Scientists at the University of Chicago Pritzker School of Molecular Engineering have developed a soft, stretchable computing patch that can run artificial intelligence directly on the human body.
The new technology could lead to wearable devices that analyze health problems instantly without needing to send data to an external computer or cloud server.
The research, carried out with scientists from Argonne National Laboratory, was published in the journal Nature Electronics.
Today’s wearable devices, such as smartwatches and fitness trackers, can collect information like heart rate, movement, and sleep patterns.
But most of the actual analysis happens somewhere else. The devices usually send the data wirelessly to a smartphone or remote server, where AI systems process the information.
That delay may not matter for counting steps or tracking exercise, but it becomes a serious problem during medical emergencies.
For example, dangerous heart rhythm problems such as ventricular fibrillation require decisions within milliseconds. Waiting for data to travel back and forth to a remote server could take too long.
The new patch is designed to solve that issue by performing the AI calculations directly on the body itself.
The device works using special stretchable electronic components called organic electrochemical transistors. Unlike the transistors inside normal computer chips, these devices process information using both electrical signals and ions moving through a gel-like material. Because of this design, the transistors can also store information in a way that resembles how brain synapses work.
One of the biggest challenges was manufacturing. Traditional computer chip production methods use heat and chemicals that would damage flexible materials designed to stretch like skin. The gel material inside the transistors also tended to spread like liquid, which could cause short circuits.
To overcome these problems, the researchers created a new type of polymer gel that hardens when exposed to ultraviolet light. This allowed them to produce highly detailed transistor arrays on flexible surfaces. The new method can create about 10,000 transistors in a single square centimeter.
To test the technology, the team used the patch to run an AI system designed to detect dangerous electrical wave patterns in the heart. Using real heart data from a human donor, the device identified the wave positions with 99.6% accuracy, even when stretched to one and a half times its normal size.
In another test, the system analyzed several health measurements, including cholesterol, blood sugar, heart rate, and ECG readings, to estimate heart attack risk. It achieved an accuracy of 83.5%.
The researchers believe this is only the beginning. Future versions could combine sensors, wireless communication, and AI processing into a single flexible device that continuously monitors a person’s health in real time.
According to the team, the goal is to create wearable or implantable systems that act almost like an “instant doctor,” capable of sensing health problems, analyzing them immediately, and responding without delay.
The breakthrough could eventually lead to smarter medical wearables that are faster, more accurate, and more deeply connected to the human body than anything available today.
