Method and system for identifying fasting and postprandial periods based on detected events and using the same to train a mathematical model
Technologies are provided for identifying fasting and postprandial periods based on detected events and using the same to train a mathematical model. Events during a period can be detected via an event detection system. Each detected event can include a specific activity that is indicative of a physical behavior of the user. Training data and the detected events can be processed to determine timing of fasting periods, timing of postprandial periods, and timing of other periods. Training data from the fasting periods and other training data from the postprandial periods can be identified. The mathematical model of the user can be trained to identify fasting parameters of a physiological blood glucose response of the user during the fasting periods, and fasting parameters of a physiological blood glucose response of the user (differently) during postprandial periods.
1 . A method, comprising:
receiving one or more streams of training data during a period, the one or more streams comprising: therapy-related data and settings from an insulin infusion device of a user;
detecting events during the period, via an event detection system;
processing the training data and the detected events to determine: timing of fasting periods within the one or more streams of training data, timing of postprandial periods within the one or more streams of training data, and timing of other periods within the one or more streams of training data, wherein the fasting periods are associated with a fasting glucose condition, wherein the postprandial periods are periods that are associated with a non-fasting glucose condition that occurs when user has consumed food during the period, and wherein the other periods are time periods that are associated with neither the fasting glucose condition nor the non-fasting glucose condition;
identifying training data from the fasting periods within the one or more streams of training data and other training data from the postprandial periods within the one or more streams of training data, the training data including labeled events or event combinations detected by the event detection system during the fasting periods and the postprandial periods, wherein each detected event comprises: a specific activity that is indicative of a physical behavior of the user;
training a mathematical model of the user that simulates a physiological blood glucose response of the user to estimate a glucose level of the user based on the training data for the fasting periods to adapt the mathematical model of the user to determine values of fasting parameters associated with the physiological blood glucose response of the user during the fasting periods;
training the mathematical model of the user based on the other training data from the postprandial periods to further adapt the mathematical model of the user to determine the values of the fasting parameters associated with the physiological blood glucose response of the user differently during the postprandial periods; and
causing administration of insulin to the user based on an estimated glucose level of the user outputted by the mathematical model, in response to one or more detected events, being greater than a target glucose setpoint value.
2 . The method according to claim 1 further comprising:
storing, in a database, the training data for the fasting periods, the detected events as labeled events, and different combinations of detected events as labeled event combinations; and
retrieving, via a system, the training data for the fasting periods, selected ones of the labeled events for the fasting periods, and selected ones of the labeled event combinations for the fasting periods from the database,
wherein training the mathematical model of the user comprises:
training the mathematical model of the user based on the training data for the fasting periods, the selected ones of the labeled events for the fasting periods, and the selected ones of the labeled event combinations for the fasting periods to adapt the mathematical model of the user to identify the fasting parameters of the physiological blood glucose response of the user during the fasting periods in response to the selected ones of the labeled events for the fasting periods and the selected ones of the labeled event combinations for the fasting periods.
3 . The method according to claim 2 , wherein training the mathematical model of the user based on the training data for the fasting periods, the selected ones of the labeled events for the fasting periods, and the selected ones of the labeled event combinations for the fasting periods comprises:
defining one or more parameters of equations of the mathematical model based on the training data for the fasting periods, the selected ones of the labeled events for the fasting periods, the selected ones of the labeled event combinations for the fasting periods, and insulin delivery to the user to adapt the mathematical model of the user to identify the fasting parameters of the physiological blood glucose response of the user during the fasting periods, wherein each parameter that is defined based on the training data for the fasting periods is modifiable as a function of one or more of the selected ones of the labeled events for the fasting periods and the selected ones of the labeled event combinations for the fasting periods.
4 . The method according to claim 2 , wherein the therapy-related data and settings from the insulin infusion device of the user describe one or more of: blood glucose levels of the user during the period and insulin delivery to the user during the period, and wherein each labeled event and each labeled event combination comprises: a label and a probability of an increased insulin delivery demand or a decreased insulin delivery demand for that particular labeled event or that particular combination of detected events, and wherein each labeled event and each labeled event combination is correlated to:
a decreased insulin delivery demand when the probability of the decreased insulin delivery demand is greater than a first threshold;
an increased insulin delivery demand when the probability of the increased insulin delivery demand is greater than a second threshold; or
an ordinary insulin delivery demand when the probability of the decreased insulin delivery demand is less than the first threshold and the probability of the increased insulin delivery demand is less than the second threshold.
5 . The method according to claim 1 , further comprising:
storing, in a database, the other training data from the postprandial periods, the detected events as labeled events, and different combinations of detected events as labeled event combinations; and
retrieving the training data for the postprandial periods, selected ones of the labeled events for the postprandial periods, and selected ones of the labeled event combinations for the postprandial periods from the database,
wherein training the mathematical model of the user comprises:
training the mathematical model of the user based on the other training data from the postprandial periods, the selected ones of the labeled events for the postprandial periods, and the selected ones of the labeled event combinations for the postprandial periods to further adapt the mathematical model of the user to identify the fasting parameters of the physiological blood glucose response of the user differently during the postprandial periods in response to the selected ones of the labeled events for the postprandial periods and the selected ones of the labeled event combinations for the postprandial periods.
6 . The method according to claim 5 , wherein training the mathematical model of the user based on the other training data from the postprandial periods, the selected ones of the labeled events for the postprandial periods, and the selected ones of the labeled event combinations for the postprandial periods, comprises:
defining one or more parameters of equations of the mathematical model based on the training data for the postprandial periods, the selected ones of the labeled events for the postprandial periods, the selected ones of the labeled event combinations for the postprandial periods, carbohydrate intake by the user and insulin delivery to the user to further adapt the mathematical model of the user to identify the fasting parameters of the physiological blood glucose response of the user differently during the postprandial periods, wherein each parameter that is defined based on the training data for the postprandial periods is modifiable as a function of one or more of the selected ones of the labeled events for the postprandial periods and the selected ones of the labeled event combinations for the postprandial periods.
7 . The method according to claim 5 , wherein the therapy-related data and settings from the insulin infusion device of the user describe one or more of: blood glucose levels of the user during the period and insulin delivery to the user during the period, and wherein each labeled event and each labeled event combination comprises: a label and a probability of an increased insulin delivery demand or a decreased insulin delivery demand for that particular labeled event or that particular combination of detected events, and wherein each labeled event and each labeled event combination is correlated to:
a decreased insulin delivery demand when the probability of the decreased insulin delivery demand is greater than a first threshold;
an increased insulin delivery demand when the probability of the increased insulin delivery demand is greater than a second threshold; or
an ordinary insulin delivery demand when the probability of the decreased insulin delivery demand is less than the first threshold and the probability of the increased insulin delivery demand is less than the second threshold.
8 . The method according to claim 1 , wherein the specific activities that are indicative of the physical behavior of the user comprise one or more of: a meal ingestion event; a sleep event; a physical activity event; an exercise event; and a work-related event.
9 . A system, comprising:
one or more hardware-based processors configured by machine-readable instructions to:
receive one or more streams of training data during a period, the one or more streams comprising: therapy-related data and settings from an insulin infusion device of a user;
detect events during the period, via an event detection system;
process the training data and the detected events to determine: timing of fasting periods within the one or more streams of training data, timing of postprandial periods within the one or more streams of training data, and timing of other periods within the one or more streams of training data, wherein the fasting periods are periods within the one or more streams that are associated with a fasting glucose condition, wherein the postprandial periods are periods within the one or more streams that are associated with a non-fasting glucose condition that occurs when user has consumed food during the period, and wherein the other periods are time periods that are associated with neither the fasting glucose condition nor the non-fasting glucose condition;
identifying training data from the fasting periods within the one or more streams of training data and other training data from the postprandial periods within the one or more streams of training data, the training data including labeled events or event combinations detected by the event detection system during the fasting periods and the postprandial periods, wherein each detected event comprises: a specific activity that is indicative of a physical behavior of the user;
train a mathematical model that simulates a physiological blood glucose response of the user to estimate a glucose level of the user based on the training data for the fasting periods to adapt the mathematical model of the user to determine values of fasting parameters associated with the physiological blood glucose response of the user during the fasting periods;
train the mathematical model based on the other training data from the postprandial periods to further adapt the mathematical model of the user to determine the values of the fasting parameters associated with the physiological blood glucose response of the user differently during the postprandial periods; and
cause administration of insulin to the user based on an estimated glucose level of the user outputted by the mathematical model, in response to one or more detected events, being greater than a target glucose value.
10 . The system according to claim 9 , further comprising:
storage configured to store a database of: the training data for the fasting periods, the detected events as labeled events, and different combinations of detected events as labeled event combinations; and
wherein the one or more hardware-based processors are further configured by machine-readable instructions to:
retrieve the training data for the fasting periods, selected ones of the labeled events for the fasting periods, and selected ones of the labeled event combinations for the fasting periods from the database; and
train the mathematical model based on the training data for the fasting periods, the selected ones of the labeled events for the fasting periods, and the selected ones of the labeled event combinations for the fasting periods to adapt the mathematical model of the user to identify the fasting parameters of the physiological blood glucose response of the user during the fasting periods in response to the selected ones of the labeled events for the fasting periods and the selected ones of the labeled event combinations for the fasting periods.
11 . The system according to claim 10 , wherein the one or more hardware-based processors are further configured by machine-readable instructions to:
define one or more parameters of equations of the mathematical model based on the training data for the fasting periods, the selected ones of the labeled events for the fasting periods, the selected ones of the labeled event combinations for the fasting periods, and insulin delivery to the user to adapt the mathematical model of the user to identify the fasting parameters of the physiological blood glucose response of the user during the fasting periods, wherein each parameter that is defined based on the training data for the fasting periods is modifiable as a function of one or more of the selected ones of the labeled events for the fasting periods and the selected ones of the labeled event combinations for the fasting periods.
12 . The system according to claim 10 , wherein the therapy-related data and settings from the insulin infusion device of the user describe one or more of: blood glucose levels of the user during the period and insulin delivery to the user during the period, and wherein each labeled event and each labeled event combination comprises: a label and a probability of an increased insulin delivery demand or a decreased insulin delivery demand for that particular labeled event or that particular combination of detected events, and wherein each labeled event and each labeled event combination is correlated to:
a decreased insulin delivery demand when the probability of the decreased insulin delivery demand is greater than a first threshold;
an increased insulin delivery demand when the probability of the increased insulin delivery demand is greater than a second threshold; or
an ordinary insulin delivery demand when the probability of the decreased insulin delivery demand is less than the first threshold and the probability of the increased insulin delivery demand is less than the second threshold.
13 . The system according to claim 9 , further comprising:
storage configured to store a database of: the other training data from the postprandial periods, the detected events as labeled events, and different combinations of detected events as labeled event combinations; and
wherein the one or more hardware-based processors are further configured by machine-readable instructions to:
retrieve the training data for the postprandial periods, selected ones of the labeled events for the postprandial periods, and selected ones of the labeled event combinations for the postprandial periods from the database; and
train the mathematical model based on the other training data from the postprandial periods, the selected ones of the labeled events for the postprandial periods, and the selected ones of the labeled event combinations for the postprandial periods to further adapt the mathematical model of the user to identify the fasting parameters of the physiological blood glucose response of the user differently during the postprandial periods in response to the selected ones of the labeled events for the postprandial periods and the selected ones of the labeled event combinations for the postprandial periods.
14 . The system according to claim 13 , wherein the one or more hardware-based processors are further configured by machine-readable instructions to:
define one or more parameters of equations of the mathematical model based on the training data for the postprandial periods, the selected ones of the labeled events for the postprandial periods, the selected ones of the labeled event combinations for the postprandial periods, carbohydrate intake by the user and insulin delivery to the user to further adapt the mathematical model of the user to identify the fasting parameters of the physiological blood glucose response of the user differently during the postprandial periods, wherein each parameter that is defined based on the training data for the postprandial periods is modifiable as a function of one or more of the selected ones of the labeled events for the postprandial periods and the selected ones of the labeled event combinations for the postprandial periods.
15 . The system according to claim 13 , wherein the therapy-related data and settings from the insulin infusion device of the user describe one or more of: blood glucose levels of the user during the period and insulin delivery to the user during the period, and wherein each labeled event and each labeled event combination comprises: a label and a probability of an increased insulin delivery demand or a decreased insulin delivery demand for that particular labeled event or that particular combination of detected events, and wherein each labeled event and each labeled event combination is correlated to:
a decreased insulin delivery demand when the probability of the decreased insulin delivery demand is greater than a first threshold;
an increased insulin delivery demand when the probability of the increased insulin delivery demand is greater than a second threshold; or
an ordinary insulin delivery demand when the probability of the decreased insulin delivery demand is less than the first threshold and the probability of the increased insulin delivery demand is less than the second threshold.
16 . At least one non-transitory computer-readable medium having instructions stored thereon that are configurable to cause at least one processor to perform operations comprising:
receiving one or more streams of training data during a period, the one or more streams comprising: therapy-related data and settings from an insulin infusion device of a user;
detecting events during the period, via an event detection system;
processing the training data and the detected events to determine: timing of fasting periods within the one or more streams of training data, timing of postprandial periods within the one or more streams of training data, and timing of other periods within the one or more streams of training data, wherein the fasting periods are periods within the one or more streams that are associated with a fasting glucose condition, wherein the postprandial periods are periods within the one or more streams that are associated with a non-fasting glucose condition that occurs when user has consumed food during the period, and wherein the other periods are time periods that are associated with neither the fasting glucose condition nor the non-fasting glucose condition;
identifying training data from the fasting periods within the one or more streams of training data and other training data from the postprandial periods within the one or more streams of training data, the training data including labeled events or event combinations detected by the event detection system during the fasting periods and the postprandial periods, wherein each detected event comprises: a specific activity that is indicative of a physical behavior of the user;
training a mathematical model of the user that simulates a physiological blood glucose response of the user to estimate a glucose level of the user based on the training data for the fasting periods to adapt the mathematical model of the user to determine values of fasting parameters associated with the physiological blood glucose response of the user during the fasting periods;
training the mathematical model of the user that simulates a physiological blood glucose response of the user based on the other training data from the postprandial periods to further adapt the mathematical model of the user to determine the values of the fasting parameters associated with the physiological blood glucose response of the user differently during the postprandial periods; and
causing administration to the user an insulin dose determined based on a difference between an estimated glucose level of the user outputted by the mathematical model in response to one or more detected events and a target glucose setpoint value.
17 . The at least one non-transitory computer-readable medium according to claim 16 , wherein the operations further comprise:
storing, in a database, the training data for the fasting periods, the detected events as labeled events, and different combinations of detected events as labeled event combinations; and
retrieving, via a system, the training data for the fasting periods, selected ones of the labeled events for the fasting periods, and selected ones of the labeled event combinations for the fasting periods from the database,
wherein training the mathematical model of the user comprises:
training the mathematical model of the user based on the training data for the fasting periods, the selected ones of the labeled events for the fasting periods, and the selected ones of the labeled event combinations for the fasting periods to adapt the mathematical model of the user to identify the fasting parameters of the physiological blood glucose response of the user during the fasting periods in response to the selected ones of the labeled events for the fasting periods and the selected ones of the labeled event combinations for the fasting periods.
18 . The at least one non-transitory computer-readable medium according to claim 17 , wherein training the mathematical model of the user based on the training data for the fasting periods, the selected ones of the labeled events for the fasting periods, and the selected ones of the labeled event combinations for the fasting periods comprises:
defining one or more parameters of equations of the mathematical model based on the training data for the fasting periods, the selected ones of the labeled events for the fasting periods, the selected ones of the labeled event combinations for the fasting periods, and insulin delivery to the user to adapt the mathematical model of the user to identify the fasting parameters of the physiological blood glucose response of the user during the fasting periods, wherein each parameter that is defined based on the training data for the fasting periods is modifiable as a function of one or more of the selected ones of the labeled events for the fasting periods and the selected ones of the labeled event combinations for the fasting periods.
19 . The at least one non-transitory computer-readable medium according to claim 16 , wherein the operations further comprise:
storing, in a database, the other training data from the postprandial periods, the detected events as labeled events, and different combinations of detected events as labeled event combinations; and
retrieving the training data for the postprandial periods, selected ones of the labeled events for the postprandial periods, and selected ones of the labeled event combinations for the postprandial periods from the database,
wherein training the mathematical model of the user comprises:
training the mathematical model of the user based on the other training data from the postprandial periods, the selected ones of the labeled events for the postprandial periods, and the selected ones of the labeled event combinations for the postprandial periods to further adapt the mathematical model of the user to identify the fasting parameters of the physiological blood glucose response of the user differently during the postprandial periods in response to the selected ones of the labeled events for the postprandial periods and the selected ones of the labeled event combinations for the postprandial periods.
20 . The at least one non-transitory computer-readable medium according to claim 19 , wherein training the mathematical model of the user based on the other training data from the postprandial periods, the selected ones of the labeled events for the postprandial periods, and the selected ones of the labeled event combinations for the postprandial periods, comprises:
defining one or more parameters of equations of the mathematical model based on the training data for the postprandial periods, the selected ones of the labeled events for the postprandial periods, the selected ones of the labeled event combinations for the postprandial periods, carbohydrate intake by the user and insulin delivery to the user to further adapt the mathematical model of the user to identify the fasting parameters of the physiological blood glucose response of the user differently during the postprandial periods, wherein each parameter that is defined based on the training data for the postprandial periods is modifiable as a function of one or more of the selected ones of the labeled events for the postprandial periods and the selected ones of the labeled event combinations for the postprandial periods.