University of Chicago researchers develop AI-powered skin patch for real-time health monitoring

Researchers at the University of Chicago Pritzker School of Molecular Engineering have developed a wearable computer patch capable of running artificial intelligence (AI) models directly on the skin, potentially transforming how health conditions are monitored and treated in real time.

The technology, detailed in the journal Nature Electronics, performs both data collection and AI analysis on the body itself. Unlike most current wearable devices, which send data to smartphones or cloud-based systems for processing, the new patch carries out computations instantly without relying on external hardware or wireless communication.

Research lead Sihong Wang described the concept as having a “personal, instantaneous doctor integrated into [users’] devices”. Although the technology is still in the research stage and not yet ready for commercial use, the team believes it could significantly improve the speed and efficiency of health monitoring systems.

On-device AI brings faster medical analysis

Modern wearable devices such as smartwatches are already capable of measuring a range of health indicators, including heart rate, movement, blood oxygen levels and electrocardiogram (ECG) signals. However, these devices generally serve as data-collection tools, transferring information elsewhere for analysis.

The newly developed patch takes a different approach by processing data directly on the skin. According to the researchers, the system can perform AI inference within milliseconds, eliminating delays associated with sending information to external devices or cloud servers.

This low-latency processing could be particularly valuable in emergency medical situations. Conditions such as ventricular fibrillation, a dangerous heart rhythm disorder, require rapid detection and intervention. Even small delays in analysing data can affect treatment outcomes.

In addition to faster response times, the researchers said the technology could reduce power consumption and improve privacy. Because sensitive health information remains on the device rather than being transmitted across networks, the risk of data exposure is reduced.

The breakthrough was made possible by stretchable transistors that can bend and conform to the skin’s surface. Traditional silicon-based computer chips are rigid and unsuitable for such applications. However, developing the flexible electronics presented its own engineering challenges.

“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 added.

High accuracy demonstrated in heart monitoring tests

A major focus of the research was the detection and monitoring of ventricular fibrillation. To evaluate the system’s capabilities, the researchers tested the patch using a donated human heart.

The study found that the device could identify fibrillation wavefront positions with an accuracy of 99.6%. This level of precision demonstrates the potential of wearable AI systems to provide detailed and reliable medical insights directly at the point of measurement.

According to the research paper, the platform supports a variety of edge-computing functions for health data analysis. These include multilayer perceptron (MLP) models for heart attack prediction, as well as convolution-based operations for tracking arrhythmia wavefronts across the heart surface.

The researchers believe the technology could eventually support a new generation of implantable devices. Future systems may be able to collect high-resolution signals directly from living organs and process that information in real time.

Such implantable computing platforms could provide doctors with highly accurate data while reducing dependence on external monitoring equipment. The concept represents a broader move towards intelligent healthcare devices that operate directly within or on the body.

Potential applications extend beyond healthcare

While healthcare remains the primary focus of the project, the researchers believe the underlying technology could have applications in several other fields.

One potential area is robotics. Wearable or implantable computing systems that process information directly at the source could help robots achieve more responsive and human-like sensory capabilities. This could be especially useful in environments where communication networks are unreliable or unavailable.

The study highlighted the possibility of using such systems in disaster recovery operations, where robots may need to operate independently without constant access to cloud-based computing resources. Local processing could allow machines to react more quickly to changing conditions and hazards.

Researchers also referenced experiments involving a small ant-like robot. In those tests, integrated computing and reinforcement learning techniques enabled the robot to navigate environments with performance comparable to traditional computer simulations.

Looking ahead, the technology reflects a broader shift in the AI industry towards edge computing. Rather than relying on distant data centres to perform calculations, future AI systems may increasingly process information locally on devices, reducing latency and improving efficiency.

Although commercial deployment is likely several years away, the research highlights how advances in flexible electronics and AI could reshape healthcare, robotics, and other industries that depend on rapid, real-time decision-making.