Tech companies are targeting heart disease — here’s how
The technology is advancing much faster than it’s used in practice
以下内容由机器翻译生成。如果您觉得可读性不好, 请阅读原文或 点击这里.
从 apps that diagnose irregular heart rhythms 自 phone cases that claim to measure blood pressure, there has been a wave of technology promising to use our everyday devices — smartphones and wearables — to fight heart disease. But why heart health, and how much can these gadgets really do?
Heart disease is the most common cause of death around the world, according to the World Health Organization, so of course companies want to work on a problem that could have a huge impact, and bring in lots of money. Apple, for example, has launched a study to identify irregular heart rhythms with its Apple Watch. Plus, many sensors and wearables lend themselves particularly well to helping with cardiovascular problems, says Greg Marcus, a cardiologist at the University of California, San Francisco.
Many phones, for example, already have accelerometers that measure physical activity like steps taken, while wearables like the Apple Watch and Fitbits use sensors to measure heart rate. These types of information are crucial when it comes to heart health. (These fitness trackers, however, are not very accurate.)
Heart irregularities can be dangerous without causing obvious symptoms, so smartphones that can diagnose them could be helpful as a prevention strategy. Tech companies are angling hard on this, but the big gap now is on the clinical side, and there needs to be more research to see whether these solutions are truly valuable.
Here’s an overview of the latest developments in this area:
IRREGULAR HEART RATE
The most common heartbeat irregularity is called atrial fibrillation, or afib. Afib happens when the two upper chambers of the heart don’t beat in sync with the two lower chambers, and can increase the risk of everything from heart attack to kidney disease to dementia. But it can be hard to detect. Various sensors, like the Zio patch 或 LifeWatch, exist, but it’s annoying to wear them continuously. “If your watch could detect it based on the rhythms that it could be measuring, then it could better allow you to get treatment early and prevent that stroke,” says Eric Peterson, a professor of medicine at Duke University and a member of the American Heart Association.
Last May, the Apple Watch app Cardiogram presented results saying that it could diagnose afib with 97 percent accuracy. Cardiogram had collaborated with Marcus, the UCSF cardiologist, as part of the Heart eHealth Study, the largest study in the field of mobile health and heart disease. And in November, Apple announced that it would partner with researchers from Stanford University to run the Apple Heart Study, to investigate afib. (Stanford will collect data from Apple until January of 2019.)
This is all very promising, but smartwatches have limitations too. The Apple Watch, for example, is being touted as an important health tool, but it’s easy to cheat. And while smartphones are extremely common, fewer people have smartwatches — and those who do are likely to be wealthier than the general population. With devices that use heart rate, the rhythm is being inferred from the pulse, and it’s possible to have an abnormal heart rhythm with a normal pulse, says Marcus. Plus, there’s a lot of variability in how good of a reading you can get, depending on motion and consistent contact with skin, and we need more research to understand how these devices can be used so that the information is most medically accurate.
HIGH BLOOD PRESSURE
High blood pressure, or hypertension, can cause strokes, heart attacks, and kidney failure. The problem is that blood pressure is notoriously difficult to take, even with the standard cuff devices at your doctor’s office.
Blood pressure varies widely during the day, and can rise if you’re stressed or even just dangle your feet off an exam table. So having a wearable that can monitor blood pressure throughout the day, and even the night, would definitely help doctors better identify at-risk patients and find the right therapy. The biggest challenge, however, is accuracy, says Bruce Alpert, a pediatric cardiologist who’s performed many validation studies for manufacturers of automated blood pressure devices.
“There have been many new technology methods invented over the last five years, especially as smartphones have become smarter and smarter,” he says. “To date, none of them has proven accurate.”
One smartphone app called Instant Blood Pressure, for instance, asked users to place their smartphone against their chest and a finger over the camera to take a measurement. But it was found to miss high blood pressure in eight patients out of 10. A new phone case claims to measure blood pressure from the fingertip, but an early study didn’t really prove whether the case is accurate enough for at-home use. And a variety of blood pressure-measuring wristbands also exists, says Peterson at Duke. The problem is always the same: accuracy.
One challenge is that all these cuffless devices don’t measure blood pressure per se, which is taken by squeezing an artery near your elbow as it’s been done for decades. They measure the pulse of blood in your finger, or wrist, and then use algorithms to correlate those numbers to a traditional blood pressure measurement. That approach leaves lots of room for errors.
People also need to be trained to take their own blood pressure correctly. In the doctor’s office, you’re asked to put your arm on a table at the same level as your heart, your back supported, and your feet flat on the ground. That’s because blood pressure can rise with the simplest motions. A wearable that takes measurements as you’re running to catch the subway or cooking dinner won’t be accurate.
What’s more, for people with hypertension, measuring blood pressure in itself can be stressful. Peterson says there’s a running joke among doctors: when patients take their blood pressure and see it’s high, they get stressed. So then they keep taking their blood pressure, and the numbers keep going up, because they get even more stressed. “The point is: Stop measuring your blood pressure!” Peterson says.
A wearable that gives you a reading every hour could stress people out, leading again to inaccurate results.
Over 30 million Americans have diabetes, meaning their blood sugar levels are too high. Over time, diabetes can damage kidneys, nerves, and eyes. It can also increase risk for heart disease and stroke, mostly because diabetes often comes with other at-risk conditions like hypertension and obesity. In fact, adults with diabetes are two to four times more likely to die from heart disease than adults without the condition, according to the American Heart Association.
People with the disease have to regularly measure their blood glucose, to regulate what they eat or even inject the hormone insulin if needed. Right now, the way to do it is invasive: patients have to prick their fingers to draw blood or get a tiny tube that goes under their skin to measure glucose in the fluid between cells.
Tech companies have been trying to create wearables that can monitor blood sugar without the use of needles. Google worked on a contact lens to detect glucose in tears, but the project is moving slowly. And Apple was rumored to also be developing some sort of needleless wearable. “It’s an incredibly difficult problem,” Mark Rice, an anesthesiologist and diabetes expert at Vanderbilt University told The Verge last year. “Everybody thinks they have a way to do it, and everybody, so far, has failed.”
Glucose is a hard molecule to detect — it doesn’t have very distinctive features. So the current tests use chemical reactions to turn glucose in a drop of blood into more easily trackable molecules. There also isn’t that much sugar in blood to begin with, and if levels drop dangerously low, it’s not going to spill over into tears, spit, sweat, or urine. So that makes any body fluid other than blood an unlikely candidate for measuring glucose.
Scientists are looking at other innovative ways to detect glucose — by looking at how infrared light shines through a thin patch of skin, like an earlobe, for instance. By calculating how much light is absorbed or scattered by glucose molecules, a device could measure blood sugar levels — at least theoretically. Analytical chemist Mark Arnold at the University of Iowa has created such a device, and it seems to be working on rat skin.
There’s only one problem: “It’s about the size of a small refrigerator,” he told The Verge last year. “Not a device that could work on somebody’s wrist.”
OTHER HEART-RELATED CONDITIONS
Earlier this week, the makers of the AliveCor KardiaBand, a sensor compatible with the Apple Watch, presented results saying that it can use a heart reading to detect dangerous levels of potassium in blood. (The US Food and Drug Administration has not yet approved KardiaBand for this purpose, though the company says it’s working on it.)
This condition, called hyperkalemia, can be caused by diabetes, dehydration, and chronic kidney disease, among other things. It can lead to kidney and heart failure. In general, hyperkalemia doesn’t cause obvious symptoms — but it does interfere with heart activity, which can show up on an electrical reading of the heart (called an electrocardiogram, or ECG).
The KardiaBand is a sensor that snaps onto the Apple Watch wristband and can take an ECG and send that information to an app. The team trained their AI with a dataset of 2 million ECGs linked to 4 million potassium values, all from Mayo Clinic. It learned to diagnose hyperkalemia with 90 to 94 percent accuracy.
Because people sometimes have hyperkalemia with a normal ECG, the KardiaBand won’t catch hyperkalemia for everyone, says William J. Brady, a professor of internal medicine at the University of Virginia School of Medicine. But in general, he says, hyperkalemia will produce an obvious abnormality in the ECG, and he has initiated treatments in patients based on the ECG before getting a blood test back to confirm. “I put a fair amount of trust and faith in the ECG in this regard,” he adds, though novice physicians or those not used to reading ECGs might find this type of interpretation more difficult.
And last year, engineers at Caltech showed that a smartphone app can accurately measure how much blood the heart pumps with each beat, called “left ventricular ejection fraction” (LVEF), as arteries expand and contract. LVEF is an important measure of heart health and is usually assessed using an ultrasound. Ultrasounds take a couple of hours and can only be done by technicians.
The app asked patients to hold a smartphone camera (an iPhone 5) up to their neck for less than two minutes. (The neck houses the carotid artery, which feeds directly into the heart, so the information from there is the most accurate.) The smartphone camera measures the expansion and contraction of the artery walls, and the algorithm inside the app analyzes the information to calculate the blood flow from the heart. The app was found to be as accurate as an ultrasound, but not as accurate as a brain scan, which is the most precise way of measuring LVEF.
This technology could be very important for children who receive chemo. Chemo can affect the heart, so patients often need to get an ultrasound every two weeks to figure out if the dosage needs to go down. But this can be difficult for those who might live in rural places, or whose parents can’t take them. “The idea here is to make sure that this kind of diagnostic is available to the masses,” says study co-author Morteza Gharib.
Whether tech companies will be able to crack the code behind monitoring the most challenging heart diseases remains to be seen. The future is definitely promising. But as more and more gadgets flood the market, it’s important that these devices are properly tested to show that they can actually work, Peterson says.
“Right now, we’re in a stage where the technology is advancing much faster than it’s used in practice, in testing,” he says. So before buying a wearable and blindly relying on it for monitoring your health, always make sure that the company has proven that the gadget is accurate. “So buyer beware!” Peterson adds.
Additional reporting by Rachel Becker.