Optimized type-2 fuzzy controller based on IoMT for stabilizing the glucose level in type-1 diabetic patients

Area of Science:

• Biomedical Engineering
• Control Systems
• Artificial Intelligence

Background:

• Advancements in the Internet of Medical Things (IoMT) enable continuous monitoring and control of blood glucose levels (BGL) for type-1 diabetic patients (T1DPs) using artificial pancreas systems.
• Regulating BGL is complex, necessitating robust controllers for exogenous insulin infusion to prevent adverse glycemic events.

Purpose of the Study:

• To propose an optimized interval type-2 fuzzy (IT2F) logic controller for an artificial pancreas system, leveraging IoMT for T1DPs.
• To enhance BGL regulation and mitigate risks of hyperglycemia and hypoglycemia during insulin therapy.

Main Methods:

• Utilized an optimized interval type-2 fuzzy (IT2F) controller integrated with an artificial pancreas system.
• Employed the grey wolf optimizer (GWO) meta-heuristic algorithm to tune the footprint of uncertainty in IT2F membership functions.
• Modeled blood glucose dynamics and meal disturbances using the nonlinear extended Bergman minimal model (EBMM) with uncertainty.
• Validated the controller's performance using simulations on the MATLAB/Simulink platform.

Main Results:

• The proposed IT2F controller effectively regulated BGL in type-1 diabetic patients.
• The controller successfully avoided both hyperglycemia and hypoglycemia under nominal model parameters.
• The system demonstrated robustness against parametric uncertainty and unpredictable meal disturbances.

Conclusions:

• The optimized IT2F-based artificial pancreas controller offers a promising solution for T1DPs.
• The integration of IoMT and advanced control strategies like GWO-tuned IT2F logic enhances glycemic management.
• This approach provides a safer and more effective method for exogenous insulin delivery, reducing the risk of severe glycemic excursions.