Researchers at the University of Chicago Pritzker School of Molecular Engineering have developed a flexible, skin-like computing patch that can analyze health data directly on the body using artificial intelligence.
The device represents a major step toward wearable and implantable technologies that can do more than simply collect information. Instead of sending data to a remote server for analysis, the new patch can process information almost instantly, making it possible to detect health problems in real time.
Most wearable devices, such as smartwatches, can track health signals like heart rate, movement, and activity levels. However, much of the deeper analysis usually happens elsewhere, often on external servers connected through wireless communication. That process can introduce delays.
In everyday use, a small delay may not seem important. But in serious medical situations, even milliseconds can matter. One example is ventricular fibrillation, a dangerous heart rhythm disorder that can become life-threatening very quickly.
The new patch was designed to overcome this delay by performing artificial intelligence calculations directly on the body. It can analyze health information in milliseconds, without needing to transmit the data wirelessly before processing it.
The device was developed and tested by researchers from the University of Chicago in collaboration with scientists at Argonne National Laboratory. The work was made possible by new manufacturing methods that allow organic electrochemical transistors to be printed onto flexible surfaces.
According to Sihong Wang, an associate professor of molecular engineering at the University of Chicago and co-senior author of the study, the goal is to make wearable and implantable devices smarter. He described the vision as creating a kind of personal, instantaneous doctor built directly into a personโs device.
Wangโs laboratory has spent years developing electronic systems that can stretch, bend, and move like human skin. The long-term aim is to create intelligent devices that can attach directly to biological tissues and function comfortably with the body.
Previous work from the group included stretchable transistor arrays and stretchable organic light-emitting diode displays. In the new study, the researchers moved further by creating a stretchable neuromorphic computing circuit.
A neuromorphic computing circuit is designed to process information in a way inspired by the human brain. Instead of relying only on conventional rigid computer chips, the system uses networks of transistors that can store and process information in a more brain-like manner.
The researchers built the device using organic electrochemical transistors. These transistors work differently from those found in standard computer chips. They process signals through both electrical currents and the movement of ions inside a gel-like electrolyte layer.
This gel layer can retain information over time, meaning each transistor can act a little like a brain synapse. In the brain, synapses strengthen or weaken as part of learning and memory. In a similar way, these transistors can hold information that helps the system recognize patterns in health data.
Creating the device was not simple. The flexible materials used in the patch are sensitive to heat and chemical solvents, so traditional chip-making techniques could not be applied easily. Another challenge was the gel electrolyte, which behaves partly like a liquid and can spread, potentially causing neighbouring devices to merge and short-circuit.
To solve this, the researchers developed a special polymer gel that hardens into precise shapes when exposed to ultraviolet light. This manufacturing breakthrough allowed them to fabricate up to 10,000 organic electrochemical transistors per square centimetre.
That level of density is important because practical artificial intelligence systems require large networks of components. Earlier studies had shown that stretchable neuromorphic devices were possible, but scaling them into a more useful system had remained difficult.
The team then tested whether the patch could handle real medical tasks.
In one experiment, the researchers used the stretchable transistor array to run a pretrained algorithm designed to support the treatment of ventricular fibrillation. This condition causes chaotic electrical activity in the heart and can quickly become fatal.
Current treatment often involves delivering a strong electrical shock to the whole heart. However, researchers are exploring more targeted approaches that could track abnormal electrical waves and apply smaller corrective pulses before the problem spreads.
For that approach to work, the system must analyze heart signals extremely quickly. The abnormal electrical wavefronts move so fast that sending data to a remote server for processing would take too long.
The new device showed that this kind of analysis could happen directly on or inside the body. Using cardiac mapping data from a donated human heart, the stretchable array identified wavefront locations with 99.6 percent accuracy, even when stretched to more than one and a half times its original length.
In another test, the researchers encoded a neural network into the array and used it to analyze vital signs and personal health information. The data included cholesterol levels, blood sugar, maximum heart rate, and electrocardiogram measurements. The system then estimated heart attack risk with an accuracy of 83.5 percent.
These results suggest that future wearable or implantable patches could monitor the body continuously, analyze signals immediately, and provide rapid medical insights without waiting for external computing systems.
The researchers are now working to combine the computing array with stretchable wireless communication systems and more advanced sensors. Their goal is to create a fully integrated health monitoring platform that can collect, analyze, and respond to health data in real time.
The broader vision is a new generation of medical devices that do not merely record information, but understand it instantly.
Instead of sending health data away to be processed elsewhere, this technology brings computation directly to the body. That could make future devices faster, more private, more responsive, and better suited for urgent medical situations.
In simple terms, the artificial intelligence skin patch brings the idea of an instant personal doctor closer to reality: a soft, flexible device that can sit on the body, read health signals, and make sense of them almost immediately.
Reference: Songsong Li, Zixuan Zhao, Max Weires, Shiyu Hu, Yang Li, Lingfeng Tang, Shilei Dai, Yahao Dai, Youdi Liu, Nan Li, Wei Liu, Naisong Shan, Junyi Yin, Xiaoao Shi, Sean Sutyak, Cheng Zhang, Jie Xu, Junhong Chen, Yuepeng Zhang, Igor R. Efimov, Fangfang Xia, and Sihong Wang, โA large-scale stretchable neuromorphic circuit for on-body edge computing,โ Nature Electronics, 20 May 2026. DOI: https://doi.org/10.1038/s41928-026-01639-8.