¶¶Òőapp

By Dave DeFusco

What if a tiny ring on your finger could tell when you’re stressed, predict your heart health and alert you to dangerous changes in your body—all in real time? That’s the bold promise behind new research presented by a Katz School team that unveiled a “digital twin” healthcare framework powered by smart rings in London at the 2025 IEEE INFOCOM Conference, a top-ranked conference on networking in the research community. 

The research, conducted by Ashikur Nobel, a Ph.D. student in Mathematics at the Katz School; Dr. Honggang Wang, chair of the Katz School’s Department of Graduate Computer Science and Engineering; and Dr. Hua Fang, professor of computer and information science at the University of Massachusetts Dartmouth and the Chan Medical School, received the only 2025 IEEE INFOCOM Best Poster Award.

“This project started with a simple idea,” said Nobel. “Can we take the data collected from smart rings, combine it with our physical biomarkers, such as weight, blood pressure and cholesterol, and use it to build a virtual version of a person’s health—a ‘digital twin’ that can learn, adapt and even predict what’s coming next?”

Originally used in industries like aviation and manufacturing, a digital twin is a virtual model of a real system that is continuously updated with live data. In the context of healthcare, it means creating a virtual model of a person’s health using real-time data from wearable sensors, like those embedded in smart rings.

Smart rings like the Oura or Samsung Galaxy Ring are small, discreet and packed with powerful sensors that can track heart rate, blood oxygen levels, sleep patterns, physical movement, and more. But Nobel wanted to go further—not just collect data but use it to understand and predict health in real time.

Their proposed framework connects a smart ring to a nearby device, like a smartphone, via Bluetooth, avoiding the need to constantly send sensitive health data to the cloud. That local processing approach makes it faster, more secure and better for user privacy.

“The ring collects raw data—your pulse, how much you’re moving, your oxygen levels,” said Nobel. “We process that data using AI models to spot trends and make predictions. For instance, we can detect if your stress is rising or if your oxygen levels drop unexpectedly. Because it’s a digital twin, it can keep monitoring and even fill in the gaps when the ring is off.”

This “filling in the gaps” is made possible by advanced AI models, particularly a type known as a Transformer model which is a widely used architecture across many applications, including tools like ChatGPT. These models can learn from patterns over time and predict future outcomes, such as the likelihood of abnormal heart rhythms or poor sleep quality.

“Let’s say the ring goes into power-saving mode or is temporarily taken off,” said Nobel. “The digital twin doesn’t just stop working; it keeps going, using what it’s already learned to estimate how your body is doing.”

That continuous prediction could be vital in early warning systems for conditions like heart attacks, breathing issues or chronic stress-related illnesses. One of the standout features of the project is how clearly it can show your health data. The team designed user-friendly visual dashboards that translate complex biological signals into simple charts and alerts. Imagine opening an app and seeing not just your heart rate, but how it’s changed over time, how it correlates with your sleep and whether stress might be affecting it.

“We wanted to make the invisible visible,” said Nobel. “The digital twin can explain what's going on inside your body in a way that makes sense, not just to doctors but to everyday users.”

Traditional health apps often rely on the cloud for heavy data processing, which raises concerns about privacy and speed. Nobel’s framework keeps most of the work local—on your phone or wearable device—ensuring faster results and better protection for personal information. That’s a big deal, said Dr. Wang.

“We are in an era where personalized healthcare is possible, but only if we protect users’ trust,” he said. “This smart ring digital twin approach is not just innovative, it’s secure, efficient and scalable. It brings together mathematics, engineering and health science to put power back in the hands of individuals.”

Wang noted that this research lays the foundation for a future where wearable technology doesn’t just track, but truly understands, the user. While the current prototype focuses on two main types of data—pulse signals from the ring’s light-based sensors and motion data from its accelerometer—the team hopes to expand it with additional sensors for skin temperature, hydration and even mood.

Initial results have been promising. The team tested their models using self-collected and real-world datasets and found strong predictive performance. While traditional models, like Long Short-Term Memory (LSTM) networks, were used in early stages, the research has since advanced to more powerful Transformer-based AI models. But as Nobel emphasized, this is just the beginning.

“We’ve shown the digital twin works in principle. Now we need to test it in clinical settings, on diverse populations and with more complex health conditions,” he said. “Our dream is to one day have a tiny ring that quietly watches over your health and warns you before a crisis happens.”