A smart artificial pancreas is transforming diabetes management thanks to innovative technology.

A research team from the University of Virginia in the United States has developed an innovative interactive artificial pancreas system based on "digital twin" technology to improve the management of type 1 diabetes.

A smart artificial pancreas is transforming diabetes management thanks to innovative technology.

Image: Natali_Mis / Gettyimages.ru

The new system uses a method known as Adaptive Biobehavioral Control (ABC), which allows the artificial pancreas to flexibly adapt to each user's physiological and behavioral changes through periodic updates and real-time data analysis.

Dr. Boris Kovachev, director of the Diabetes Technology Center at the University of Virginia and lead author of the study, explained that this innovation improves the insulin delivery algorithm every two weeks, increasing the efficiency of the artificial pancreas and making it more responsive to patients' needs.

He added that the system also provides an interactive digital simulation tool, allowing users to safely test different settings—such as changing nighttime insulin doses—in a virtual environment before actually implementing them on their devices.

The concept of a "digital twin" is based on virtual models that simulate the performance of real-world systems and processes. This technology was first used in NASA's Apollo space program in the 1960s, but for the first time, it is being used on a personal, individual basis for diabetes patients via dedicated cloud models.

This technology allows users to understand how their daily adjustments—such as exercise or dietary changes—affect the performance of the artificial pancreas, enhancing treatment accuracy and reducing risks.

A six-month clinical study showed that users who adopted ABC increased the time they spent within the safe blood sugar range from 72% to 77%, and their average HbA1c level decreased from 6.8% to 6.6%.

Kovachev emphasized that the new system addresses two major challenges that have made it difficult for previous systems to achieve optimal results: the first is managing large fluctuations in blood sugar throughout the day, caused by eating or physical activity, and the second is what is known as the "plateau phase," a point at which users' improvement declines after an initial successful period.

The research team explained that the system offers two strategies to address these challenges, by combining the digital twin with an algorithm that learns from the user's behavior and adapts to it over time.

Kovachev concluded, "Co-adaptation between humans and machines is essential in conditions such as type 1 diabetes, where artificial intelligence and the user share treatment decisions. Digital twin technology is a key pillar in enabling this intelligent interaction."

It's worth noting that the artificial pancreas is an advanced medical system used to help type 1 diabetes patients automatically regulate their blood sugar levels by mimicking the function of the body's natural pancreas, which fails to produce insulin.

Source: Interesting Engineering - Published on 11-06-2025 - https://ar.rt.com/101vn