system and terminate the arrhythmias.

(27:42) It was very nice because we started it from a theory point of view, then doing some numerical simulations. And then we went to experiments in vitro and then in vivo where actually we were able to defibrillate hearts with using just 10% of the energy. So instead of using the large energy shocks, you can use 10%, a couple of shocks like that and defibrillate. So imagine, what do you prefer, if you were to be hit by Mike Tyson: Do you prefer one shock or a couple of slaps? So it’s probably better to have a few slaps even though they are probably they’re very painful, but less painful than the one big hit, right? So that’s the main idea that you can do small little shocks and then control the system. We’ve been working on different methods to apply that idea. And we’ve been successful so far.

Strogatz (28:28): Is the idea something like, I know sometimes when you have a spiral wave, they come with a handedness that there might be some that are, so to speak, right-handed and others are left-handed. They come in pairs, often. And if you hit a left-handed one with a right-handed one, they’ll both disappear. Is that the kind of thing you’re doing? Are you trying to inject a spiral wave to bang into an existing spiral wave? Or are you trying to push a spiral wave off the heart, or what?

Fenton (28:53): Well, it turns out that actually, that’s what happens at the end, or that’s the requirement at the end, that you need to defibrillate. Every time you have fibrillation, you have many spiral waves. And you have spiral waves that rotate clockwise and spiral waves that rotate anticlockwise. And when they appear, you have to terminate all of them. And the way you terminate all of them is by matching each one with their counterpart. So when you do the big shock, that’s actually effectively what you’re doing. You excite all the tissue, such that you’re connecting all the spiral waves from one direction to the spiral waves from the other direction. So if you excite all the tissue, you do that instantaneously.

(29:27) So we came out actually with a theory recently that using the dynamics of phase space where you can actually map the dynamics of the system not in physical space, but in a space of the dynamics of the variables of the system. That actually can tell you where to perturb. And it turns out when you go back to the physical space, the easiest way to terminate an arrhythmia is precisely to do a stimulus along the edge behind the wave that connects one spiral wave with their counterpart spiral wave. And that mechanism actually we call it “teleportation” because a spiral wave that is in one point, you can actually effectively with a stimulus move it somewhere else. So instantaneously you can move it from one place in the domain to another space by a stimulus that is well designed across the back of the wave, of the spiral wave. So in order to defibrillate, you want to teleport all the spiral waves that are clockwise with counterparts that are anticlockwise. And if you do that with the lowest energy, you succeed to defibrillate with the lowest energy.

Strogatz (30:26): This sounds very interesting. But I can imagine doctors objecting to it for a few reasons. For one thing, you know, when it’s a matter of life and death, like a person only has a few seconds, you can see why they want to use the paddles. It does work. If you’re describing something that requires precise measurements and timing, won’t they say to you, “We don’t have time for that. We can’t make those measurements. This person is lying on the ground.”

Fenton (30:52): Right. No, you’re totally right. When you’re talking about some things that are really life and death. So that is the case here. With the low energy defibrillation methods that we have developed, they work obviously computationally, and they work in the lab. But to make sure that they work all the time, everywhere, it needs to be designed in a particular way. So right now, when you have the manufacturers that do defibrillators, they say, “Right now it works,” right? The main idea, it works right now, why do we want to lower? What we need to do is to make sure that we can develop these theories applicable that always succeed even at the low energy.

(31:28) So you can try to first do that as a first approximation. And if not, if it fails, then you do the big shock. You need to make sure that when you apply these, you don’t wait too long so then it’s too hard to defibrillate at the end.

(31:39) Right now, a lot of the implantable defibrillators, what they do is before they do a big shock, they try to do what is called ATP, anti-tachycardia pacing. When an arrhythmia forms, in general it starts with a spiral wave that develops into multiples. So the main idea is that once the algorithm detects that you have fibrillation in the ventricles, then they’re gonna try to pace a little faster than the rotation of the spiral waves to see they can affect the wave and terminate it. So they do a little bit of this ATP. And then if it doesn’t work, then they go for the shock.

Strogatz (32:11): I’m just thinking that it’s interesting the psychology of the people making the implantable defibrillators or the doctors who use them in their patients, that they’re willing to do this sort of gentler — I assume this ATP, the anti-tachycardia pacing, that that’s a sort of a gentler or more benign attempt to rescue the heart before you give it the blast, the implantable blast. So they already seem open to this kind of idea to try something milder before you bring out the big gun.

Fenton (32:41): Right. So this is something in between. The ATP is done with an electrode that is attached to the… So the defibrillators in general, they do an electric shock between an electrode that is in the ventricles and the defibrillator itself. So that’s how they do an electric field between those two. And when they do the ATP, they just base from the electrode that is on the ventricle. One of the ideas we have is to try to do the low energy defibrillation by using just instead of those two — the electrode and the base of the defibrillator — to do the low energy shocks. But it still requires some time, and we need to work on making sure that we always show that it always is safe and successful to defibrillate.

(33:19) But from my point of view of a physicist, I think it’s amazing that we have been able to understand a lot of the dynamics of arrhythmias from just using concepts of excitable systems that exist, that have been in existence for many years, to describe the dynamics of chemical oscillators and things like that. And the theory can be applied. And actually, numerically, we can always see that it works. And we go to the experiments. And we can actually see that it works.

(33:44) So the exciting thing is that just using those concepts can be applied to develop new techniques of new ways to defibrillate, that is not just the method that it was just discovered, that just one big shock will work, as long as it’s very strong. So there’s still a long way to go to make sure this is applicable. But the theory is there. And that I think that’s what is exciting. Then the engineering part of how to make it work successfully and reliable, yeah, it’s a long road to go. But we have the background of where to start.

Strogatz (34:14): Oh, it’s great. You know, as a person who does math myself, I’m thrilled that you have these theoretical ideas that may turn out to be lifesaving or improve the quality of life for the people who need them. So let me ask you just a few questions about the details of this. I want to ask you about the computer part and about the experimental part. So why don’t we start with the computer? These computations sound like they would be hard. I read somewhere that you need something like 40 or 50 differential equations, nonlinear differential equations for each cell because I assume you’re keeping track of ionic conductances and voltages and concentrations. So they’re hard equations just for one cell and then you have a lot of cells to deal with. How do you do these computations? Are you using supercomputers? Are you using graphics cards, or what?

Fenton (35:04): Yes, in order to quantify the electrical voltage in the cell, you have to account for all the ion channels that exist and all the currents that pass through, in addition to the calcium dynamics. So you can go as a physicist looking for the spherical cow — in this case, the spherical cell. And do simpler models that have only like two variables. Two variables is enough to give you dynamics of how will happen in general. But as you want to go more precise to the cell dynamics and all the complexity that exists there, you can start using models that people have developed over the years. Biomedical engineers have developed complex models. Some of the models include up to 100 differential equations just for one cell. You can imagine the number of variables you have is in the thousands. The larger the number of variables you have, [the larger] the number of experimental data you need to make sure you’re not in a real minimum, not in a local minimum. But regardless, these models need to be used to investigate a little bit of how it works.

(36:16) So when you study these models in space, now you have to account for all the cardiac cells. And sometimes numerically, you have to go even discretize shorter than the discretization of the cardiac cells because of how we model the diffusion of the electrical activity through the cell. So you ended up having millions of cardiac cells you had to simulate when you go to 2D or 3D realistic hearts. So most of the time people use supercomputers. So you need to use supercomputers to simulate the electrical activity. And sometimes it takes many, many hours to do just a couple of seconds. It is a big, a big problem. There are methods to try to accelerate the dynamics. You can use adaptive-in-time-and-space methods and more complicated ways which we and many other people have done. So you can run simulations faster.

(36:46) Over the last 10-15 years, the develop of the graphic cards for gaming has actually allowed to do the supercomputer simulations very often. Sometimes you can do them in a PC or even in a laptop, you can have thousands of processors in the GPU, and develop the simulations using these programs that can have access not to the CPU, but the GPU, which has multiple processors to accelerate the dynamics that they are plotting on the screen.

(37:12) So what happens is that instead of using the pixels that you used to plot it on the screen for coloring, you can use them to account for the variables for the model that you want to do. So you can then use the pixels’ information for each variable of the system. And then multiple variables are the cells in the GPU and run it really, really fast in parallel.

Strogatz (37:33): I want to underline that point. I think it’s remarkable for people who haven’t thought about it that you think of video games as fun, but kind of frivolous. You know, this is like kids wasting their time or just goofing around, or grownups too. But they have come up with a technology to make the games play very fast that can be useful to people like you who are trying to calculate things very fast about heart cells spread out in space. It’s just not something you would have imagined, maybe, as the way forward but it turned out to be very valuable and a very creative use of this technology developed for video games.

Fenton (38:08): Yeah, a lot of people have been using them for high- performance computing since the last 15 years. And then Nvidia developed their own language to do that, this CUDA language. It can be a C compiler in CUDA or even a Fortran compiler in CUDA. Yeah, my old advisor Robert Gilmour used to say “Before you used to spend a lot of money on the computers. Now you have to spend a lot of money on the people who code the software for the computers.” Because now it’s, you don’t need to spend money on the supercomputer — you can spend the money on a cheap GPU computer. But now it’s more complicated to write the codes to do that.

(38:41) But there are many different languages out there. And we started working with one called WebGL that allows you to run codes directly through the web browser. So you can actually run the simulations in a browser. So they’re independent of the operating system and independent of the device. So as long as the GPU can handle the memory of the program that you want to run, you can even run it on the cell phone. The cell phone is so powerful. Your cell phones now, they are much more powerful than all the computation that was on the lunar modules that we sent. It is just amazing how much power they have. So you can actually do a simulation of a 3D heart — a small 3D heart, like a rabbit heart — on a high-end cell phone. We can do simulations in real time.

(39:23) Also, the nice thing is that they can, because they’re using the pixels that you show on the screen to do the simulations, you can interactively stimulate the tissue directly or change parameters in the simulation and see what happens as you investigate the dynamics of the system. And it can be done for any reaction-diffusion system or any partial differential equations. So you can do it for fluid dynamics, you can do it for crystal grow. So, this is the nice thing about GPUs, graphic cards. It is so powerful now. We can do those kinds of simulations now.

Strogatz (39:52): So let me close with you by just bringing up the question of the experiments. You mentioned them. But in my introduction, I mentioned that you’ve been using, in addition to animal hearts… Of course, the really dramatic thing is to use actual human hearts. And so I understand that you’ve gotten access to human hearts from donors, organ donors. Could you tell a little bit about that and what they’ve taught you?

Fenton (40:15): Yes, so for many years, we have always used animal hearts — rabbits, guinea pigs, pigs sometimes. And when I was at Cornell, we actually used even horse hearts, and a horse heart is huge. It is bigger than two basketballs together. They’re made for running. So when you open a horse, the inside you can see is mostly lungs and the heart.

(40:35) The main idea is to try to minimize the use of animals. And also, most importantly, the main case, what we wanted to study is the human heart. So when I came here to Georgia Tech, I tried to collaborate, I did collaborate with cardiologists at Emory, at the university of Emory in the hospital. And finally, after a few years, collaborating with a couple of cardiologists, we were able to write some of the protocols with the lawyers and permissions with the patients. Very often when a patient gets a heart transplant, we can get the hearts from the patient. They call us and we’re waiting outside the operating room. As soon as the new heart arrives and they take out the heart of the patient, they give it to us. So I can prepare it and bring it to Georgia Tech, which is 10 minutes from Emory, from the hospital. So I bring it into our lab, and I can perfuse it with something similar to blood. You can use what is called Tyrode’s solution, it’s a solution that has all the ions necessary to keep the heart alive. And we can revive the heart. It’s kind of like a — really like a Frankenstein. Really is. It’s just, it’s alive! It’s just, bring the heart, just start perfusing it, that comes back alive. And it starts to contract. And then we can do experiments there.

(40:40) To visualize electrical signals, we use something called optical mapping. You put a dye that is a voltage dye that goes into the membrane of the heart. And these dyes absorb light at one frequency and emit at a different frequency. That the peak emission is a function of the voltage. So as the voltage changes, the emission spectra changes. So the amount of light that you get at a given frequency changes. So, you can put some filters into the camera and then you visualize directly in whole space the electrical signal as a change in the intensity of the light. And then we can visualize the spiral waves that form. We can see the spiral waves actually in real life, we can see the spiral waves rotating, spiral waves breaking, how they initiate and how can actually when we do the stimulus, how are we perturbing them such that they can either continue or terminate. So it’s pretty amazing that we can actually now do those experiments in the real hearts, in the real human hearts.

Strogatz (42:38): It’s really amazing. I mean, because this is something, you know, I personally have been interested in this kind of question about excitable media and heart arrhythmias ever since I worked with a gentleman named Art Winfree a long time ago, back in the early 1980s. And in those days, there were just the beginnings of visualization of waves on hearts. But mostly it was theoretical. We imagined spiral waves. Our math and chemical analogues told us there should be spiral waves or their three-dimensional generalization, scroll waves. But the idea that you could actually see one on a human heart was pretty fantastical. And now you’re doing it. We should probably end by trying to think about the future. What do you picture down the road, the theoretical and experimental work that you and your group have been doing? What’s your dream about where this could lead?

Fenton (43:27): I think, what we’re all looking for, is that we can defibrillate hearts before they start and know when something is going to develop defibrillation and how to terminate them with very low energy pulses. It turns out that now there’s another way to try to defibrillate — is using light. So there’s some groups that have been working on adding ion channels into the cells into cardiac cells that can be excited with light. So you can actually stimulate or unstimulate with light, depending on the intensity and the wavelength. It seems like it’s possible in the future that you can create this application of making cells excitable with light. And then at some point, maybe you can even defibrillate with just putting a light internally into your system and defibrillate that way without having to use an electric shock. This is called optogenetics. And there are many groups in the U.S. and in Europe working on that.

Strogatz (44:19): Wow. That, that is really futuristic thinking. Amazing! Let me just say thank you very much, Flavio, for joining us today. This has been a fascinating conversation. So, we’ve been talking here with Flavio Fenton, who studies cardiac dynamics at the School of Physics at Georgia Tech. Thanks again so much for joining us, Flavio.

Fenton (44:37): Oh, Steve. This has been my pleasure. Thank you so much for having me.

Announcer (44:44): Space travel depends on clever math. Find unexplored solar systems in Quanta Magazine’s new daily math game, Hyperjumps. Hyperjumps challenges you to find simple number combinations to get your rocket from one exoplanet to the next. Spoiler alert: There’s always more than one way to win. Test your astral arithmetic at hyperjumps.quantamagazine.org.

Strogatz (45:16): The Joy of Why is a podcast from Quanta Magazine, an editorially independent publication supported by the Simons Foundation. Funding decisions by the Simons Foundation have no influence on the selection of topics, guests or other editorial decisions in this podcast or in Quanta Magazine. The Joy of Why is produced by Susan Valot and Polly Stryker. Our editors are John Rennie and Thomas Lin, with support by Matt Carlstrom, and Zach Savitsky [as well as Nona McKenna and Saugat Bolakhe]. Our theme music was composed by Richie Johnson. Julian Lin came up with the podcast name. The episode art is by Peter Greenwood and our logo is by Jaki King. Special thanks to Bert Odom-Reed at the Cornell Broadcast Studios. I’m your host, Steve Strogatz. If you have any questions or comments for us, please email us at [email protected] Thanks for listening.