System, method and computer readable medium for dynamical tracking of the risk for hypoglycemia in type 1 and type 2 diabetes
A system, method and non-transient computer readable medium for tracking hypoglycemia risk in patients with diabetes exercise. A system may include a digital processor configured to execute instructions to receive an input from each available data source of a plurality of intermittently available data sources; determine a plurality of probability signals for impending hypoglycemia, wherein each probability signal is based on one or more of the inputs from the available data sources or a lack of input from an unavailable data source; wherein a probability signal for each unavailable data source is assigned a value corresponding to a zone of uncertainty; and determine an aggregate risk of hypoglycemia based on the plurality of intermittently data sources by aggregating the plurality of probability signals.
1. A method for tracking hypoglycemia risk comprising:
obtaining an input from each available data source of a plurality of intermittently available data sources;
determining, for each of the plurality of intermittently available data sources, one or more probability signals for impending hypoglycemia, wherein each probability signal for an available data source is based on one or more of the inputs therefrom and each probability signal for an unavailable data source is based on a lack of input therefrom and assigned a probability value corresponding to a zone of uncertainty of risk of hypoglycemia; and
determining an aggregate risk of hypoglycemia by aggregating the probability signals for the plurality of intermittently available data sources.
2. The method of claim 1 , wherein one data source of the plurality of intermittently available data sources comprises self-monitoring blood glucose (SMBG) data.
3. The method of claim 2 , wherein determining the one or more probability signals for impending hypoglycemia includes determining a chronic risk of hypoglycemia based on the SMBG data by the formulas
e ChronicRisk( t 0 )= f Chronic (SMBG t 0 )
e ChronicRisk( t )=β2· e ChronicRisk( t− 1)+β1· f Chronic (SMBG t )
wherein β1 and β2 are predefined constants.
4. The method of claim 2 , wherein determining the one or more probability signals for impending hypoglycemia includes determining an acute risk of hypoglycemia based on the SMBG data by the formulas
e AcuteRisk( t 0 )=α1· f Acute (SMBG t 0 )
e AcuteRisk( t )=α2· e AcuteRisk( t −1)+α1· f Acute (SMBG t )
wherein α1 and α2 are predefined constants.
5. The method of claim 2 , wherein obtaining the self-monitoring blood glucose (SMBG) data comprises receiving a blood glucose signal from a continuous blood glucose monitor.
6. The method of claim 1 , wherein the plurality of intermittently available data sources includes one or more of: a physical activity indication, an insulin delivery indication, a carbohydrate indication, and a non-insulin medicine indication.
7. The method of claim 6 , wherein the physical activity indication comprises a signal from at least one sensor configured to detect when the user begins to exercise.
8. The method of claim 1 , wherein one or more of the plurality of intermittently available data sources are automatically monitored and reported.
9. The method of claim 1 , wherein one or more of the plurality of intermittently available data sources are self-reported by a user.
10. The method of claim 1 , wherein determining the one or more probability signals for impending hypoglycemia comprises translating each input from each available data source into a respective probability signal for impending hypoglycemia.
11. The method of claim 1 , wherein the probability signal for impending hypoglycemia is standardized on a scale where minimal risk of hypoglycemia is mapped to zero, maximal risk of hypoglycemia is mapped to 1, a cutoff value differentiating no-risk and elevated risk is mapped to 0.5, and the zone of uncertainty in determining risk of hypoglycemia is mapped to 0.5.
12. The method of claim 1 further comprising: using the aggregate risk of hypoglycemia to estimate the probability of a hypoglycemic event.
13. The method of claim 1 , wherein aggregating the probability signals includes combining the probability signals using the Bayes formula.
14. The method of claim 1 , wherein aggregating the probability signals includes the steps of:
determining an individual's chronic risk of hypoglycemia based on self-monitored blood glucose data by the formula
P 1 hypo =P 1 ( e ChronicRisk)
if a probability signal for acute risk of hypoglycemia based on self-monitored blood glucose data is available, updating the aggregate risk of hypoglycemia by the formula
P
hypo
2
=
P
hypo
1
·
P
2
(
eAcuteRisk
)
P
hypo
1
·
P
2
(
eAcuteRisk
)
+
(
1
-
P
hypo
1
)
·
(
1
-
P
2
(
eAcuteRisk
)
)
and, for each additional probability signal, updating the aggregate risk of hypoglycemia by the formula
P
hypo
3
=
P
hypo
2
·
P
3
(
Exercise
)
P
hypo
2
·
P
3
(
Exercise
)
+
(
1
-
P
hypo
2
)
·
(
1
-
P
3
(
Exercise
)
)
,
where “Exercise” indicates one of the data sources of the additional probability signal.
15. The method of claim 1 further comprising: displaying an alert on a display of a portable computing device based on the determined aggregated risk of hypoglycemia.
16. The method of claim 1 further comprising: communicating an instruction to an insulin pump based on the determined aggregated risk of hypoglycemia.
17. A system for tracking hypoglycemia risk comprising:
a digital processor;
a memory in communication with the digital processor, wherein the memory contains instructions configured to be executed by the processor to
receive an input from each available data source of a plurality of intermittently available data sources;
determine, for each of the plurality of intermittently available data sources, one or more probability signals for impending hypoglycemia, wherein each probability signal for an available data source is based on one or more of the inputs therefrom and each probability signal for an unavailable data source is based on a lack of input therefrom and assigned a probability value corresponding to a zone of uncertainty of risk of hypoglycemia; and
determine an aggregate risk of hypoglycemia by aggregating the probability signals for the plurality of intermittently available data sources.
18. The system of claim 17 further comprising:
a display; and
wherein the digital processor is configured to generate an alert on the display if the determined aggregate risk of hypoglycemia indicates a probability of a hypoglycemic event exceeds a predetermined threshold.
19. The system of claim 17 further comprising:
a continuous blood glucose monitoring sensor in communication with the digital processor, the continuous blood glucose monitoring sensor configured to generate self-monitored blood glucose data and communicate said data to the digital processor.
20. The system of claim 17 further comprising:
an insulin pump in communication with the digital processor and configured to dispense or not dispense insulin in response to the determined aggregate risk of hypoglycemia.