Gluclas

Gluclas generates glucose infusion rate (GIR) adjustments using a proportional-integral-derivative (PID) control strategy to maintain target blood glucose (BG) concentrations during glucose clamp experiments.

Key Features:

• Proportional-Integral-Derivative (PID) controller: Employs a PID controller to compute real-time GIR suggestions from BG measurements.
• Anti-Wind-Up Scheme: Implements an anti-wind-up mechanism to prevent integrator saturation when actuator limits are reached.
• Sampling Jitter Correction: Applies corrections for sampling jitter to compensate timing variability in BG measurement and actuation.

Scientific Applications:

• In Silico Validation: Simulations on 50 virtual subjects using a glucose metabolism simulator that included models for measurement error and sampling delay produced plateau coefficient of variation (CV) consistently below 5%.
• Hyperglycemic Clamp: Achieved an average BG level of 12.18 mmol/l (target: 12.4 mmol/l).
• Euglycemic Clamp: Reached an average BG level of 4.92 mmol/l (target: 5.5 mmol/l).
• Hypoglycemic Clamp (in silico): Maintained an average BG level of 2.38 mmol/l (target: 2.5 mmol/l).
• In Vivo Hypoglycemic Trial: First in vivo trials in three subjects produced average BG levels of 2.56–2.68 mmol/l (target: 2.5 mmol/l) with CV below 5%.

Methodology:

Computational methods explicitly include a proportional-integral-derivative (PID) controller that generates real-time GIR suggestions, an anti-wind-up scheme for actuator limits, corrections for sampling jitter, and in silico validation using a glucose metabolism simulator modeling measurement error and sampling delay on 50 virtual subjects.