Anas El Fathi, PhD
Anas El Fathi, PhD, assistant professor with the University of Virginia Center for Diabetes Technology, received a Career Development Award grant totaling nearly $1 million from Breakthrough T1D to develop next-generation artificial intelligence technology designed to simplify daily management for people living with Type 1 Diabetes.
The five-year award will support a research project titled “ABX: An AI-Driven Automated Bolus Extension to Eliminate Meal Announcements in Hybrid Closed-Loop Systems.” The project will run from March 1, 2026, through February 28, 2031.
The research team aims to develop and evaluate Automated Bolus eXtension (ABX), a software upgrade for existing insulin pump systems that could automatically deliver mealtime insulin without requiring users to count carbohydrates, enter meal information, or manually deliver insulin doses. By learning from each user’s glucose responses, daily habits, and insulin patterns, ABX is designed to continuously adapt and make safe insulin dosing decisions throughout the day.
Even with today’s advanced automated insulin delivery systems, people with type 1 diabetes must still manually estimate carbohydrates and announce meals to their insulin pump. This step can be stressful, error-prone, and is often skipped — especially among teens and young adults— leading to dangerous spikes in blood glucose and increased risk of long-term complications.
“Mealtime insulin dosing remains one of the most burdensome aspects of diabetes care,” said El Fathi. “Our goal is to create a system that learns from each person’s real-world data and safely automates this process, reducing the daily mental workload while improving glucose control.”
The study will enroll 48 adolescents and young adults who use commercially available insulin pumps. Over a 20-week clinical study, researchers will evaluate how well the ABX system improves glucose outcomes and reduces the daily burden of diabetes management. Participants will compare their usual hybrid closed-loop insulin pump system with and without ABX, and later test an enhanced version of the technology that uses advanced artificial intelligence to refine insulin dosing decisions over time.
Researchers will measure improvements in Time-in-Range, the number of hours per day blood glucose remains in a healthy range, along with reductions in glucose spikes after meals and episodes of dangerously low blood sugar.
Filed Under: Research