10/22/2026
Wearable sensors coupled with artificial intelligence can generate blood pressure readings nearly as well as the invasive and risky arterial lines relied on in intensive care units and operating rooms, new work by Johns Hopkins University researchers demonstrates.
Successful initial patient tests suggest the system could become an alternative to arterial catheters and offer hospitals the ability to monitor arterial blood pressure beyond intensive care units. The sensors could also give people with hypertension a way to continuously monitor their health, like wearable glucose monitors for people with diabetes.
"Patients in the ICU need continuous blood pressure monitoring to catch problems early, but it means an arterial line, which comes with a risk of bleeding, clotting, and infection," says lead author Carl Harris, a PhD student in Johns Hopkins Biomedical Engineering. "We wanted to find a better way."
The findings are published in Computers in Biology and Medicine.
Patients in intensive care often experience high degrees of fluctuation in their blood pressure. If it's too high, it can lead to stroke, heart attacks, and kidney damage. Too low might mean not enough blood getting to the brain and vital organs.
The current standard for continuously monitoring blood pressure is an arterial line—a catheter inserted directly into an artery, typically in the arm or the groin, that provides continuous, real-time pressure tracking. The procedure works very well, but comes with a high risk of bleeding, clotting, and infection. It also limits mobility.
Familiar blood pressure cuffs that wrap around the arm aren't invasive, but they only provide intermittent readings. The team created a system called MOSAIC based on two sensors—one placed on a patient's chest, and the other on a finger. Together the sensors record the heart's electrical activity and the flow of blood through the body and relay these signals to a deep learning model that generates a waveform, a continuous readout of blood pressure over time.
"We reconstruct waveform data in a way that's meaningful, accurate, reliable and, most importantly, non-invasive," says senior author Robert David Stevens, an affiliate member of the Malone Center and chief of the Division of Informatics, Integration, and Innovation at Johns Hopkins Medicine. "It's a possible solution for avoiding the current standard of care for measuring blood pressure, arterial lines—a very invasive procedure with a risk of many complications."
In an initial study of 28 patients in the intensive care unit at the Johns Hopkins Hospital, the system generated waveforms that closely matched those recorded by traditional arterial catheters.
"We're very close to hitting that gold standard," Harris says. "This demonstrates we can do what we set out to do—and pretty well."
The team is now validating the sensors and algorithm in a larger cohort of Johns Hopkins ICU patients. Longer term, the researchers hope the system could reduce the need for invasive monitoring and become a way to continuously measure blood pressure in a range of settings—in regular hospital wards and at home.
The team envisions people who have hypertension, one of the most common and the deadliest diseases worldwide, could wear the sensors daily to keep an eye on their blood pressure. They hope the sensors could bring about a paradigm change for the millions of people living with hypertension, the way glucose monitors have for people with diabetes.
The sensors could also be used on healthy people to gain insights into daily blood pressure fluctuations and trends during daily activities.
"We observe sick patients in the intensive care unit, but we have no idea what's going on with blood pressure in a healthy person who's just living their life, going to work and being with their family," Stevens says. "What happens to their blood pressure day after day? Nobody really knows."
Additional authors include: Johns Hopkins graduate student Bright Nnadi; former graduate student Sampath Rapuri; former PhD student John Rattray; Johns Hopkins Applied Physics Lab senior engineer Francesco Tenore; and engineer Ralph Etienne-Cummings.
The work was supported by the Johns Hopkins Institute for Clinical and Translational Research, through the National Institutes of Health's National Center for Advancing Translational Sciences award 1UM1TR004926, NIH Roadmap for Medical Research, and National Science Foundation Graduate Research Fellowship DGE2139757.
Image Credit: Getty Images
