Closed loop control of physiological glucose

1. A system to control glucose in a patient, the system comprising:

a medication delivery device configured to deliver a medication dose to the patient; and

a controller operably coupled to the medication delivery device and including control logic operative to:

execute a multi-model predictive controller algorithm at least in part by executing a plurality of models each comprising a plurality of model parameters having a set of values, wherein executing the plurality of models includes propagating multiple state vectors,

select one of the plurality of executed models based at least in part on differences between glucose values of the propagated state vectors and previously measured glucose values,

calculate a basal deviation using the selected model, wherein the basal deviation is a dosage amount above or below a predefined basal profile, and

determine a first medication dose based at least in part on the calculated basal deviation and the predefined basal profile.

2. The system of claim 1 , wherein the control logic is further operative to:

predict future glucose levels of the patient based at least in part on the selected executed model, wherein the determination of the first medication dose is based at least in part on the predicted future glucose levels.

3. The system of claim 2 , wherein the control logic is further operative to:

determine a difference between the predicted future glucose levels and a nominal target glucose value, wherein the determination of the first medication dose is based at least in part on the difference between the predicted future glucose levels and the nominal target glucose value.

4. The system of claim 2 , wherein the prediction of future glucose levels is based at least in part on the predefined basal profile.

5. The system of claim 2 , wherein the prediction of future glucose levels is based at least in part on the predefined basal profile and without meal boluses.

6. The system of claim 1 , wherein the control logic being operative to select one of the plurality of executed models comprises the control logic being operative to:

select one of the plurality of executed models in response to comparing a calculated level of interstitial glucose and a measured level of interstitial glucose.

7. The system of claim 6 , wherein the measured level of interstitial glucose comprises a plurality of past measured levels of interstitial glucose.

8. The system of claim 1 , wherein the control logic being operative to select one of the plurality of executed models comprises the control logic being operative to:

select one of the plurality of executed models based at least in part on a nominal target glucose value.

9. The system of claim 8 , wherein the nominal target glucose value is increased in response to an exercise announcement.

10. The system of claim 8 , wherein the nominal target glucose value is increased in response to a meal announcement.

11. The system of claim 1 , wherein the control logic is further operative to:

re-execute the multi-model predictive controller algorithm,

select a different one of the plurality of executed models, and

determine a second medication dose based at least in part on the selected executed model.

12. The system of claim 1 , wherein the control logic is further operative to:

re-execute the multi-model predictive controller algorithm including re-executing the plurality of models using the plurality of model parameters and the set of values,

select a different one of the plurality of executed models, and

determine a second medication dose based at least in part on the selected executed model.

13. The system of claim 1 , wherein the plurality of model parameters includes at least one of an insulin sensitivity, an insulin time constant, a meal action time constant, a sensor time constant, an insulin-to-carbohydrate ratio, an input from a user interface, and a controller gain value.

14. The system of claim 1 , wherein the plurality of model parameters includes an insulin sensitivity, an insulin time constant, a meal action time constant, a sensor time constant, and an insulin-to-carbohydrate ratio.

15. The system of claim 1 , wherein each of the plurality of models includes a set of linear differential equations that calculate levels of physiological glucose and interstitial glucose.

16. The system of claim 15 , wherein the set of linear differential equations model storage and transportation of insulin in the patient.

17. The system of claim 1 , wherein the controller is further operative to transmit a request including the determined first medication dose to the medication delivery device, the medication delivery device includes insulin, and the medication delivery device is configured to deliver the insulin to the patient based at least in part on the determined first medication dose.

18. The system of claim 1 , further comprising:

a user interface operably coupled to the controller and configured to receive input from the patient.

19. The system of claim 1 , further comprising:

a glucose measurement device operably coupled to the controller and configured to measure glucose data associated with the patient and transmit the measured glucose data to the controller, wherein the determination of the first medication dose is based at least in part on the measured glucose data.

20. A method to control glucose in a patient, the method comprising:

executing, using one or more controllers, a multi-model predictive controller algorithm by executing a plurality of models each of which comprise a plurality of model parameters having a set of values, wherein executing the plurality of models includes propagating multiple state vectors;

selecting, using the one or more controllers, one of the plurality of executed models based at least in part on differences between glucose values of the propagated state vectors and previously measured glucose values;

calculating, using the one or more controllers, a basal deviation using the selected model, wherein the basal deviation is a dosage amount above or below a predefined basal profile;

determining, using the one or more controllers, a first medication dose based at least in part on the calculated basal deviation and the predefined basal profile; and

transmitting, using the one or more controllers, the first medication dose to a medication delivery device configured to deliver the medication dose to the patient.

21. The method of claim 20 , further comprising:

predicting, using the one or more controllers, future glucose levels of the patient based at least in part on the selected executed model, wherein the determination of the first medication dose is based at least in part on the predicted future glucose levels.

22. The method of claim 21 , further comprising:

determining, using the one or more controllers, a difference between the predicted future glucose levels and a nominal target glucose value, wherein the determination of the first medication dose is based at least in part on the difference between the predicted future glucose levels and the nominal target glucose value.

23. The method of claim 21 , wherein the prediction of future glucose levels is based at least in part on the predefined basal profile.

24. The method of claim 20 , further comprising:

re-executing the multi-model predictive controller algorithm including re-executing the plurality of models using the plurality of model parameters and the set of values;

selecting a different one of the plurality of executed models; and

determining a second medication dose based at least in part on the selected executed model.

25. The method of claim 20 , wherein the plurality of model parameters includes an insulin sensitivity, an insulin time constant, a meal action time constant, a sensor time constant, and an insulin-to-carbohydrate ratio.

26. The method of claim 20 , wherein executing a plurality of models includes executing a set of linear differential equations to calculate levels of physiological glucose and interstitial glucose, wherein the set of linear differential equations model storage and transportation of insulin in the patient.

27. The method of claim 20 , wherein selecting one of the plurality of executed models is in response to comparing a calculated level of interstitial glucose and a measured level of interstitial glucose.

28. The method of claim 20 , wherein selecting one of the plurality of executed models is based at least in part on a nominal target glucose value.