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

Most current model-based approaches to closed-loop artificial pancreas systems rely on mathematical equations describing the human glucoregulatory system; however, incorporating the various physiological parameters (e.g., illness, stress) into these models has been problematic. We evaluated a fully automated "fuzzy logic" (FL) closed-loop insulin dosing controller that does not require differential equations of the glucoregulatory system and allows clinicians to personalize dosing aggressiveness to meet individual patient requirements.

This pilot study evaluated the FL controller in the setting of bed rest in a very controlled environment. Two carbohydrate-controlled meals were given (30 g at 8 a.m. and 60 g at 2 p.m. without meal announcement or premeal bolus. The primary end point of the study was avoidance of hypoglycemia, defined at <60 mg/dL. Multiple end points related to the frequency and severity of hyperglycemia and hypoglycemia were also assessed.

Of the 12 subjects we recruited, 10 were enrolled, and seven completed the study. Two of the enrolled subjects were discontinued because of hypoglycemia; the other was discontinued because of sensor failure. Seven of the 10 subjects who completed the study had average blood glucose values of 165 mg/dL and were within a specified target blood glucose range (70-200 mg/dL) for 76% of the 24-h study period.

Our findings suggest that the FL controller provides a viable alternative to model-based controllers as a component of a closed-loop insulin delivery system. Furthermore, our FL controller allows clinicians to easily specify the level of glucose control based on each patient's clinical needs.