07-07-2012, 11:43 AM
Model Reference Adaptive Scheme for Multi-drug Infusion for
Blood Pressure Control
Model Reference Adaptive Scheme.pdf (Size: 2.46 MB / Downloads: 58)
Introduction
If blood pressure is controlled and oscillations in the homodynamic variables are reduced, patients
experience fewer complications after surgery. In clinical practice, this is usually achieved
using manual drug delivery. Given that different patients have different sensitivity and reaction
time to drugs, determining manually the right drug infusion rates may be difficult. This is a
problem where automatic drug delivery can provide a solution, especially if it is designed to
adapt to variations in the patient’s model.
Various automatic control techniques have been used to control the hemodynamic variables.
Many studies have focused on the infusion of a single drug to lower the patients’ blood pressure
and maintain it at the desired level using in particular the vasoactive drug sodium nitroprusside
SNP [1–8, 13]. Slate and Sheppard [3] used a one-drug patient model and implemented a
nonlinear digital proportional-integral-derivative (PID) controller to regulate the mean arterial
pressure.
Patient Response Model
The objective of the control system is to decrease the patient’s mean arterial pressure and increase
the cardiac output to the desired value by tracking the reference signals. The patient
hemodynamic model [12] is defined by the linear small-signal first-order transfer function matrix
given in equation (1). The drugs used to control the variables CO and MAP are dopamine
(DPM) and sodium nitroprusside (SNP). DPM increases both CO and MAP while SNP increases
CO and decreases MAP. The drug infusion rates are measured in (mg=min:kg). Cardiac
output is measured in (ml=min:kg). Mean arterial pressure is measured in millimeters of
mercury (mmHg).
Simulation Results
The proposed algorithm has been implemented, and tested through a set of experiments by
considering two different control objectives as described in section 4.1 and 4.2.
Table 1 lists the nominal values and the range of the parameters of the patient’s model which
were simulated using Matlab Simulink as shown in figure 3. The simulations were done for
different patient’s sensitivity. The MRAC has been implemented to compute the drugs infusion
rates which are the inputs to the MIMO patient model.
Simulations were conducted for all the parameters values in the range to find by trial and error
the best weighting matrices A and B. The proportional and integral gains were then calculated
from equation (8) and (9) using these weighting matrices.
Conclusions
The paper has presented an adaptive multi-drug control scheme for blood pressure control. The
proposed scheme was designed and evaluated by simulating variations in patient sensitivity and
disturbances in cardiac output and mean arterial pressure. Two drugs were used; dopamine and
nitroprusside. The simulation results have confirmed that MRAC is potentially useful for regulating
MAP and CO by computing the DPM and SNP infusion rates. The proposed algorithm
demonstrated better performance as compared to an non-adaptive PID controller. In the simulations
studies the proposed scheme produced better performance compared to reported results,
particularly for the updated controller’s gain K22 (mean arterial pressure gain). This includes
shorter settling time and very small or no overshoot when the patient sensitivity K22 less than
or equal -20. For further work, the proposed controller will be tested with a wide range of
patients’ sensitivities using more than two drugs.