CN117546251A - Event-oriented blood glucose response prediction - Google Patents

Event-oriented blood glucose response prediction Download PDF

Info

Publication number
CN117546251A
CN117546251A CN202280041824.6A CN202280041824A CN117546251A CN 117546251 A CN117546251 A CN 117546251A CN 202280041824 A CN202280041824 A CN 202280041824A CN 117546251 A CN117546251 A CN 117546251A
Authority
CN
China
Prior art keywords
event
person
glucose
physiological
insulin
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202280041824.6A
Other languages
Chinese (zh)
Inventor
A·米赫诺
钟宇翔
D·Y·康
M·P·斯通
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Medtronic Minimed Inc
Original Assignee
Medtronic Minimed Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from US17/852,878 external-priority patent/US20230000447A1/en
Application filed by Medtronic Minimed Inc filed Critical Medtronic Minimed Inc
Publication of CN117546251A publication Critical patent/CN117546251A/en
Pending legal-status Critical Current

Links

Landscapes

  • Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)

Abstract

Disclosed herein are techniques related to event-guided prediction of glycemic response. In some embodiments, the techniques may involve obtaining a predictive model that correlates a person's glycemic response to events with a physiological parameter of the person during the events. These techniques may also involve obtaining glucose level measurements of the person during the event. Additionally, the techniques may involve determining a physiological parameter of the person during the event based on the glucose level measurement. Furthermore, the techniques may involve predicting a glycemic response of the person to the event based on applying the predictive model to the physiological parameter.

Description

Event-oriented blood glucose response prediction
Cross Reference to Related Applications
This application claims the benefit and priority of the following applications: U.S. provisional application 63/216,865 filed on day 6 and 30 of 2021, U.S. provisional application 63/304,090 filed on day 1 and 28 of 2022, and U.S. patent application 17/852,878 filed on day 29 of 2022 and 6. The entire contents of each of the foregoing applications are hereby incorporated by reference.
Technical Field
The present disclosure relates generally to the medical arts, and more particularly to event-guided blood glucose response prediction.
Background
Humans may use insulin therapy to control type I or type II diabetes. Insulin therapy may include the use of an insulin infusion system to deliver or dispense insulin. Insulin infusion systems may include an infusion device that typically includes a small motor and drive train components configured to deliver insulin from a reservoir into a person's body, for example, via a percutaneous needle or cannula placed in subcutaneous tissue. Insulin infusion systems may be beneficial for diabetes control in certain individuals.
Disclosure of Invention
Disclosed herein are techniques related to event-guided prediction of glycemic response. These techniques may be practiced in a variety of ways, such as using: a processor-implemented method; a system comprising one or more processors and one or more processor-readable media; and/or one or more (non-transitory) processor-readable media.
In accordance with aspects of the present disclosure, these techniques may involve obtaining a predictive model that correlates a person's glycemic response to events with a person's physiological parameters during those events. These techniques may also involve obtaining glucose level measurements of a person during an event. Additionally, the techniques may involve determining physiological parameters of the person during the event based on glucose level measurements. Furthermore, the techniques may involve predicting a person's glycemic response to an event based on applying a predictive model to a physiological parameter.
Drawings
The foregoing and other aspects and features of the present disclosure will become more apparent in view of the following detailed description when taken in conjunction with the accompanying drawings in which like reference numerals identify like elements.
Fig. 1 is a diagram of an exemplary therapy delivery system according to aspects of the present disclosure;
FIG. 2 is a diagram of a plurality of exemplary event data, according to aspects of the present disclosure;
FIG. 3 is a diagram of exemplary event data content in accordance with aspects of the present disclosure;
FIG. 4 is a block diagram of an exemplary technique for blood glucose response prediction in accordance with aspects of the present disclosure;
FIG. 5 is a diagram of an exemplary physiological model in accordance with aspects of the present disclosure;
FIG. 6 is an illustration of an exemplary technique for determining a physiological state of a person during an event in accordance with aspects of the present disclosure;
FIG. 7 is an illustration of an exemplary technique for identifying similar events in accordance with aspects of the present disclosure;
FIG. 8 is a diagram of a table showing exemplary similarity determination results for events, in accordance with aspects of the present disclosure;
FIG. 9 is a chart illustrating an exemplary grouping of similar events, in accordance with aspects of the present disclosure;
FIG. 10 is a diagram of an exemplary predictive model in accordance with aspects of the present disclosure;
11A-11B are illustrations of exemplary techniques for deriving a glucose response curve based on the output of a predictive model, in accordance with aspects of the present disclosure; and is also provided with
Fig. 12 is an illustration of an exemplary insulin delivery device according to aspects of the present disclosure.
Detailed Description
In terms of diabetes treatment, a great deal of effort has been expended to develop closed-loop insulin delivery systems. In particular, there is interest in developing a closed loop system that can predict the glycemic response to various events. As used herein, an event may refer to a meal event, an exercise event, a sleep event, a disease event, or any other event that may affect glucose levels. The term "glycemic response" refers to a change in glucose level caused by an event and may be expressed as a series of glucose levels over time.
Some methods for predicting glycemic response involve: glucose dynamics are modeled based on glucose levels measured prior to an event. For example, the glycemic response to a meal event may be predicted based on: extrapolated results from glucose levels measured prior to meal intake. However, predictions made in such a manner tend to lose accuracy as events progress. This is due, at least in part, to variability between events. For example, due to differences in meal characteristics (e.g., carbohydrate intake may be different at lunch versus dinner) and physiological environments (e.g., metabolism of a person may be different at lunch versus dinner), the glycemic response to a lunch meal may be different than the glycemic response to a dinner meal. Thus, modeling the same glucose dynamics for all events may result in the prediction of glycemic response becoming less and less accurate over time.
Thus, disclosed herein are event-directed methods for predicting glycemic response. Such methods enable prediction of glycemic response, which accounts for variability between events. For example, such methods may take event characteristics (e.g., carbohydrate content of a meal) and physiological environment (e.g., metabolic rate of a person while eating) into account when predicting a glycemic response. Thus, the technology disclosed herein provides a robust method that achieves an increase in the accuracy of the blood glucose response prediction even a few hours after the start of the event.
The present disclosure is described primarily with respect to insulin delivery systems. Aspects and embodiments of the present disclosure may be practiced with one or more types of insulin (e.g., fast acting insulin, medium acting insulin, and/or slow acting insulin). Unless the context indicates otherwise, terms such as "dose", "insulin", "basal" and "bolus" may not denote a particular type of insulin. For example, fast acting insulin may be used for both basal and bolus doses. As used herein, the term "basal" refers to and includes insulin delivered in amounts and frequencies intended to correspond to healthy body between meals and during sleep. The term "bolus" refers to and includes insulin delivered in amounts and at timings intended to correspond to a healthy body releasing insulin to counteract high glucose levels such as caused by consumption of food and drink. Catering may include any type or amount of food or beverage consumption, including breakfast, lunch, dinner, snack, beverage, and the like.
Although the present disclosure may be described primarily with respect to insulin delivery systems, the scope of the present disclosure is not limited to insulin delivery systems. Rather, the present disclosure is equally applicable to and may be practiced with respect to other therapy systems. Any aspects, embodiments and descriptions intended to be relevant to insulin delivery systems should be equally applicable to other types of therapy delivery systems.
Although the disclosure is not limited in this respect, discussions utilizing terms such as, for example, "processing," "computing," "determining," "establishing", "analyzing", "checking", or the like, may refer to the operation(s) and/or process (es) of a computer, computing platform, computing system, or other electronic computing device, that manipulate and/or transform data represented as physical (e.g., electronic) quantities within the computer's registers and/or memories into other data similarly represented as physical quantities within the computer's registers and/or memories or other non-transitory information storage medium that may store instructions to perform the operations and/or processes. As used herein, "exemplary" does not necessarily mean "preferred" and may simply refer to an example unless the context clearly indicates otherwise. Although the present disclosure is not limited in this respect, the terms "plurality" and "plurality" as used herein may include, for example, "multiple" or "two or more". The terms "plurality" or "plurality" may be used throughout the specification to describe two or more components, devices, elements, units, parameters, or the like. The term "set" as used herein may include one or more items. Unless explicitly stated, the methods described herein are not limited to a particular order or sequence. In addition, some of the methods or elements of the methods may occur or be performed concurrently or in parallel.
Referring to fig. 1, a diagram of an exemplary therapy delivery system 100 for a person 101 is shown. The system 100 may be an insulin delivery system. The illustrated therapy delivery system 100 includes a delivery device 102, a monitoring device 104, a computing device 106, and optionally a remote or cloud computing system 108. The delivery device 102, monitoring device 104, and computing device 106 may be embodied in various ways, including being disposed in one or more device housings. For example, in various embodiments, all of the devices 102-106 may be provided in a single device housing. In various embodiments, each of the devices 102-106 may be provided in a separate device housing. In various embodiments, two or more of the devices 102-106 may be disposed in the same device housing, and/or a single device 102, 104, or 106 may have two or more portions disposed in two or more housings. Such embodiments and combinations thereof are contemplated as being within the scope of the present disclosure.
Fig. 1 also shows communication links 112-118. The communication links 112-118 may each be a wired connection and/or a wireless connection. In the case where the two devices are located in the same device housing, the communication link may comprise, for example, wires, cables, and/or a communication bus located on a printed circuit board, etc. In the case where two devices are separated from each other in different device housings, the communication link may be a wired connection and/or a wireless connection. The wired connection may include, but is not limited to, an ethernet connection, a USB connection, and/or another type of physical connection. The wireless connection may include, but is not limited to, a cellular connection, a Wi-Fi connection, A connection, a mesh network connection, and/or another type of connection using a wireless communication protocol. Various embodiments of communication links 112-118 may use direct connections such as +.>Connections, and/or connections routed through one or more networks or network devices (not shown) may be used, such as an ethernet network, a Wi-Fi network, a cellular network, a satellite network, an intranet, an extranet, the internet, and/or other types of networks such as the internet backbone. Various combinations of wired and/or wireless connections may be used for communication links 112-118.
Aspects of insulin delivery system 100 are described below. Additional aspects and details can be described in the following U.S. patent nos.: 4,562,751;4,685,903;5,080,653;5,505,709;5,097,122;6,485,465;6,554,798;6,558,320;6,558,351;6,641,533;6,659,980;6,752,787;6,817,990;6,932,584; and 7,621,893. The entire contents of each of the aforementioned U.S. patents are hereby incorporated by reference.
The delivery device 102 is configured to deliver a therapeutic substance (e.g., insulin) to the person 101. The delivery device 102 may be secured to the person 101 (e.g., to the body or clothing of the person 101) or may be implantable on or in the body of the person 101. In various embodiments, the delivery device 102 may include a reservoir, an actuator, a delivery mechanism, and a cannula (not shown). The reservoir may be configured to store a quantity of the therapeutic substance. In various embodiments, the reservoir may be refillable or replaceable. The actuator may be configured to drive the delivery mechanism. In some examples, the actuator may include a motor, such as an electric motor. The delivery mechanism may be configured to move the therapeutic substance from the reservoir through the cannula. In some examples, the delivery mechanism may include a pump and/or a plunger. The cannula may facilitate a fluid connection between the reservoir and the body of the person 101. The cannula and/or needle may facilitate delivery of the therapeutic substance to a tissue layer, vein, or body cavity of the person 101. During operation, the actuator may drive the delivery mechanism in response to a signal (e.g., a command signal) to move the therapeutic substance from the reservoir through the cannula and into the body of the person 101.
The components of the delivery device 102 described above are exemplary. The delivery device 102 may include other components such as, but not limited to, a power source, a communication transceiver, computing resources, and/or a user interface, among others. Those skilled in the art will recognize various implementations of delivery device 102 and components of such implementations. All such implementations and components are contemplated as being within the scope of the present disclosure.
With continued reference to fig. 1, the monitoring device 104 is configured to detect a physiological condition (e.g., glucose concentration level) of the person 101, and may also be configured to detect other events. The monitoring device 104 may be secured to the body of the person 101 (e.g., secured to the skin of the person 101 via an adhesive) and/or may be at least partially implanted in the body of the person 101. Depending on the particular location or configuration, the monitoring device 104 may be in contact with biological matter (e.g., interstitial fluid and/or blood) of the person 101.
The monitoring device 104 includes one or more sensors (not shown), such as, but not limited to, electrochemical sensors, electrical sensors, and/or optical sensors. As will be appreciated by those of skill in the art, the electrochemical sensor may be configured to respond to the interaction or binding of the biomarker with the substrate by generating an electrical signal based on the electrical potential, conductance, and/or impedance of the substrate. The substrate may comprise a material selected to interact with a specific biomarker such as glucose. The potential, conductance, and/or impedance may be proportional to the concentration of a particular biomarker. In the case of an electrical sensor, and as will be appreciated by those skilled in the art, the electrical sensor may be configured to respond to the electrical biological signal by generating an electrical signal based on the amplitude, frequency, and/or phase of the electrical biological signal. The electrical biological signal may comprise a change in current produced by the sum of potential differences across tissue, such as the nervous system, of the person 101. In various embodiments, the electrical biological signal may include portions generated by the heart of the person 101 over time, for example, as a potential change in an electrocardiogram recorded as indicative of the glucose level of the person 101. In the case of an optical sensor, the optical sensor may be configured to respond to the interaction or binding of the biomarker with the substrate by generating an electrical signal based on a change in brightness of the substrate, as will be appreciated by those skilled in the art. For example, the substrate may include a material selected to fluoresce in response to contact with a selected biomarker, such as glucose. Fluorescence may be proportional to the concentration of the selected biomarker.
In various embodiments, the monitoring device 104 may include other types of sensors that may be worn by, carried by, or coupled to the person 101 to measure activities of the person 101 that may affect the glucose level or glycemic response of the person 101. For example, the sensor may comprise an acceleration sensor configured to detect acceleration of the person 101 or a part of the person 101, such as a hand or foot of the person. Acceleration (or lack of acceleration) may be indicative of exercise, sleep, or food/beverage consumption activities of person 101, which may affect the glycemic response of person 101. In various embodiments, the sensor may include a heart rate and/or a body temperature, which may be indicative of an amount of physical exercise experienced by the person 101. In various embodiments, the sensor may include a GPS receiver that detects GPS signals to determine the location of the person 101.
The above-described sensors are exemplary. Other sensors or other types of sensors for monitoring physiological conditions, activity, and/or position, etc. will be recognized by those skilled in the art and are contemplated as within the scope of the present disclosure. For any sensor, the signal provided by the sensor shall be referred to as a "sensor signal".
The monitoring device 104 may include components and/or circuitry configured to pre-process the sensor signals. Preprocessing may include, but is not limited to, amplification, filtering, attenuation, scaling, isolation, normalization, transformation, sampling, and/or analog-to-digital conversion, etc. Those skilled in the art will recognize various implementations of such preprocessing, including but not limited to implementations using processors, controllers, ASICs, integrated circuits, hardware, firmware, programmable logic devices, and/or machine-executable instructions, etc. The type of preprocessing and its implementation are exemplary. Other types of pre-processing and implementations are contemplated as falling within the scope of the present disclosure. In various embodiments, the monitoring device 104 may not perform preprocessing.
As used herein, the term "sensed data" shall mean and include information represented by sensor signals or by preprocessed sensor signals. In various embodiments, the sensed data may include a glucose level within the person 101, an acceleration of a portion of the person 101, a heart rate of the person 101, a temperature of the person 101, and/or a geographic location (e.g., GPS location) of the person 101, among others. The monitoring device 104 may communicate the sensed data to the delivery device 102 via the communication link 112 and/or to the computing device 106 via the communication link 114. The use of sensed data by delivery device 102 and/or computing device 106 will be described later herein.
The computing device 106 provides processing power and may be implemented in various ways. In various embodiments, the computing device 106 may be a consumer device such as a smart phone, a computerized wearable device (e.g., a smart watch), a tablet computer, a laptop computer, or a desktop computer, etc., or may be a dedicated device (e.g., a portable control device) provided by, for example, the manufacturer of the delivery device 102. In various embodiments, the computing device 106 may be a "processing circuit" (defined below) integrated with another device, such as the delivery device 102. In various implementations, the computing device 106 may be fixed to the person 101 (e.g., fixed to the body or clothing of the person 101), may be at least partially implanted in the body of the person 101, and/or may be held by the person 101.
For each of the embodiments of computing device 106, computing device 106 may include various types of logic circuits including, but not limited to, microprocessors, controllers, digital Signal Processors (DSPs), application Specific Integrated Circuits (ASICs), field Programmable Gate Arrays (FPGAs), central Processing Units (CPUs), graphics Processing Units (GPUs), programmable logic devices, memories (e.g., random access memories, volatile memories, non-volatile memories, etc.), or other discrete or integrated logic circuits, as well as combinations of such components. The term "processing circuitry" may generally refer to any of the foregoing logic circuitry alone or in combination with other logic circuitry, or any other circuitry for performing calculations.
Aspects of the delivery device 102, the monitoring device 104, and the computing device 106 have been described above. One or more of the devices 102-106 may include a user interface (not shown) that presents information to the person 101 and/or receives information from the person 101. The user interface may include a Graphical User Interface (GUI), a display device, a keyboard, a touch screen, a speaker, a microphone, a vibration motor, buttons, switches, and/or other types of user interfaces. Those skilled in the art will recognize various types of user interfaces that may be used, and all such user interfaces are contemplated as falling within the scope of the present disclosure. For example, where the computing device 106 is a consumer device such as a smart phone, tablet computer, laptop computer, etc., the user interface will include a display device, physical and/or virtual keyboard, and/or audio speakers, etc., provided by such a consumer device. In various embodiments, the user interface may notify the person 101 of sensed data (e.g., glucose levels) and/or insulin delivery data (e.g., historical, current, or future insulin delivery rates) and may present an alert to the person 101. In various embodiments, the user interface may receive inputs from person 101, which may include, for example, requested insulin delivery changes and/or dining indications, etc. The above description and embodiments of the user interface are exemplary and other types and other uses of the user interface are contemplated as falling within the scope of the present disclosure.
The communication between the devices 102-106 and the cooperation between the devices 102-106 are described below with respect to insulin delivery. As shown in FIG. 1, and as described above, devices 102-106 may communicate with each other via communication links 112-116. In various embodiments, the computing device 106 may control the operation of the delivery device 102 and/or the monitoring device 104. For example, computing device 106 may generate one or more signals (e.g., command signals) that cause delivery device 102 to deliver insulin to person 101, e.g., at a basal dose and/or bolus dose. In various embodiments, the computing device 106 may receive data associated with insulin delivery (e.g., insulin delivery data) from the delivery device 102 and/or sensed data (e.g., glucose level) from the monitoring device 104, and may perform calculations to control the delivery device 102 based on the insulin delivery data, sensed data, and/or other data. Insulin delivery data may include, but is not limited to, the type of insulin being delivered, the historical insulin delivery rate and/or amount, the current insulin delivery rate and/or amount, and/or user input affecting insulin delivery. As will be appreciated by those of skill in the art, in the closed loop mode of operation, the computing device 106 may transmit a dose command to the delivery device 102 based on a difference between a current glucose level in the body of the person 101 (e.g., received from the monitoring device 104) and a target glucose level (e.g., determined by the computing device 106). The dose command may indicate an amount of insulin to be delivered and/or an insulin delivery rate, and may adjust the current glucose level toward a target glucose level. Examples of closed loop operation for insulin infusion systems are described in the following U.S. patent numbers: 6,088,608, 6,119,028, 6,589,229, 6,740,072, 6,827,702, 7,323,142 and 7,402,153, described in U.S. patent application publication nos. 2014/0066887 and 2014/0066889. The entire contents of each of the foregoing patents and publications are hereby incorporated by reference.
With continued reference to fig. 1, the remote or cloud computing system 108 may be a proprietary remote/cloud computing system or a commercial cloud computing system that includes one or more server computing devices. When the computing resources of a client computing device (e.g., computing device 106) are insufficient, the remote/cloud computing system 108 may provide additional computing resources as needed. The computing device 106 and the remote/cloud computing system 108 may communicate with each other via a communication link 118, which may traverse one or more communication networks (not shown). The communication network may include, but is not limited to, an ethernet network, a Wi-Fi network, a cellular network, a satellite network, an intranet, an extranet, the internet, and/or an internet backbone, among other types of networks. Those skilled in the art will recognize the specific implementation of remote/cloud computing system 108 and how to interact with such systems over various types of networks. For example, the remote/cloud computing system 108 may include an array of processing circuits (defined above) and may execute machine-readable instructions. Such implementations, interfaces, and networks are contemplated as within the scope of the present disclosure.
Thus, an exemplary therapy delivery system has been described above. For convenience, the following description may refer primarily to insulin delivery systems as examples of therapy delivery systems. However, any aspect, embodiment or description intended to be relevant to an insulin delivery system should be applicable to a therapy delivery system that delivers therapies other than insulin.
Aspects of the present disclosure relate to event-directed prediction of glycemic response. These events may be specific to one person rather than a group of people. In various embodiments, events may have various degrees of specificity. For example, meal events may include breakfast, lunch, dinner, snack and/or beverage meal events, other levels of specificity. Sleep events may include, for example, a doze event and/or an overnight sleep event, and so forth. The exercise events may include, for example, weight training, walking, jogging, and/or swimming exercise events, and the like. Disease events may include, for example, fever, vomiting, and/or drug ingestion, among others. Such events are exemplary, and other events and/or other levels of specificity for an event are considered to be within the scope of the present disclosure.
According to aspects of the present disclosure, information related to an event may be stored as event data. Referring to fig. 2, a plurality of event data 200 for a plurality of events is shown. The plurality of event data 200 may be stored in a database, such as a relational database or a NoSQL database. The plurality of event data 200 may span any time period, such as days to months to years, and other time periods, and may span different types of events, such as meal events, exercise events, sleep events, and/or disease events, and so forth. In various embodiments, the plurality of event data 200 may be stored by a computing device (such as computing device 106 of fig. 1) and/or by a remote or cloud computing system (such as remote or cloud computing system 108 of fig. 1).
Fig. 3 is an illustration of exemplary content of event data. Exemplary event data includes event characteristic data 310, therapy data 320, and physiological measurement data 330. Optionally, the event data may include physiological state data 340. It should be understood that in various embodiments, the event data may include different content, and that variations of fig. 3 are considered to be within the scope of the present disclosure. For example, in some embodiments, the treatment data 320 may be excluded from the event data and/or additional data may be included in the event data.
Event characteristic data 310 may include data corresponding to one or more variable characteristics that may be used to distinguish events. Non-limiting examples of event characteristic data 310 include meal content information (e.g., carbohydrate content information of a meal), meal type classifications (e.g., breakfast, snack, etc.), exercise activity classifications (e.g., aerobic, anaerobic, jogging, swimming, etc.), activity level information (e.g., accelerometer data and/or heart rate data), and/or disease-related information (e.g., body temperature), among others. Event characteristic data 310 may be obtained based on user input and/or data determined by the device. For example, the computing device 106 may obtain a carbohydrate count provided via a user interface and/or automatically estimated based on time of day and similar meals.
The therapy data 320 may include data related to drug delivery, such as by the delivery device 102 of fig. 1. For example, the therapy data 320 may include the amount and timing of insulin actually delivered to the person by the delivery device in relation to (e.g., during) the event.
The physiological measurement data 330 may include one or more measurements of a physiological characteristic. The physiological measurement data 330 may include, for example, glucose level measurements of a person obtained from one or more monitoring devices (e.g., 104 of fig. 1) during an event.
Event characteristic data 310, therapy data 320, and/or physiological measurement data 330 for a person (e.g., as event data for a person) may be collected and/or stored in various ways and in association with an event. In various embodiments, this may be accomplished based on a manual indication by a user, such as by the user manually intervening in a user interface of any of the devices 102-106 of fig. 1 to indicate the start and/or end of an event. In various embodiments, the event characteristic data 310, the therapy data 320, and/or the physiological measurement data 330 for a person may be automatically collected and/or stored as event data without human intervention. For example, any one of the devices 102-106, or some combination thereof, may automatically detect an ongoing event and may collect and store data 310-330 based on the automatic detection. Additionally or alternatively, one or more of the devices 102-106 may automatically collect and/or store any of the data 310-330 during post-processing of the event (e.g., processing after the event occurs). For example, the data 310-330 of the devices 102-106 may be transmitted to and stored by a remote or cloud computing system 108, which may process the data 310-330 to detect the beginning and ending of an event and may differentiate the data 310-330 according to the corresponding event. Those skilled in the art will recognize various techniques for automatically detecting certain events, such as using accelerometer data to determine various activities. Such and other techniques may also be used to process and differentiate data according to corresponding events. In various embodiments, some combination of manual indication and automatic processing may be used to collect and store event data associated with the corresponding event. Such and other embodiments are contemplated as being within the scope of the present disclosure.
The physiological state data 340 may represent at least a portion of a physiological state of a person during an event. For example, a physiological model of a person may include a set of equations with various parameters, which may vary from person to person and event to event. One or more of these parameters may be derived and stored as physiological state data 340 for the event. Non-limiting examples of physiological state data 340 include metabolic parameters corresponding to the rate of absorption of carbohydrates into the body (denoted as m 1 ) And the rate of conversion of carbohydrates to glucose (expressed as k m ). Notably, parameter m for a particular person 1 And k m The value of (a) may vary over time (e.g., may fluctuate throughout the day).
For example, based on direct measurements using sensors or other devices, it may be difficult to monitor a person's physiological state during an event. Thus, the physiological state data 340 may be derived during post-event processing and included in the event data. In various embodiments, physiological state data 340 may be derived based on a physiological model that may be used to simulate the physiological and glycemic response of a person for whom event characteristic data 310, therapy data 320, and physiological measurement data 330 are collected. As will be described in more detail in connection with fig. 6, physiological models may be used along with event characteristic data 310, therapy data 320, and/or physiological measurement data 330 to derive a physiological state of a person.
FIG. 4 illustrates a flow chart of an exemplary technique for blood glucose response prediction. The technique of fig. 4 may be performed by a remote or cloud computing system (e.g., 108 of fig. 1). In various embodiments, if a different computing system (e.g., computing system 106 of fig. 1) has sufficient computing resources, some or all of the techniques of fig. 4 may be performed on such a computing system.
An overview of the blocks of fig. 4 is provided below. Blocks 410-430 relate to the generation of a person-specific predictive model that correlates a person's glycemic response to an event with a physiological parameter of the person during the event. Blocks 440-460 relate to the application of the predictive model. The following figures describe these blocks in more detail.
At block 410, event data for events of a person is acquired. The event data may be historical data for events occurring in the past. In various embodiments, the acquired event data may be filtered for a particular type of event, such as event data for meal-only events, exercise-only events, sleep-only events, or illness-only events, or the like. In various embodiments, the acquired event data may be filtered for specific groupings of events, such as breakfast events, lunch events, dinner events, snack events, drink events, snooze events, overnight sleep events, jogging events, swimming events, fever or vomiting events, and so forth. The type or grouping of events may be specified by user input or may be automatically determined by one or more devices (e.g., 102-108 of fig. 1) as a type or grouping required for a certain glycemic response prediction model.
At block 420, similar events for the person are identified. As used herein, events are identified as "similar" if their associated glycemic responses are similar (e.g., substantially identical). However, it may be difficult to identify similar events due to variability in the physiological state and/or variability in the therapy provided in connection with the event (e.g., the amount of insulin delivered in connection with a meal event may vary). Such variability can lead to significantly different glycemic responses, even for otherwise identical events occurring at different times of the day. Thus, to facilitate identification of similar events, blood glucose responses may be normalized for comparison (e.g., by using the same physiological state data and/or the same therapeutic data).
For example, ronad lunch and dinner may eat a cheese hamburger. The actual glycemic response of ronard may be graphically represented as a curve generated based on plotting glucose levels (e.g., stored as physiological measurement data 330) as a function of time. Although the meal content is the same for both meals, the actual blood glucose response curves for both meals may have significantly different shapes. This may be due to the different metabolic rates of lunch versus dinner and/or the different amounts of insulin delivered by lunch versus dinner. To account for such differences, based on the metabolic rate of alternative ronad at lunch, the metabolic rate of ronad at lunch can be derived (e.g., using techniques that will be described in more detail in connection with fig. 6) and used to generate a hypothetical blood glucose response curve for dinner (e.g., using techniques that will be described in more detail in connection with fig. 7). Additionally or alternatively, the amount of insulin delivered for lunch may be used to generate a hypothetical blood glucose response curve for dinner. Thus, if the actual blood glucose response curve for a lunch is similar to the hypothetical blood glucose response curve for a dinner, the two meals may be identified as similar.
Various techniques may be used to make comparisons of glycemic responses. In some embodiments, a physiological model of a person may be used to simulate a blood glucose response curve for comparison. Additionally or alternatively, comparing glycemic responses may involve calculating Mean Absolute Relative Differences (MARD) of corresponding glucose levels at different time points. Additionally or alternatively, comparing blood glucose responses may involve calculating a sum of absolute values of differences in corresponding glucose levels at different time points. Additionally or alternatively, comparing blood glucose responses may involve applying a weighting function to differences in corresponding glucose levels at different time points. For example, the weighting function may penalize differences that exceed a predetermined threshold. Those skilled in the art will recognize various techniques for comparing and/or weighting differences in blood glucose responses and all such techniques are considered to be within the scope of the present disclosure.
At block 430, a predictive model is developed based on the events determined to be similar events at block 420. As described above, similar events result in similar glycemic responses if variability between events is taken into account (e.g., by normalizing glycemic responses using the same physiological status data and/or therapy data). Based on this, similar events can be analyzed to determine how sources of variability (e.g., physiological status, treatment, and/or event characteristics) contribute to differences in glycemic response. As will be explained in more detail in connection with fig. 9 and 10, machine learning and/or statistical techniques may be used to develop predictive models that take into account sources of variability. Such a predictive model achieves higher accuracy in the prediction of glycemic response because it is robust enough to account for variability between events.
One non-limiting example of such a predictive model is a set of one or more equations that may be used to predict a person's glycemic response to a meal event. The predictive model can be used for determining the metabolic rate m of a person during dining 1 And k m As input, an estimated amount of carbohydrates contained in the meal, and/or an amount of time spent since the beginning of the meal. The predictive model may provide as outputs: information indicating the person's glucose level caused by a meal (e.g., information indicating the cumulative total of glucose that is predicted to have occurred in the person's blood after the amount of time has elapsed since the beginning of the meal). This information may be used to generate a blood glucose response curve representing the person's blood glucose response to a particular meal event (e.g., a curve representing the person's glucose level as a function of time after the start of a particular meal eventA wire).
As described above, blocks 440-460 relate to the application of a predictive model. At block 440, event data for a new event (e.g., an ongoing event that the person is experiencing) is obtained. The event data may include event characteristic data, therapy data, and/or physiological measurement data. For example, if the new event is eating, the event data may include an estimated amount of carbohydrates in the meal, an amount of insulin delivered to the person, and/or a glucose level measured by a glucose sensor.
At block 450, physiological state data of the person may be determined or acquired upon occurrence of a new event (e.g., 340 of fig. 3). In an exemplary context of a dining event, the physiological state data may include one or more metabolic rates (e.g., m 1 And k m ) For a person, they may fluctuate throughout the day. Based on applying the physiological model of the person to the event data obtained for the new event, physiological state data may be determined (e.g., estimated). For example, a physiological model of a person may generally output an estimated glucose level based on inputs including event characteristic data and based on physiological state data of the person at the time of the event. Thus, by providing physiological measurement data and event characteristic data as inputs to a physiological model of a person, physiological state data of the person may be back calculated.
In various embodiments, the physiological state data may be determined based on physiological state data of other events. For example, rate m 1 Sum rate k m The value of (2) may be based on the time of day and the rate m of other events based on the time of day 1 Sum rate k m To estimate. Additionally or alternatively, the rate m 1 Sum rate k m The value of (2) may be based on the rate m of the previous event 1 Sum rate k m Is estimated by the value of (a). Other types of physiological parameters may be estimated in other ways as will be appreciated by those skilled in the art.
At block 460, a person's glycemic response to the new event is predicted based on applying the predictive model to the physiological state data determined or acquired at block 450. The physiological state data may be provided as input to a predictive model, which may output information indicative of a person's glycemic response to a new event. For example, if the new event is eating, the predictive model may output a value corresponding to the cumulative total of glucose that is predicted to have occurred in the human blood after a certain amount of time has elapsed since the beginning of the meal. This information may be used to determine a point along the glucose response curve that corresponds to a particular amount of time that has elapsed. Other predictive models may be used to determine other points along the glucose response curve (each model outputting a value for a different amount of time that has elapsed since the beginning of a meal). The prediction of a person's glycemic response to a new event will be further described in connection with fig. 11.
The techniques of fig. 4 are exemplary and variations are considered to be within the scope of the disclosure. For example, the variation may exclude one or more blocks (e.g., blocks 410-430) and/or include one or more additional blocks (e.g., blocks corresponding to determining an amount of insulin to be delivered to the person based on a predicted glycemic response).
Fig. 5 shows a block diagram of an exemplary physiological model 500 for a person. As described above, the physiological model 500 may be used to determine a physiological state of a person at the time of an event (as will be further discussed in connection with fig. 6) and/or to identify similar events (as will be further discussed in connection with fig. 7). The instructions related to the physiological model 500 may be implemented in various programming languages and may be executed on a remote or cloud computing system (e.g., 108 of fig. 1). In various embodiments, if a different computing system (such as computing system 106 of fig. 1) has sufficient computing resources, some or all of the instructions related to physiological model 500 may be executed on such a computing system.
The physiological model 500 of a person includes input variables 502, physiological model parameters 504, one or more glucose increase models 506, and one or more glucose decrease models 508. Glucose increase model 506 may be one or more models that may be used to simulate mechanisms that result in an increase in glucose levels in a person, such as a meal consumption model and/or certain exercise models, among others. Glucose reduction model 508 may be one or more models that may be used to simulate a mechanism that results in a reduction in a person's glucose level, such as a model for simulating delivery of insulin to the body by a delivery device (e.g., 102 of fig. 1), and so forth. Various models may be used to simulate glucose increases and/or decreases in a human body, and one of ordinary skill in the art will recognize such models and embodiments thereof. Modeling and simulating a glycemic response using a glucose increase model and a glucose decrease model may be described below, but it is intended that other types of models may be used to simulate a glycemic response. For example, some models that may be used to simulate a glycemic response to a particular event (e.g., exercise) may take into account both glucose-increasing and glucose-decreasing effects, and such models may be used in addition to and/or in lieu of glucose-increasing and glucose-decreasing models. The present disclosure relating to a glucose-increasing model and a glucose-decreasing model should also apply to such other models and embodiments.
In accordance with aspects of the present disclosure, the glucose increase model 506 and the glucose decrease model 508 are configured to use the input variables 502 and the physiological model parameters 504. The input variables 502 may include event data (e.g., data 310-330) that may be used to simulate a glycemic response to an event. For example, for a meal event, input variables 502 may include meal content information (e.g., protein, fat, and/or carbohydrate count) and/or a start time of a meal. For example, for a disease event, input variables 502 may include a body temperature and/or a start time of fever. For example, for an exercise event, input variables 502 may include the type of exercise. Any event may include data indicative of the amount and timing of insulin delivered to a person by, for example, a delivery device (e.g., 102 of fig. 1). Other types of events and other types of input variables 502 are considered to be within the scope of the present disclosure.
The physiological model parameters 504 may be equation parameters reflecting the physiological state of a person during an event, and thus may vary from person to person and/or over time for a particular person. At least some of the physiological model parameters 504 may correspond to the physiological state data 340 (fig. 3) of the event. The physiological model parameters 504 may be used by a glucose increase model 506 and a glucose decrease model 508 Mimicking the mechanism of glucose increase or decrease in humans. For example, for a meal event, the physiological model parameters 504 may include the rate at which carbohydrates are absorbed into the body (m 1 ) And the rate of conversion of carbohydrates to glucose (k m ). Various physiological model parameters 504 may relate to how protein and fat affect the carbohydrate absorption rate (m 1 ) Etc. For exercise events, physiological model parameters 504 may include endogenous glucose production rates attributable to physical exercise. For any event, the event data may include treatment data (e.g., 320 of fig. 3), which may include the amount and/or timing of insulin delivered to the person (e.g., by delivery device 102 of fig. 1). For insulin delivery, the physiological model parameters 504 may include the rate of insulin absorption in the body, insulin sensitivity index, and the like. Various models can be used to simulate glucose increases and decreases in a human body, and those skilled in the art will recognize physiological model parameters 504 and embodiments thereof for such models.
Using the input variables 502 and the physiological model parameters 504, the glucose increase model 506 may be used to simulate an increase in glucose over time 516 (e.g., glucose is present in human blood). Using the input variables 502 and the physiological model parameters 504, the glucose reduction model 508 may be used to simulate the reduction of glucose over time 518 (e.g., glucose metabolism in the presence of insulin in human blood). The simulated increase in glucose over time 516 and the simulated decrease in glucose over time 518 may be combined and may be applied to the starting body glucose level 522 to provide a body glucose simulated level over time 520 (e.g., a person's glycemic response to an event). While fig. 5 depicts the simulated level of body glucose over time 520 as a juxtaposition of simulated glucose increase over time 516 and simulated glucose decrease over time 518, it should be appreciated that the simulated level of body glucose over time 520 may be depicted as a single curve (e.g., 640 of fig. 6) obtained by subtracting the simulated glucose decrease over time 518 from the simulated glucose increase over time 516.
The illustration and description of fig. 5 is exemplary. In various embodiments, a single model may be used to simulate multiple types of events (e.g., a combined sleep and disease model), or multiple models may be used to simulate glycemic response to a single type of event. In various embodiments, body glucose analog levels over time 520 may be provided without separately modeling glucose increase and glucose decrease. Such and other variations are contemplated as being within the scope of the present disclosure.
Referring now to fig. 6, a diagram of an exemplary technique for determining physiological state data (e.g., physiological model parameters reflecting at least a portion of a physiological state of a person) during an event using a physiological model is shown. The techniques may be implemented in a variety of programming languages and may be executed on a remote or cloud computing system (e.g., 108 of FIG. 1). In various embodiments, if a different computing system (such as computing system 106 of fig. 1) has sufficient processing resources, some or all of these techniques may be performed on such a computing system.
Conceptually, the physiological state of a person may be determined based on a comparison of an actual glucose level to a simulated glucose level. For example, at least a portion of the actual blood glucose response 630 may be compared to at least a portion of the simulated blood glucose response 640. The actual glucose level may be obtained based on physiological measurement data 614 (e.g., one or more glucose levels measured by a glucose sensor), and the simulated glucose level may be obtained based on physiological model 620. The physiological model parameters 624 may be adjusted until the simulated glucose level and the actual glucose level match (e.g., have substantially similar values), at which point the physiological model parameters 624 should reflect the physiological state of the person for the event. The physiological state data 618 may be identified from physiological model parameters 624 that simulate glucose levels that match actual glucose levels.
More specifically, at least some of the event data 610 (event characteristic data 612, physiological measurement data 614, and/or therapy data 616) may be provided as input variables 622 to the physiological model 620. The physiological model parameters 624 may initially be set to default values, values for previous events, or values that are roughly estimated based on the event data 610. Based on the input variables 622 and physiological model parameters 624, a glucose increase model 626 and a glucose decrease model 628 may be used to provide simulated glucose increases and decreases over time, which may be combined and applied to the starting body glucose level to provide a simulated blood glucose response 640. The simulated blood glucose response 640 may be compared to the actual blood glucose response 630 of the person to determine a difference metric 650 therebetween, such as a Mean Absolute Relative Difference (MARD) or another difference metric. The physiological model parameters 624 may be iteratively adjusted in accordance with physiological constraints on the values of these parameters to minimize the difference metric 650. The physiological model parameters 624 corresponding to the minimum difference metric 650 may be identified and at least some of the physiological model parameters 624 may be stored as physiological state data 618 representing a physiological state of the person during the event.
All or part of the glycemic responses 630 and 640 may be compared. For example, to determine the physiological status of a person for a past event, blood glucose reactions 630 and 640 may be compared overall, while to determine the physiological status of a person for an ongoing event, corresponding portions of blood glucose reactions 630 and 640 may be compared.
Those skilled in the art will recognize various ways to adjust the physiological model parameters 624 to minimize the difference metric. Such and other techniques are considered to be within the scope of this disclosure. The illustration and description of fig. 6 are exemplary and variations are considered to be within the scope of the disclosure.
FIG. 7 illustrates an exemplary technique for identifying similar events using physiological models. The technique depicted in fig. 7 involves event data for a plurality of events, including a first event 710 and a second event 720, each having associated event characteristic data (e.g., 712 and 722), physiological measurement data (e.g., 714 and 724), treatment data (e.g., 716 and 726), and/or physiological status data (e.g., 718 and 728). However, to identify similar events while taking into account variability between events, normalized glycemic responses may be compared. Normalization may involve using the same physiological state data and/or the same treatment data for each of a plurality of events.
For example, fig. 7 depicts normalization based on using physiological state data 718 of the first event 710 for both the first event 710 and the second event 720. The technique of fig. 6 may be used to obtain physiological state data 718 for the first event 710. In the example of fig. 7, the physiological model 730 is used to model the glycemic response 750 to the second event 720 using at least the event characteristic data 722 and the physiological measurement data 724 of the second event 720 while borrowing the physiological state data 718 of the first event 710.
The simulated blood glucose response 750 to the second event 720 may be compared to the actual blood glucose response 740 to the first event 710 (e.g., obtained based on the physiological measurement data 714) to determine a difference metric 760, which may be indicative of a degree of similarity between the blood glucose responses. The difference metric 760 may be determined in a variety of ways. For example, the difference metric 760 may be a Mean Absolute Relative Difference (MARD) calculation for a pair of glycemic responses. Other techniques for determining the difference metric are also contemplated. For example, the difference metric 760 may be the sum of absolute differences between the actual blood glucose response 740 and the simulated blood glucose response 750. In various implementations, the difference metric 760 may be calculated based on applying a weighting function to the differences such that differences exceeding a predetermined threshold are penalized.
The difference metric 760 may be compared to a predetermined threshold to determine whether a pair of events are similar. For example, if the difference measure 760 is less than a predetermined threshold, the second event 720 may be identified as similar to the first event 710. However, if the difference metric is greater than or equal to the predetermined threshold, the second event 720 may be identified as dissimilar to the first event 710. Other ways of determining similarity are considered to be within the scope of this disclosure.
In various embodiments, the difference metric 760 may be calculated for the entirety of the blood glucose reactions 740 and 750. In various embodiments, the difference metric 760 may be calculated for corresponding portions of the blood glucose reactions 740 and 750. For example, fig. 7 depicts both blood glucose reactions 740 and 750 as including a portion from the beginning of the event to time T, and these portions may be compared to calculate a difference metric 760.
According to aspects of the present disclosure, the technique of fig. 7 may be performed for more than two sets of event data (e.g., for each pair of event data depicted in fig. 2). For example, if there are m items of event data, the technique of fig. 7 may be performed for each pair of event data (each pair including event data i and event data j), where i and j are both within the range [1, m ], and i+.j.
The event similarity determination results may be organized into a table, such as the table shown in fig. 8, where circles indicate similar events and "x" indicate dissimilar events. The table of fig. 8 includes similarity determination results based on comparing blood glucose responses over a particular duration (e.g., duration from the start of an event to time T). According to aspects of the present disclosure, a separate table may be generated to include similarity determination results based on comparing blood glucose responses over different durations (e.g., durations from the start of an event to different times T').
The table of fig. 8 includes a plurality of rows. Each row may correspond to a similarity determination using different physiological state data. For example, a second row may correspond to a similarity determination of physiological state data using event 1, while a third row may correspond to a similarity determination of physiological state data using event 2.
As described above, similar events may be analyzed to determine a predictive model that may be used to predict a glycemic response based on physiological state data. The analysis may involve generating a graphical representation of the similar event using physiological state data for the similar event. Fig. 9 provides one example of such a graphical representation. Although fig. 9 depicts a three-dimensional representation, it should be understood that techniques described herein may be practiced using graphical representations of differing numbers of dimensions (e.g., less than three dimensions or more than three dimensions).
Each axis 902-906 may correspond to a different input variable of the predictive model. However, at least one axis may correspond to one parameter included in the physiological state data. For example, use inIn the example of a meal event, axis 902 may correspond to parameter k m And axis 904 may correspond to the parameter m 1 Is a value of (2). The optional axis 906 may correspond to a value of the carbohydrate count. Events may be based on their respective k m Value, m 1 Values and/or carbohydrate counts.
The graphical representation of fig. 9 includes a single row of event groupings according to the table of fig. 8 that have been identified as similar events 910. The grouping distinguishes similar events 910 from dissimilar events 920. According to aspects of the present disclosure, the similar events 910 may be used to determine a predictive model based on fitting a curve or surface to the similar events 910. For a two-dimensional graphical representation, a curve may be fitted to a set of similar events, while for a three-dimensional graphical representation, a curved surface may be fitted to a set of similar events. In general, the term "hypersurface" may be used herein to refer to fitting terms in the n dimensions and thus may refer to straight lines, curves or surfaces. Those skilled in the art will appreciate various ways of fitting hypersurface to data, including automated procedures and/or algorithms for performing such fits.
To provide a clear example, the graphical representation of fig. 9 includes only a set of similar events 910. However, it should be understood that in some embodiments, the graphical representation may include multiple groupings, and each grouping may correspond to a different row of the table of fig. 8. Furthermore, a separate hypersurface may be fitted to each grouping such that the graphical representation includes multiple hypersurfaces. From the plurality of hypersurfaces, a consistent hypersurface may be determined, for example, using machine learning or statistical techniques.
Consistent hypersurface may be reduced to one or more equations to be included in the predictive model, according to aspects of the disclosure. The one or more equations may relate the output variable to at least one of the input variables corresponding to one axis of the graphical representation. For example, using a meal event as an example, one or more equations may relate the output variable to k m 、m 1 And/or carbohydrate count correlations. The output variable may indicate (e.g., predict) a person versus event (e.g., an ongoing eventPart) of the blood glucose response. For example, using a meal event as an example, the output variable may indicate all or a portion of the area under the curve as a function of time and representing a predicted rate of glucose occurrence in human blood.
As used herein, fAUC refers to the fraction of area under the curve, and when the output variable indicates a portion of the area under the curve, the output variable may correspond to one fAUC value. Many types of fAUC values are cited herein. For example, fAUC Ra Refers to a portion of the area under the curve representing the rate at which glucose appears in human blood as a function of time, and fAUC Ip Refers to a portion of the area under the curve representing the rate at which insulin appears in human blood as a function of time. In some embodiments, the fAUC value may correspond to the time T 1 And T is 2 The fraction of area under the curve in between. In such embodiments, the fAUC T1，T2 May refer to an area fraction. When T is 1 Corresponding to the start time of an event, fAUC T1，T2 Can be simply written as fAUC T Wherein T is 2 Simply written as T and may correspond to any point in time between the start and end of an event (e.g., T may correspond to a time that generally corresponds to the end of an event).
Notably, the fAUC values can be used to predict glucose levels in humans at a later time. For example, fAUC Ra T The value may provide the cumulative total of glucose that is expected to occur in human blood from the start of the event to time Texpected, and fAUC Ip T The value may provide a cumulative total of insulin that is expected to occur in the human blood from the start of the event to time Texpected. Will fAUC Ip T Multiplying the value by the human insulin sensitivity index will yield the cumulative total amount of glucose expected to be metabolized from the beginning of the event to time Texpected, and this cumulative total amount of glucose expected to be metabolized can be calculated from fAUC Ra T The value is subtracted to derive the person's glucose level at time T (relative to the person's glucose level at the beginning of the event). Thus, the glucose level of the person at time T may be obtained from the output of the predictive model.
The predictive model may be generated in a variety of ways, including through the use of machine learning techniques and/or statistical analysis techniques (e.g., regression analysis), and so forth. Those skilled in the art will recognize such techniques and implementations thereof.
Fig. 10 illustrates an exemplary predictive model 1000. As described above, the predictive model 1000 may be generated using an algorithm configured to fit one or more hypersurfaces to one or more groupings of similar events (910 of fig. 9), determine a consistent hypersurface, and/or reduce the consistent hypersurface to one or more equations. In various embodiments, the predictive model 1000 may be determined based on executing at least a portion of an algorithm on a neural network. In various embodiments, at least a portion of the algorithm may be executed on a remote or cloud computing system (e.g., 108 of fig. 1). In various embodiments, if a different computing system (such as computing system 106 of fig. 1) has sufficient computing resources, some or all aspects of the algorithm may be executed on such a computing system. Those skilled in the art will appreciate various types of machine learning algorithms and implementations thereof.
Event data input 1012 may be provided as an input to predictive model 1000. The event data input 1012 may include some or all of the physiological state data (340 of fig. 3) and may optionally include event characteristic data (310 of fig. 3). For example, if the predictive model 1000 is developed to predict glycemic response to a meal event, the event data input 1012 may include the physiological parameter k m And m 1 And may optionally include an estimated total amount of carbohydrates consumed during a meal event.
Based on the event data input 1012 for an event, the predictive model 1000 derives an output value 1022 indicative of a glycemic response to the event (e.g., information indicative of an accumulated amount of glucose predicted to have occurred in the human blood since the event). The output value 1022 may be specific to a particular duration from the start of the event (e.g., the duration from the start of the event to time T). Thus, different predictive models may be used to determine output values corresponding to different durations from the beginning of an event.
As aboveThe output value 1022 may be used to derive a glycemic response to the event. For example, output value 1022 may be fAUC Ra T Values that can be used to determine glucose values along a glucose response curve.
Fig. 11A-11B depict exemplary techniques for deriving a blood glucose response curve. The glucose response curve may be derived based on the physiological measurement data (e.g., 330 of fig. 3) and the output values 1022 of the predictive model 1000. For example, physiological measurement data may be used to determine a point near the beginning of the glucose response curve, and output value 1022 may be used to predict a point near the end of the glucose response curve. As described above, the output value 1022 may be specific to a particular duration from the beginning of the event. Thus, the output value 1022 may be used to determine a glucose value corresponding to a particular time value T after the start of the event. Thus, different predictive models may be used to obtain different output values, each corresponding to a different time value after the start of an event.
For example, fig. 11A-11B depict a blood glucose response curve generated based on two different output values obtained from two different predictive models. FIG. 11A depicts a time T from the beginning of an event based on specificity a A portion of a glycemic response curve generated based on output values specific to duration of time from the beginning of the event to time T is depicted in FIG. 11B b A portion of a glucose response curve generated from the output value of the duration of (a). While fig. 11A-11B depict two different time values, it should be appreciated that the techniques disclosed herein may be practiced with any number of time values (e.g., more than two time values or less than two time values).
In FIGS. 11A-11B, glucose values 1104a-B correspond to, respectively, time values T along the glucose response curve a And T b Points at. Glucose values 1104a-b may be determined based on subtracting fAUC values 1106a-b from fAUC values 1102a-b, respectively. For example, curve 1102 may represent the rate at which glucose occurs in human blood as a function of time, and fAUC value 1102a may correspond to the time value T from the beginning of the event a A cumulative total amount of glucose expected to occur in human blood;whereas curve 1106 may represent the rate at which glucose is metabolized in human blood as a function of time, and fAUC value 1106a may correspond to the expected time value T from the beginning of the event a A cumulative total amount of blood glucose metabolized therebetween. Thus, subtracting fAUC value 1106a from fAUC value 1102a will yield the expected time value T a A net change in glucose level occurs (relative to the person's glucose level at the beginning of the event 1110).
As described above, the fmac values 1102a-b may be obtained from different predictive models. For example, the first predictive model may derive a model specific to the time value T from the beginning of the event a The first output value of the duration of (a) and the second predictive model may derive a value specific to the time value T from the start of the event b A second output of the duration of (a). In some embodiments, each predictive model may be generated at a remote or cloud computing system (e.g., one or more server computers) and transmitted to a different computing system (e.g., a smart phone), where the predictive model is stored locally for application at the different computing system.
The fAUC values 1106a-b may be determined based on providing therapy data (e.g., 320 of FIG. 3) to the physiological model 500 to simulate glucose metabolism. For example, model 500 may provide an equation or curve representing the rate at which glucose is metabolized in the presence of insulin, and through a time value T a And T b To integrate the equation or curve, which will yield the expected slave time value T, respectively a And T b The cumulative total amount of glucose that begins to be metabolized.
For the avoidance of doubt, the techniques depicted in fig. 11A-11B are provided as examples only, and other techniques may be used to derive the blood glucose response curve. For example, some techniques may involve using fAUC values 1102a-b to determine glucose values 1104a-b without regard to glucose metabolism (e.g., in the absence of therapeutic data).
In some embodiments, the output value 1022 of the predictive model 1000 may be used for therapy determination. For example, the output value 1022 may be used to determine an amount of insulin delivery (e.g., a size of bolus to be delivered to a person associated with a meal event) to maximize a time range. In such embodiments, the predicted glycemic response will be the desired glycemic response. Thus, glucose values 1104a-b and fAUC values 1102a-b may be used to calculate fAUC values 1106a-b in the reverse direction, which may be used to determine the total amount of glucose to be metabolized to achieve a desired glycemic response. For example, fAUC values 1106a-b may be used to derive a glucose metabolism rate curve, and calculating the area under the curve will yield the total amount of glucose metabolized. To determine the amount of insulin delivered, the total amount of glucose metabolized may be divided by the insulin sensitivity index.
In some embodiments, the treatment may be administered based on transmitting the treatment determination toward the treatment delivery device. One non-limiting example of such an apparatus is described below in connection with fig. 12.
As described above, the therapy determination may be communicated toward the insulin delivery device 1200 (e.g., from the cloud computing system 108 via an intermediary computing device 106 communicatively coupled to the device 1200). In such devices, insulin delivery may be performed based on internal communications between a central computing module (e.g., a microcontroller for the entire device 1200) and insulin delivery modules (e.g., including a microcontroller, motor, and pump). For example, insulin delivery may be caused by the central computing module transmitting a delivery command in the form of an electrical signal that travels to the insulin delivery module via the communication structure. The central computing module may also be configured to communicate (e.g., via a transceiver) with a computing device (e.g., 106 of fig. 1) communicatively coupled to a remote or cloud computing system (e.g., 108 of fig. 1). In accordance with the techniques described herein, the insulin delivery device 1200 may transmit various event data towards a remote or cloud computing system, which may transmit insulin delivery determination results towards the insulin delivery device 1200.
Insulin delivery device 1200 may provide rapid-acting insulin through a small tube 1210 configured for fluid connection with a subcutaneously inserted cannula. The device 1200 may deliver two types of doses: a basal dose that can be delivered in minute amounts periodically (e.g., every five minutes) throughout the day and night; and bolus doses to cover the increase in blood glucose caused by eating and/or correct for hyperglycemic levels. The illustrated insulin delivery device 1200 includes a user interface with button elements 1220 that can be manipulated to administer insulin bolus injections, change therapy settings, change user preferences, select display features, and the like. The insulin delivery device 1200 also includes a display device 1230 that may be used to present various types of information or data to a user. According to aspects of the present disclosure, a user of the insulin delivery device 1200 may use the button element 1220 to input certain event data (e.g., event type, event start time, event details, etc.), and may use the display device 1230 to confirm the user input. The illustrated insulin delivery device 1200 of fig. 12 is exemplary and other types of insulin delivery devices and other techniques other than the above described techniques are considered to be within the scope of the present disclosure.
The embodiments disclosed herein are examples of the present disclosure and may be embodied in various forms. For example, while certain embodiments herein are described as separate embodiments, each of these embodiments herein may be combined with one or more of the other embodiments herein. Specific structural and functional details disclosed herein are not to be interpreted as limiting, but rather as a basis for the claims and as a representative basis for teaching one skilled in the art to variously employ the present disclosure in virtually any appropriately detailed structure. Throughout the description of the drawings, like reference numerals may refer to like elements.
The phrases "in one embodiment," "in an embodiment," "in various embodiments," "in some embodiments," or "in other embodiments" may each refer to one or more of the same or different embodiments in accordance with the present disclosure. The phrase in the form "a or B" means "(a), (B) or (a and B)". The phrase of the form "at least one of A, B or C" means "(a); (B); (C); (A and B); (A and C); (B and C); or (A, B and C) ".
Any of the techniques, operations, methods, programs, algorithms, or code described herein may be converted to or expressed in a programming language or computer program embodied on a computer, processor, or machine readable medium. As used herein, the terms "programming language" and "computer program" each include any language for specifying instructions for a computer or processor, and include, but are not limited to, the following languages and derivatives thereof: assembler, basic, batch file, BCPL, C, c+, c++, delphi, fortran, java, javaScript, machine code, operating system command language, pascal, perl, PL1, python, scripting language, visual Basic, meta language itself specified programming, and all first, second, third, fourth, fifth, or higher generation computer languages. Databases and other data schemas, as well as any other meta-languages, are included. No distinction is made between interpretation, compilation language, or language that uses both compilation and interpretation methods. No distinction is made between the compilation and source version of the program. Thus, reference to a program that a programming language may exist in more than one state (such as source, compiled, target, or linked) is a reference to any and all such states. References to programs may encompass actual instructions and/or the intent of such instructions.
It should be understood that the foregoing description is only illustrative of the present disclosure. To the extent consistent, any or all of the aspects detailed herein may be used in combination with any or all of the other aspects detailed herein. Various alternatives and modifications can be devised by those skilled in the art without departing from the disclosure. Accordingly, the present disclosure is intended to embrace all such alternatives, modifications and variances. The embodiments described with reference to the figures are presented only to illustrate certain examples of the present disclosure. Other elements, steps, methods, and techniques that are not significantly different than those described above and/or in the appended claims are also intended to be included within the scope of this disclosure.
Although several embodiments of the present disclosure have been illustrated in the accompanying drawings, it is not intended to limit the disclosure thereto, as it is intended that the disclosure be as broad in scope as the art will allow and should be read in the same way. Therefore, the above description should not be construed as limiting, but merely as exemplifications of particular embodiments. Those skilled in the art will envision other modifications within the scope and spirit of the claims appended hereto.

Claims (20)

1. A system, the system comprising:
one or more processors; and
one or more processor-readable media storing instructions that when executed by the one or more processors cause performance of the following:
obtaining a predictive model, the predictive model correlating:
a glycemic response of a person to an event, and
a physiological parameter of the person during the event;
obtaining a glucose level measurement of the person during the event;
determining a physiological parameter of the person during the event based on the glucose level measurement; and
based on applying the predictive model to the physiological parameter, a glycemic response of the person to the event is predicted.
2. The system of claim 1, wherein the event is a meal event.
3. The system of claim 1, wherein the physiological parameter is a metabolic parameter.
4. A system according to claim 3, wherein the metabolic parameter is the rate at which carbohydrates are absorbed into the human body.
5. A system according to claim 3, wherein the metabolic parameter is the rate of conversion of carbohydrates to glucose.
6. The system of claim 1, wherein the one or more processor-readable media further store instructions that when executed by the one or more processors cause the following to be performed:
carbohydrate content information for the event is obtained.
7. The system of claim 1, wherein the one or more processor-readable media further store instructions that when executed by the one or more processors cause the following to be performed:
insulin delivery information is obtained, the insulin delivery information indicating an amount of insulin delivered to the person during the event.
8. The system of claim 7, wherein predicting a patient's glycemic response comprises accounting for glucose metabolism caused by the amount of insulin delivered to the person during the event.
9. The system of claim 1, wherein the one or more processor-readable media further store instructions that when executed by the one or more processors cause the following to be performed:
determining an amount of insulin delivered to the person during the event based on the predicted glycemic response; and
Based on the amount of insulin delivered towards the insulin delivery device, delivery of the amount of insulin is caused.
10. The system of claim 9, wherein transmitting the amount of insulin toward the insulin delivery device comprises transmitting the amount of insulin to an intermediary computing device communicatively coupled to the insulin delivery device.
11. The system of claim 1, wherein determining the physiological parameter comprises:
acquiring a physiological model of the person; and
the physiological parameter is derived based on applying the physiological model to the glucose level measurement.
12. The system of claim 1, wherein predicting the patient's glycemic response comprises:
obtaining a value indicative of a cumulative total of glucose that is predicted to have occurred in the human blood after a certain amount of time has elapsed since a start time of the event; and
based on the values, a curve is generated representing the glucose level of the person during the particular amount of time.
13. The system of claim 12, wherein the value corresponds to a fraction of an area under a curve representing a rate at which glucose appears in the human blood as a function of time.
14. The system of claim 1, wherein the predictive model is configured to output a value corresponding to a cumulative total of glucose that is predicted to have occurred in the human blood after a specified amount of time has elapsed since a start time of the event.
15. The system of claim 1, wherein the predictive model is generated based on a set of past events that are identified as eliciting a similar glycemic response from the person if the physiological parameter of the person is the same during each of the set of past events.
16. The system of claim 15, wherein the set of past events includes a first event and a second event, an actual glycemic response of the person to the first event is identified as similar to a hypothetical glycemic response of the person to the second event, the hypothetical glycemic response to the second event determined based on application of a physiological model of the person to physiological parameters of the person during the first event.
17. The system of claim 15, wherein one or more of the physiological parameters of the person during each past event in the set of past events are used to generate a graphical representation, and wherein a hypersurface is fitted to the graphical representation to generate the predictive model.
18. The system of claim 1, wherein the predictive model is specific to the person.
19. A processor-implemented method, the processor-implemented method comprising:
obtaining a predictive model, the predictive model correlating:
a glycemic response of a person to an event, and
a physiological parameter of the person during the event;
obtaining a glucose level measurement of the person during the event;
determining a physiological parameter of the person during the event based on the glucose level measurement; and
based on applying the predictive model to the physiological parameter, a glycemic response of the person to the event is predicted.
20. One or more processor-readable media storing instructions that when executed by one or more processors cause performance of the following:
obtaining a predictive model, the predictive model correlating:
a glycemic response of a person to an event, and
a physiological parameter of the person during the event;
obtaining a glucose level measurement of the person during the event;
determining a physiological parameter of the person during the event based on the glucose level measurement; and
Based on applying the predictive model to the physiological parameter, a glycemic response of the person to the event is predicted.
CN202280041824.6A 2021-06-30 2022-06-30 Event-oriented blood glucose response prediction Pending CN117546251A (en)

Applications Claiming Priority (5)

Application Number Priority Date Filing Date Title
US63/216,865 2021-06-30
US63/304,090 2022-01-28
US17/852,878 US20230000447A1 (en) 2021-06-30 2022-06-29 Event-oriented predictions of glycemic responses
US17/852,878 2022-06-29
PCT/US2022/035651 WO2023278653A1 (en) 2021-06-30 2022-06-30 Event-oriented predictions of glycemic responses

Publications (1)

Publication Number Publication Date
CN117546251A true CN117546251A (en) 2024-02-09

Family

ID=89794337

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202280041824.6A Pending CN117546251A (en) 2021-06-30 2022-06-30 Event-oriented blood glucose response prediction

Country Status (1)

Country Link
CN (1) CN117546251A (en)

Similar Documents

Publication Publication Date Title
US11998322B2 (en) Intermittent monitoring
US10842419B2 (en) Method, system and computer program product for CGM-based prevention of hypoglycemia via hypoglycemia risk assessment and smooth reduction insulin delivery
JP6986007B2 (en) Systems, devices and methods of dynamic glucose profile response to physiological parameters
US20220273204A1 (en) Intermittent Monitoring
US20210256872A1 (en) Devices, systems, and methods for predicting blood glucose levels based on a personalized blood glucose regulation model
KR20200066204A (en) Method for providing recommendations for maintaining a healthy lifestyle basing on user's daily activity parameters, automatically tracked in real time, and corresponding system
JP2023509974A (en) Bolus dose administration
CN117413319A (en) Personalized food recommendation based on sensed biomarker data
US20230000447A1 (en) Event-oriented predictions of glycemic responses
Cescon et al. Linear modeling and prediction in diabetes physiology
US20210177313A1 (en) Continuous health monitoring system
CN117546251A (en) Event-oriented blood glucose response prediction
US20210398639A1 (en) Default carbohydrate consumption counts based on population carbohydrate consumption models
CN109788911A (en) For determining the method and system and non-transitory computer-readable medium of carbohydrate intake event from the glucose monitoring data of instruction glucose level
JP2023513805A (en) Decision support and therapeutic delivery systems
US20230270362A1 (en) Continuous health monitoring system
US11806137B2 (en) Real-time meal detection based on sensor glucose and estimated plasma insulin levels
WO2022180184A1 (en) Continuous health monitoring system
US20240325639A1 (en) Blood glucose estimation based on rate of change information
US20220230751A1 (en) Decision support system, and method in relation thereto
WO2023102498A1 (en) Real-time meal detection based on sensor glucose and estimated plasma insulin levels
CN118302825A (en) Meal delivery to correct bolus
Islam Integrated physiological model of virtual diabetic patient for machine learning research
WO2023219847A1 (en) Cost-effective therapy recommendations

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination