Diabetes if untreated results in prolonged hyperglycaemia which can lead to diabetic retinopathy, neuropathy, and nephropathy. Treatments for diabetes include oral medication and insulin therapy. The risk associated with insulin therapy is hypoglycaemia. Diabetics monitor their blood glucose levels using capillary blood
glucose monitors or Continuous Glucose Monitoring Systems (CGMS).
The aim of the artificial pancreas is to improve the quality of diabetics’ lives. An
electromechanical artificial pancreas essentially consists of a glucose sensor, drug
delivery system, and a controller. An issue with subcutaneous glucose sensing and
insulin delivery is the time lags between venous and interstitial fluid glucose
concentrations. An implantable artificial pancreas employing intravenous glucose
monitoring and insulin delivery more closely mimics pancreatic insulin secretions
to the portal vein.
Model Predictive Control (MPC) can be extended for hypoglycaemia avoidance
and to take various constraints into account. Parameters of models used have
inter-patient variance; hence, patient-specific models are preferred. Patient parameters can also change over time with physiological changes. An artificial pancreas
system that adaptively tunes the patient-specific controller is proposed.
The artificial pancreas is to be implanted in the ilium and the device has to fit the
space and encapsulate the controller and the insulin delivery system. Automatic
insulin dosage is handled by a low-power microcontroller, with wireless communication for data logging. As an implantable glucose sensor is planned to be used in
the artificial pancreas, two fluid flow connections integrating intravenous glucose
sensing and insulin delivery were investigated and compared.