Page 289 - Glucose Monitoring Devices

296 CHAPTER 15 Automated closed-loop insulin delivery

Despite the advancement in AP systems, a clinically significant number of hypo-
glycemic events were reported during tests of closed-loop delivery systems [41]. To
further reduce the risk of hypoglycemia, dual-hormone closed-loop delivery systems
are developed that combine insulin delivery with subcutaneous glucagon delivery
[42e45]. Glucagon, a hormone produced in the alpha cells of the pancreas, counters
the effects of insulin through promoting the breakdown of glycogen to glucose by
the liver, thus stabilizing glucose concentrations and preventing hypoglycemia.
Other approaches to reduce the risk of hypoglycemia include suggesting rescue car-
bohydrates with hypoglycemia, which is predicted by the control algorithms.
The incorporation of additional physiological variables in mAP systems and the
estimation of model parameters and meal consumption enhance the available infor-
mation on the physiologic and metabolic state of the subject. In addition to these
innovations, the amount of previously administered insulin that is present in the
blood or the subcutaneous space, called as the insulin on board (IOB), must be quan-
tified to prevent overdosing. The IOB is typically determined in infusion pumps
through static approximations of the insulin action curves [30,46,47]. The insulin
action curves, though convenient, do not account for the time-varying dynamics
and kinetics of the metabolic states of individuals due to the diurnal variations in
the insulin diffusion, absorption, and utilization. Therefore the insulin decay profiles
and action curves used in calculating IOB are not accurate enough over the diverse
conditions encountered throughout the day to be reliably used in an mAP system. In
contrast, accurate estimates of the insulin concentration in the bloodstream can be
obtained using CGM measurements with adaptive observers designed for simulta-
neous state and parameter estimation. The estimated plasma insulin concentration
(PIC) can be subsequently used to design a predictive control algorithm that is
dynamically constrained by the estimated PIC and thus explicitly considers the
insulin concentration in the bloodstream as part of the optimal control solution
[48]. Incorporating PIC constraints in the optimal control problems can prevent
insulin stacking that may lead to hypoglycemia.
The remainder of the chapter describes the various facets of the AP systems in
detail. First, the popular closed-loop glycemic control algorithms, including PID
control and MPC, are discussed. Then, dynamic glucose-insulin modeling methods
and adaptive control techniques are presented, followed by a review of state and
parameter estimation techniques. Finally, closed-loop glycemic control results are
demonstrated using an adaptive MPC approach and possible future research direc-
tions are briefly outlined.

Closed-loop glycemic control algorithms
Proportional-integral-derivative control
PID control is widely adopted in various industries to regulate an output variable by
manipulating an input variable. PID control computes the control action based on the