A new skin-like computing patch developed at the University of Chicago Pritzker School of Molecular Engineering (UChicago PME) can analyze health data using artificial intelligence in an unprecedented way. Unlike today's wearable devices, it carries out its AI computations directly on the body, in mere milliseconds, without relying on a wireless connection.
While your current smartwatch may be able to track your heart rate or movements, it doesn't analyze what it finds. The analysis happens elsewhere, after it shuttles data to an external server. In some situations—detecting ventricular fibrillation in the heart, for instance—that few-seconds lag to communicate with the server is too long.
The new device, designed and tested in collaboration with researchers at Argonne National Laboratory, was made possible by the development of manufacturing processes that allow organic electrochemical transistors to be printed onto flexible surfaces.
"The future that we're trying to realize is to make wearable and implantable devices smarter," said Sihong Wang, an associate professor of molecular engineering at UChicago PME and co-senior author of the new study, published in Nature Electronics . "It's helping people have a personal, instantaneous doctor integrated into their devices."
Manufacturing stretchy transistors
For years, Wang's lab has been working to create electronic components that can stretch and bend like human skin, with the ultimate goal of creating smart devices that adhere to human tissues. The group previously developed methods for fabricating stretchable transistor arrays and a stretchable OLED display .
In the new work, Wang and his colleagues set out to build a stretchable neuromorphic computing circuit—a large array of transistors that can run analyses of health data. Earlier work had demonstrated that the concept was theoretically possible with a small number of transistors but hadn't scaled it up to a practical size.
The transistors the team wanted to use, called organic electrochemical transistors, work differently from those inside a standard computer chip. They process information using both electrical current and the movement of ions through a gel-like electrolyte layer. The electrolytes give each transistor a built-in memory, letting them store numerical values stably over time, much the way a synapse in the brain can be strengthened or weakened to encode a learned pattern.
These components, however, presented a manufacturing challenge. The flexible surface layer is sensitive to heat and solvents and so can't be fabricated using standard chip production techniques. At the same time, the gel electrolyte layer has a tendency to move like a liquid, merging with neighboring devices and causing short circuits.
"What we had to ask was whether we could use or change the properties of these polymers to make them compatible with photolithography—the main patterning method used in the microelectronics industry," Wang said.
The team solved the challenge by engineering a new type of polymer gel that could be hardened into precise patterns through exposure to ultraviolet light. The result is a fabrication method that can produce 10,000 organic electrochemical transistors per square centimeter.
"As computer scientists, we're used to thinking of a neural network weight as just a number," said Zixuan Zhao, a graduate student at UChicago CS and co-first author of the study. "In hardware, it's a material—with variability, history, and physical limits. The challenge was to hold those constraints in mind and still compute with enough precision to matter."
Saving lives with speedy computing
To test the utility of the new devices, Wang's team used one of their new stretchable arrays to run a pre-trained algorithm designed to help treat ventricular fibrillation. This dangerous electrical storm in the heart can be fatal and is most often treated with a one-size-fits-all defibrillator shock that delivers a massive jolt of electricity to the entire heart. Researchers have proposed a more precise treatment: mapping abnormal waves of electricity as they move through the heart and delivering small, precise pulses just ahead of them before they can continue.
However, the obstacle has been time. Wavefronts move through the heart so fast that the entire analysis must be completed within milliseconds — far too quickly for data to be transmitted to an external computer and back.
"This is a situation where it's not feasible to have remote computing. It just takes too long," said Wang. "But if you have a computing device that can do the analysis within the body, it could be possible."
Using real cardiac mapping data from a donor human heart, the team showed that the stretchable array could locate wavefront positions with 99.6% accuracy, even while the device was stretched to more than one and a half times its normal length.
In a separate demonstration, a neural network encoded in the array analyzed a combination of vital signs and personal health data—including cholesterol levels, blood sugar, maximum heart rate, and ECG readings—to assess a patient's risk of heart attack, achieving 83.5% accuracy.
Wang sees this computing array as one component of a fully integrated, body-compatible health platform. His lab is now working to pair the computing array with stretchable wireless communication components and improved sensors, moving toward a system that can sense, analyze, and respond to health data as a fully integrated whole.
"Instead of sending data away to a remote server, we can begin making sense of it right where life is happening," said Fangfang Xia, a computer scientist at Argonne National Laboratory and co-senior author of the study.
Citation: "A large-scale stretchable neuromorphic circuit for on-body edge computing," Li et al, Nature Electronics, May 20, 2026. DOI: 10.1038/s41928-026-01639-8
Funding: This work and the researchers involved were supported by the US Office of Naval Research (N00014-21-1- 2266, N00014-21-1-2581), the University of Chicago Joint Task Force Initiative, the National Institutes of Health (1DP2EB034563, R01-HL141470, R01 HL165002), Argonne National Laboratory, the U.S. Department of Energy (DE-AC02-06CH11357, DE‐SC0014664), the Leducq Foundation, and the CZ Biohub.
