Data-Driven Robust Control for Type 1 Diabetes Under Meal and Exercise Uncertainties

TL;DR

This paper introduces a fully automated, data-driven robust model-predictive control system for an artificial pancreas that personalizes insulin delivery for Type 1 diabetes patients by accounting for meal and exercise uncertainties.

Contribution

It develops a novel control framework that learns uncertainty sets from data to enable personalized, automated insulin regulation without manual meal or activity input.

Findings

Successfully regulates glucose levels in virtual patients under high carbohydrate disturbances.

Demonstrates robustness to individual variations in meal and exercise patterns.

Achieves high accuracy in in silico simulations without explicit meal announcements.

Abstract

We present a fully closed-loop design for an artificial pancreas (AP) which regulates the delivery of insulin for the control of Type I diabetes. Our AP controller operates in a fully automated fashion, without requiring any manual interaction (e.g. in the form of meal announcements) with the patient. A major obstacle to achieving closed-loop insulin control is the uncertainty in those aspects of a patient's daily behavior that significantly affect blood glucose, especially in relation to meals and physical activity. To handle such uncertainties, we develop a data-driven robust model-predictive control framework, where we capture a wide range of individual meal and exercise patterns using uncertainty sets learned from historical data. These sets are then used in the controller and state estimator to achieve automated, precise, and personalized insulin therapy. We provide an extensive in silico evaluation of our robust AP design, demonstrating the potential of this approach, without explicit meal announcements, to support high carbohydrate disturbances and to regulate glucose levels in large clusters of virtual patients learned from population-wide survey data.