Use of a "fuzzy logic" controller in a closed-loop artificial pancreas

Area of Science:

• Biomedical Engineering
• Control Systems
• Diabetes Technology

Background:

• Current artificial pancreas systems often use complex mathematical models that struggle with physiological variability.
• Fuzzy logic (FL) controllers offer a model-free approach, simplifying integration of factors like illness and stress.
• FL controllers allow for personalized dosing aggressiveness based on individual patient needs.

Purpose of the Study:

• To evaluate a fully automated fuzzy logic (FL) closed-loop insulin dosing controller.
• To assess the controller's ability to personalize insulin delivery without requiring differential equations.
• To determine the efficacy of the FL controller in maintaining target blood glucose levels.

Main Methods:

• A pilot study was conducted in a controlled environment with subjects on bed rest.
• Participants received two carbohydrate-controlled meals without announcement or pre-meal insulin bolus.
• The primary endpoint was the avoidance of hypoglycemia (<60 mg/dL), with secondary endpoints assessing hyperglycemia and hypoglycemia frequency/severity.

Main Results:

• Seven out of ten subjects who completed the study maintained average blood glucose levels within the target range (70-200 mg/dL) for 76% of the 24-hour period.
• Two subjects were discontinued due to hypoglycemia, and one due to sensor failure.
• The FL controller demonstrated effective glucose management in a controlled setting.

Conclusions:

• The fuzzy logic controller presents a promising alternative to model-based controllers in closed-loop artificial pancreas systems.
• This FL controller facilitates easy customization of glucose control levels for individual patient requirements.
• The findings support the viability of FL controllers for advanced diabetes management.