System and method for adjusting insulin delivery
The embodiments described herein may relate to methods and systems for adjusting insulin delivery. Some methods and systems may be configured to adjust insulin delivery to personalize automated insulin delivery for a person with diabetes. Such personalization may include receiving and/or determining one or more user specific dosage parameters.
1. A method of delivering insulin, the method comprising:
generating an indication that a probability of occurrence of an undesirable event exceeds a threshold responsive to a prediction by an event model; and
generating a recommended change to a personalization setting of an insulin therapy responsive to the indication,
wherein the event model is configured to predict an occurrence of the undesirable event responsive to a difference between a personalized model that represents a user's physiological response to insulin therapy and a generalized model that represents a generalized physiological response by a general population to an insulin therapy.
2. A method of delivering insulin, the method comprising:
generating an indication that a probability of occurrence of an undesirable event exceeds a threshold responsive to a prediction by an event model; and
generating a recommended change to a personalization setting of an insulin therapy responsive to the indication,
wherein the event model is configured to predict an occurrence of the undesirable event responsive to a difference between a personalized model that represents a user's physiological response to insulin therapy and a generalized model that represents a generalized physiological response to an insulin therapy, and
wherein the personalized model comprises a first relationship between a daily basal rate (BR) for a user and at least one of a carbohydrate-to-insulin ratio (CR) and an insulin sensitivity factor (ISF) for the user, wherein the generalized model comprises a second relationship between a general BR, a general CR, and a general ISF, and wherein the difference between the personalized model and the generalized model comprises a distance between a probability distribution of the second relationship and the first relationship.
3. The method of claim 2 , wherein generating the indication includes generating a visual depiction of the probability distribution of the general CR and general ISF for the BR of the user and an indicator of a location of the first relationship within the visual depiction.
4. The method of claim 2 , wherein generating the indication includes generating a plurality of curves illustrating the general BR relative to at least one of the general CR and the general ISF and an indicator of a location relative to the plurality of curves for at least one of the CR and the ISF for the user.
5. The method of claim 2 , wherein generating the indication comprises:
generating the probability distribution of the second relationship between the general BR, the general CR, and the general ISF as a general ellipsoidal distribution having a major axis and two minor axes, an intersection of the major axis and the two minor axes defining a midpoint of the probability distribution;
determining a location represented by the first relationship within the probability distribution; and
determining a distance between the location and the midpoint of the probability distribution.
6. The method of claim 5 , wherein the probability distribution of the second relationship comprises a multivariate normal distribution of logarithms of the general BR, the general CR, and the general ISF.
7. The method of claim 5 , wherein the probability distribution of the second relationship comprises a plurality of concentric ellipsoidal regions.
8. The method of claim 7 further comprising, responsive to determining that the location represented by the first relationship falls within an outer ellipsoidal region, determining that the first relationship exceeds a threshold distance.
9. The method of claim 8 , further comprising, generating a visual depiction of the probability distribution, the plurality of concentric ellipsoidal regions, and the location represented by the first relationship.
10. The method of claim 2 , wherein if the CR for the user is not received, the method further comprising determining the CR for the user based on the BR for the user according to CR=a*BR-b where a is a number between approximately 114 and 126, and b is a number between approximately 0.785 and 0.815.
11. The method of claim 2 , wherein if the ISF for the user is not received, the method further comprising determining the ISF for the user based on the BR for the user according to ISF=x*BR-y where x is a number between approximately 1115 and 1140,and y is a number between approximately 1.00 and 1.06.
12. The method of claim 11 , further comprising determining a probability that the first relationship falls outside the probability distribution, the probability distribution following a multivariate normal distribution with mean u and covariation matrix Σ and a distance D m of the first relationship (X) from the probability distribution Dm=x-uTΣ-1(x-μ), the probability that the first relationship falls outside the probability distribution determined by P(x-μΣ≤Dm)=γ( 32 ,Dm22)Γ(32) where γ is a lower incomplete Gamma function and Γ is a Gamma function.
13. The method of claim 12 , wherein the mean u is approximately μ=3.01112.37573.8645 and the covariation matrix Σ is approximately Σ=0.2843-0.1657-0.2216-0.16570.19780.1863-0.22160.18630.2968.
14. The method of claim 12 , wherein a magnitude of the probability that the first relationship falls outside the probability distribution is depicted visually by one of a color or a numerical score.
15. A method of delivering insulin, the method comprising:
determining a carbohydrate-to-insulin ratio (CR) based on a received daily basal rate (BR) for a user;
calculating a bolus dose of insulin to account for food to be ingested by the user based on the CR;
delivering the bolus dose of insulin; and
generating a message to an electronic insulin delivery device to deliver the bolus dose of insulin,
wherein determining the CR is performed according to CR=a*BR-b where a is a number between approximately 114 and 126, and b is a number between approximately 0.785 and 0.815.
16. A method of delivering insulin, the method comprising:
determining an insulin sensitivity factor (ISF) based on a received daily basal rate (BR) for a user;
calculating a bolus dose of insulin based on the ISF, the bolus dose of insulin to account for a blood glucose reading beyond a threshold variation from a target blood glucose level; and
delivering the bolus dose of insulin,
wherein determining the ISF is performed according to ISF=x*BR-y where x is a number between approximately 1115 and 1140, and y is a number between approximately 1.00 and 1.06.
17. The method of claim 16 , further comprising generating a message to an electronic insulin delivery device to deliver the bolus dose of insulin.
18. A method of delivering insulin, the method comprising:
determining a personalized model that represents a user's physiological response to insulin therapy;
predicting, via an event model, a probability of an occurrence of an undesirable event responsive to a difference between the personalized model and a generalized model that represents a generalized physiological response by a general population to an insulin therapy; and
generating an indication that a probability of occurrence of the undesirable event exceeds a threshold responsive to a prediction by an event model.
19. A method of delivering insulin, the method comprising:
determining a personalized model that represents a user's physiological response to insulin therapy;
predicting, via an event model, a probability of an occurrence of an undesirable event responsive to a difference between the personalized model and a generalized model that represents a generalized physiological response to an insulin therapy; and
generating an indication that a probability of occurrence of the undesirable event exceeds a threshold responsive to a prediction by an event model,
wherein determining the personalized model comprises:
receiving at least a daily basal rate (BR) for the user and only one of a carbohydrate-to- insulin ratio (CR) and an insulin sensitivity factor (ISF) for the user;
if the CR for the user is not received, determining the CR for the user based on the BR for the user;
if the ISF for the user is not received, determining the ISF for the user based on the BR for the user; and
determining a first relationship between the BR and at least one of the CR and the ISF for the user.
20. The method of claim 19 , wherein predicting, via the event model, the probability of the occurrence of the undesirable event responsive to the difference between the personalized model and the generalized model comprises:
comparing the first relationship with a probability distribution of a second relationship of the generalized model between a general BR, a general CR, and a general ISF; and
determining that the first relationship exceeding a threshold distance from the probability distribution.
21. The method of claim 20 , wherein generating the indication comprises: generating, at a user interface, the indication that the first relationship exceeds the threshold distance and providing a recommended variation in at least one of CR and ISF.
22. The method of claim 21 , further comprising generating a message to an electronic insulin delivery device to deliver insulin.
23. The method of claim 21 , wherein generating the indication includes generating a visual depiction of the probability distribution of the general CR and general ISF for the BR of the user and an indicator of a location of the first relationship within the visual depiction.
24. The method of claim 21 , wherein generating the indication includes generating a plurality of curves illustrating the general BR relative to at least one of the general CR and the general ISF and an indicator of a location relative to the plurality of curves for at least one of the CR and the ISF for the user.
25. The method of claim 19 , wherein generating the indication comprises:
generating the probability distribution of a second relationship between the general BR, the general CR, and the general ISF as a general ellipsoidal distribution having a major axis and two minor axes, an intersection of the major axis and the two minor axes defining a midpoint of the probability distribution;
determining a location represented by the first relationship within the probability distribution; and
determining a distance between the location and the midpoint of the probability distribution.
26. The method of claim 25 , wherein the probability distribution of the second relationship comprises a multivariate normal distribution of logarithms of the general BR, the general CR, and the general ISF.
27. The method of claim 25 , wherein the probability distribution of the second relationship comprises a plurality of concentric ellipsoidal regions.
28. The method of claim 27 , further comprising, responsive to determining that the location represented by the first relationship falls within an outer ellipsoidal region, determining that the first relationship exceeds the threshold distance.
29. The method of claim 28 , further comprising, generating a visual depiction of the probability distribution, the plurality of concentric ellipsoidal regions, and the location represented by the first relationship.
30. A method of delivering insulin, the method comprising:
determining a personalized model that represents a user's physiological response to insulin therapy; and
determining a difference between the personalized model and a generalized model that represents a generalized physiological response by a general population to an insulin therapy.
31. A method of delivering insulin, the method comprising:
determining a personalized model that represents a user's physiological response to insulin therapy; and
determining a difference between the personalized model and a generalized model that represents a generalized physiological response to an insulin therapy,
wherein determining the personalized model comprises:
receiving at least a daily basal rate (BR) for a user and only one of a carbohydrate-to-insulin ratio (CR) and an insulin sensitivity factor (ISF) for the user; and
determining a first relationship between the BR and at least one of the CR and the ISF for the user.
32. The method of claim 31 , wherein determining a difference between the personalized model and the generalized model comprises:
comparing the first relationship with a probability distribution of a second relationship of the generalized model between a general BR, a general CR, and a general ISF; and
determining a distance between the first relationship and a midpoint of the probability distribution of the second relationship of the generalized model.
33. The method of claim 32 , further comprising:
determining that the distance between the first relationship and the midpoint of the probability distribution exceeds a threshold distance; and
generating an indication that the first relationship exceeds the threshold distance and generating a recommended variation in at least one of CR and ISF.
34. The method of claim 33 , wherein generating the indication includes generating a visual depiction of the probability distribution of the general CR and general ISF for the BR of the user and an indicator of a location of the first relationship within the visual depiction.
35. The method of claim 34 , wherein generating the indication includes generating a plurality of curves illustrating the general BR relative to at least one of the general CR and the general ISF and an indicator of a location relative to the plurality of curves for at least one of the CR and the ISF for the user.
36. The method of claim 35 , wherein generating the indication comprises:
generating the probability distribution of the second relationship between the general BR, the general CR, and the general ISF as a general ellipsoidal distribution having a major axis and two minor axes, an intersection of the major axis and the two minor axes defining the midpoint of the probability distribution;
determining a location represented by the first relationship within the probability distribution; and
determining the distance between the location and the midpoint of the probability distribution.
37. The method of claim 36 , wherein the probability distribution of the second relationship comprises a multivariate normal distribution of logarithms of the general BR, the general CR, and the general ISF.
38. The method of claim 36 , wherein the probability distribution of the second relationship comprises a plurality of concentric ellipsoidal regions.
39. The method of claim 38 , further comprising, responsive to determining that the location represented by the first relationship falls within an outer ellipsoidal region, determining that the first relationship exceeds the threshold distance.
40. The method of claim 39 , further comprising, generating a visual depiction of the probability distribution, the plurality of concentric ellipsoidal regions, and the location represented by the first relationship.