Share this post:

Share on email
Share on facebook
Share on twitter
Share on linkedin

Biological age and other AI tools in preventive cardiology


Doctors have observed for many years that people age at different speeds, according to Dr. Francisco Lopez-Jimenez, chair of preventive cardiology at Mayo Clinic in Rochester, Minnesota. The vascular system can also show signs of age and can offer clues about risk of developing a heart attack, vascular disease or stroke.

In this Mayo Clinic Cardiovascular CME podcast, Dr. Lopez-Jimenez explains what vascular age means, how it can be measured using an electrocardiogram and how people have been able to lower their vascular age through lifestyle changes. He also discusses how he and other doctors at Mayo Clinic are using ECG to identify heart conditions and the future impact artificial intelligence may have in medicine.

Dr. Kopecky: Hello, I’m Stephen L. Kopecky, M.D., a preventive cardiologist at Mayo Clinic in Rochester, Minnesota. My pleasure to be speaking today with the chair of Preventive Cardiology in Rochester, Francisco Lopez-Jimenez, M.D., M.B.A. Welcome, Francisco.

Dr. Lopez-Jimenez: Thank you, Steve. Pleasure being here.

Dr. Kopecky: Now, Francisco, you’ve been helping to head up our artificial intelligence tools in preventive cardiology. And you’ve come up with this concept of vascular age. Can you explain that to us? And how is that different from, like, chronologic age?

Dr. Lopez-Jimenez: Sure. So we know that people age at different speeds. And this has been observed for many, many, many years, that some people just don’t age as fast as others, or the opposite, that some people actually age fairly quickly. So you see a person that is 60 and actually looks like 80, or the other way around.

So we know that the aging happens at different speeds. So what we call as vascular age is what we believe determines the aging of our vascular system, including the heart, arteries, veins. And might predispose people to develop heart attacks or peripheral vascular disease or even developing strokes. So this is something that has been discussed at least for three decades, but we believe that we are getting into the secrets of how, how to uncover or how to describe and measure vascular age better.

Dr. Kopecky: And what tools are you using here, at Mayo, to kind of look at vascular age?

Dr. Lopez-Jimenez: Yeah, what we’re doing right now is we are, we are, we can calculate vascular age using the electrocardiogram. But to understand that, let me give you a little bit of background in how vascular age has been trying to, to be calculated in the past.

So about 20 years ago, some investigators noticed that people who smoke actually have arteries that look older. And also people who have diabetes or people with high cholesterol. So at the beginning, the first efforts to determine vascular age were based essentially on counting the number of risk factors, and according to the number of risk factors and how bad those risk factors were, it would be a way to sort of estimate the age of the arteries.

However, you know, the problem is that that doesn’t necessarily add much to what we know. We know that people who have risk factors have arteries that are more likely to be in trouble. So what we did here at Mayo, a few, a year ago, actually, was to, we noticed that we can actually predict the age using the electrocardiogram, but we notice that the prediction was not perfect. So was close, but in some people was not that close. In some people will be like 10 or 15 years different than the actual age. So we ask ourselves, so is this just error, trying to measure, calculating age using the electrocardiogram or is that error telling us something?

So what we did then was we grouped people who, who were — the electrocardiogram will tell us that they were much older than their chronological age. And also we grouped people where the electrocardiogram will tell us that there were much younger than the chronological age. And what we saw was that those who are older by ECG actually die earlier and those who seemed to be younger by the electrocardiogram actually live longer compared to their peers of the same age. So to add that was an unequivocal signal that the electrocardiogram can actually tell us the vascular age. And then subsequent studies actually showed the same for heart attacks and other cardiovascular events.

Dr. Kopecky: So a patient comes to you, you do the electrocardiogram and they have an older vascular age than a chronologic age. What do you tell them to do? What can a patient do to reverse that or change that?

Dr. Lopez-Jimenez: Well, I think the good news is that it seems like what we recommend for people to prevent heart attacks are exactly the same things that might get people to be younger. And we have seen more than a few anecdotal situations where we ask the patient to exercise regularly and to lose weight and to follow a healthier diet. And six months later, the electrocardiogram shows that they are five years younger. So it seems that what we had been recommending all along actually makes our arteries become younger.

Dr. Kopecky: So in just six months you can see a difference in the ECGA?

Dr. Lopez-Jimenez: Absolutely, yeah, absolutely. So some people, you, in some patients, you might actually wonder if the graph is wrong because it seems to be going in the wrong direction, you know, the patient is actually becoming younger and younger as time goes. But what happens is that they are implementing significant changes that are just helping the arteries to heal and to essentially become younger.

Dr. Kopecky: Fascinating. So what other tools may be helpful with artificial intelligence to prevent, you know, look at age? Vascular age.

Dr. Lopez-Jimenez: As you know, artificial intelligence has been evolving very fast over the last several years in medicine in particular. And we at Mayo had been doing a lot of experiments and studies trying to find better ways to diagnose conditions. And actually our focus has been on using the electrocardiogram, a tool that has been used for more than 100 years. But trying to see if the electrocardiogram can identify other conditions that it was not intended to identify.

For example, we were able to demonstrate that the electrocardiogram can find people with a low ejection fraction or when the heart is just not pumping good enough. And even when the electrocardiogram seems completely normal in some situations, the artificial intelligence is telling us that that particular patient surely has a failing heart. So that was one and the other one that was, to me was fascinating, was to tell us when a patient has atrial fibrillation or one of those abnormal rhythms that can lead to strokes, even though the electrocardiogram seems normal. So to identify those individuals might actually be a game changer in the future, because if we can prove that those individuals are more likely to have strokes, even though the heart rhythm seems to be normal now, but the computer is telling us that this person, at some point during the day or during the month, is actually having atrial fibrillation leading to stroke, we might be able to treat patients way in advance and to know who might be at risk for strokes.

Dr. Kopecky: So you can find atherosclerosis with an ECG and find hypertrophy with ECG, so hypertension, you can look at things like heart failure. What about some of the infiltrative diseases? Amyloid, sarcoid? Can you?

Dr. Lopez-Jimenez: Absolutely. Yeah, we’re actually just finishing a project right now that, it seems like we can identify those with cardiac amyloidosis. And again, what is fascinating is to see cases where the electrocardiogram will seem completely normal otherwise. But the computer is telling us that this patient has a very high probability to have cardiac amyloidosis. And we actually have several cases that electrocardiogram was already showing that signal years before, years before the patient was diagnosed with cardiac amyloidosis. So that’s another situation that might help us to identify patients where the cardiac condition was not even suspected. But the electrocardiogram might be telling us that years in advance, years before the official diagnosis is made.

Dr. Kopecky: That’s really fascinating. What a great tool and, you know, relatively inexpensive and certainly low-risk tests like electrocardiogram. Congratulations on all that. Any other parting words for us about artificial intelligence, where do we go from here?

Dr. Lopez-Jimenez: Well, I anticipate that there will be a lot of developments in the next few years. And certainly artificial intelligence will not replace doctors. I think it’s going to just make our lives easier and it is going to be, is going to make patients to be better off to live longer and will make medicine more advanced and more accurate.

Dr. Kopecky: Great. Well, it’s a wonderful message. Thank you, Francisco, for joining us today.

Dr. Lopez-Jimenez: Thank you, Steve. Thank you for the invitation.


Francisco Lopez-Jimenez, M.D.

Dr. Lopez-Jimenez is chair of preventive cardiology at Mayo Clinic in Rochester, Minnesota.

Recent Posts

Share this post:

Share on facebook
Share on twitter
Share on linkedin
Share on email
Related Content