Research Database
Facilitating diabetes management with an AI-personalized carbohydrate-free bolus calculator for smart insulin pens.
Anas El, PhD
Institution:
University of Virginia
Grant Number:
4-22-PDFPM-16
Type of Grant:
Translational
Diabetes Type:
Type 1 Diabetes
Therapeutic Goal:
Manage Diabetes
Project Date:
-
Project Status:
active

Research Description

Most patients with type 1 diabetes use multiple daily injections of insulin to treat this disease. They need to self-administer, usually using an insulin pen, an insulin dose before consuming their meals. Patients calculate mealtime insulin dose to cover meal-nutrients, principally, carbohydrates. However, this is a challenging task that requires education in carbohydrate counting and good numeracy skills. This is particularly problematic for patients using insulin pens because of the lack of a built-in bolus calculator. The new generation of insulin pens, called smart insulin pens, can connect with smartphones and use a bolus calculator. Still, the existing bolus calculators are not designed for this population. This project proposes a new bolus calculator for users of smart insulin pens and glucose sensors that do not require counting carbohydrates. Instead, it only requires a qualitative meal selection, e.g., less than usual, usual, or more than usual meal portions. This simplification will be achieved by leveraging an artificial intelligence approach that recognizes non-obvious individual patterns in data the user’s habits. Successful completion of this project will result in the development and clinical investigation of an AI-based bolus calculator that facilitates mealtime insulin calculation. This approach will be suited for users with of smart insulin pen with a glucose sensor, as it will be safe and improve their quality of life and glycemic outcomes.

Research Profile

What area of diabetes research does your project cover? What role will this particular project play in preventing, treating and/or curing diabetes?

My project is about optimizing and facilitating glucose regulation after meal consumption for people with type 1 diabetes using insulin bolus injections. In recent years, we have seen tremendous progress in optimizing the therapy for type 1 diabetes patients using closed-loop insulin delivery systems. However, most people around the world still use injections. My project aims to leverage artificial intelligence techniques with a combination of novel sensing technologies (intermittently or real-time continuous glucose monitoring systems and connected smart insulin pens) to introduce a novel simplified bolus calculator for this population. The bolus calculator will only require the users to qualify their meals in terms of carbohydrate content instead of exactly quantifying the carbohydrate contents of their meals.

If a person with diabetes were to ask you how your project will help them in the future, how would you respond?

This project introduces an AI-driven bolus calculator with a simplified qualitative meal approach. The short-term purpose of this project is to simplify the therapy for people using insulin bolus injections. However, this technology can be included in closed-loop insulin delivery systems. In short, for a person with diabetes, this technology aims on improving post-meal glucose control.

Why important for you, personally, to become involved in diabetes research? What role will this award play?

I have been working in diabetes technology research since 2016. During this time, I truly experienced how people with diabetes live with the life-long burden of optimizing their insulin doses. I desire to work towards alleviating this burden using novel technologies, and this project is a step towards achieving this goal.

In what direction do you see the future of diabetes research going?

Diabetes research spans multiple subfields. Regarding the field of new technologies for diabetes, I see future research focused on more patient-centric technologies. These are algorithms will accompany the user in making insulin-related decisions and tailor the treatment for their sake. This should increase adherence to treatment and result in more people achieving glucose targets and fewer people with long-term macrovascular complications.