Decision support and therapy management system

Abstract

Techniques for data analysis and user guidance are provided for determining and providing one or more therapies to a user based on where the user is or will be in their menstrual cycle. In certain embodiments, a method of personalizing diabetes therapy based on information related to a user's menstrual cycle is provided. The method includes measuring blood glucose measurements of the user using a glucose monitoring system. The method also includes receiving information related to the user's menstrual cycle. The method further includes determining a therapy for the user to achieve a target blood glucose during a sub-phase or phase of the user's menstrual cycle based on at least one of historical data associated with the user and historical data associated with a stratified group.

Description

Technical Field

[0001] This application claims the rights to U.S. Application Serial No. 62 / 976778, entitled "Women's Health and Glycemic Control," filed February 14, 2020, and U.S. Application Serial No. 63 / 011175, entitled "Decision Support and Treatment Management System," filed April 16, 2020. The aforementioned provisional applications are incorporated herein by reference in their entirety. Background Technology

[0002] This application generally relates to medical devices such as analyte sensors, including systems and methods for providing treatment to patients using the medical device.

[0003] Related technical specifications

[0004] Diabetes is a metabolic condition related to the body's production or use of insulin. Insulin is a hormone that allows the body to use glucose as energy or store glucose as fat.

[0005] When a person eats a meal containing carbohydrates, the food is processed by the digestive system, producing glucose in the bloodstream. Blood glucose can be used as energy or stored as fat. The body normally maintains blood glucose levels within a range sufficient for energy to support bodily functions and avoid problems that can occur when glucose levels are too high or too low. The regulation of blood glucose levels depends on the production and use of insulin, which regulates the movement of blood glucose into cells.

[0006] When the body doesn't produce enough insulin, or when it can't effectively use the insulin it has, blood sugar levels rise outside the normal range. This condition, where blood sugar levels are higher than normal, is called "hyperglycemia." Chronic hyperglycemia can lead to many health problems, such as cardiovascular disease, cataracts and other eye problems, nerve damage (neuropathy), and kidney damage. Hyperglycemia can also cause acute problems, such as diabetic ketoacidosis—a condition where the body becomes excessively acidic due to the presence of blood sugar and ketones, which are produced when the body cannot use glucose. A condition where blood sugar levels are lower than normal is called "hypoglycemia." Severe hypoglycemia can lead to acute crises, such as seizures or death.

[0007] People with diabetes can receive insulin to control their blood sugar levels. For example, insulin can be received manually via injection with a needle. Wearable insulin pumps are also available. Diet and exercise also affect blood sugar levels.

[0008] Diabetes is sometimes referred to as "type 1" and "type 2." People with type 1 diabetes are usually able to use insulin when it's available, but due to a problem with the pancreas's beta cells that produce insulin, the body cannot produce enough. People with type 2 diabetes may produce some insulin, but due to decreased insulin sensitivity, they have become "insulin resistant." As a result, even when insulin is present, the body doesn't use it effectively to regulate blood sugar levels.

[0009] Managing diabetes presents complex challenges for patients, clinicians, and caregivers because many factors combine to influence a patient's glucose levels and glucose trends. For example, a woman's menstrual cycle can significantly affect her insulin resistance, depending on which stage of her menstrual cycle she is in.

[0010] More specifically, a menstrual cycle typically lasts 21 to 35 days, with an average of about 28 days. During this cycle, hormonal fluctuations not only trigger ovulation and menstruation but also affect the body's insulin resistance. The menstrual cycle generally consists of four phases: menstruation, the follicular phase, ovulation, and the luteal phase. During the luteal phase of the menstrual cycle, a hormone called progesterone is released, which can lead to insulin resistance, resulting in more frequent episodes of high blood sugar, even if the patient follows the same exercise, diet, and / or insulin therapy regimen. Therefore, during the luteal phase of a patient's menstrual cycle, treatments recommended by decision support systems to help women control their glucose levels may be less effective.

[0011] This background is provided to provide a brief context for the following invention and detailed description. This background is not intended to help define the scope of the claimed subject matter, nor should it be construed as limiting the claimed subject matter to embodiments that address any or all of the aforementioned drawbacks or problems. Summary of the Invention

[0012] One aspect is a system comprising: a glucose monitoring system including: a glucose sensor configured to measure a user's blood glucose; a sensor electronics module configured to transmit sensor data corresponding to the blood glucose measurement value provided by the glucose sensor to a processor; and storage circuitry; the processor being configured to: receive information related to the user's menstrual cycle, and, based on at least one of historical data associated with the user and historical data associated with the user's stratified group, determine a treatment for the user to achieve a target blood glucose level in a sub-phase or phase of the menstrual cycle, wherein: the historical data associated with the user includes blood glucose measurements of the user provided by the glucose monitoring system; and the historical data associated with the user is structured to indicate the following At least one of the following: a pattern of at least one of the user's blood glucose measurements and insulin resistance during a sub-phase or phase of the user's menstrual cycle, and a pattern of the physiological effect of treatment on the user's blood glucose measurements during a sub-phase or phase of the user's menstrual cycle, the pattern indicating the effectiveness of treatment in achieving target blood glucose; and historical data associated with the user's stratification group being structured to indicate at least one of the following: a pattern of at least one of the user's blood glucose measurements and insulin resistance during a sub-phase or phase of the menstrual cycle, and a pattern of the physiological effect of treatment on the user's blood glucose measurements during a sub-phase or phase of the menstrual cycle, the pattern indicating the effectiveness of treatment in achieving target blood glucose.

[0013] In the aforementioned system, the processor is configured to provide treatment. The treatment in the aforementioned system includes a dose of insulin. In the aforementioned system, the insulin dose is higher than the average insulin dose administered to the user during the non-luteal phase or phase of their menstrual cycle. In the aforementioned system, the processor is configured to send a signal to a drug delivery device to administer the insulin dose to the user.

[0014] In the aforementioned system, the processor is configured to provide a treatment recommendation to a user or another person, the recommendation indicating an insulin dosage. In the aforementioned system, the processor is configured to provide a treatment recommendation to a user or another person, the recommendation indicating at least one of the amount, type, duration, and intensity of exercise. In the aforementioned system, the processor is configured to provide a treatment recommendation to a user or another person, the recommendation indicating at least one of the amount and type of food.

[0015] On the other hand, there is a method for personalizing diabetes treatment based on information related to a user's menstrual cycle, the method comprising: measuring the user's blood glucose level using a glucose monitoring system; receiving information related to the user's menstrual cycle in a processor communicating with the glucose monitoring system; and, in the processor, determining a treatment for the user to achieve a target blood glucose level during a sub-phase or phase of the user's menstrual cycle based on at least one of historical data associated with the user and historical data associated with the user's stratified group, wherein: the historical data associated with the user includes the user's blood glucose level measurements provided by the blood glucose monitoring system; and the historical data associated with the user is structured to indicate at least one of the following: Patterns of at least one of a user's blood glucose measurements and insulin resistance during sub-phases or phases of the user's menstrual cycle, and patterns of the physiological effects of treatment on the user's blood glucose measurements during sub-phases or phases of the user's menstrual cycle, indicating the effectiveness of treatment in achieving target blood glucose; historical data associated with the user's stratification group are structured to indicate at least one of the following: patterns of at least one of a user's blood glucose measurements and insulin resistance during sub-phases and phases of the menstrual cycle, and patterns of the physiological effects of treatment on the user's blood glucose measurements during sub-phases or phases of the menstrual cycle, indicating the effectiveness of treatment in achieving target blood glucose.

[0016] The above methods also include providing treatment. In the above methods, the treatment includes administering a dose of insulin. In the above methods, the insulin dose is higher than the average insulin dose administered to the user during non-luteal phases or menstrual cycles. In the above methods, providing treatment also includes signaling to a drug delivery device to administer the insulin dose to the user. In the above methods, providing treatment further includes providing a treatment recommendation to the user or other individual, which indicates the insulin dose.

[0017] In the above methods, providing treatment further includes providing treatment recommendations to the user or other individual, which instruct at least one of the amount, type, duration, and intensity of exercise. In the above methods, providing treatment also includes providing treatment recommendations to the user or other individual, which instruct at least one of the amount and type of food.

[0018] Another aspect is a non-transitory computer-readable medium having instructions stored thereon, which, when executed by a system, cause the system to perform a method comprising: measuring a user's blood glucose measurement using a glucose monitoring system; receiving, in a processor communicating with the glucose monitoring system, information related to the user's menstrual cycle; and, in the processor, determining a treatment for the user to achieve a target blood glucose level during a sub-phase or phase of the user's menstrual cycle based on at least one of historical data associated with the user and historical data associated with the user's stratified group, wherein: the historical data associated with the user includes the user's blood glucose measurements provided by the glucose monitoring system; and the historical data associated with the user is structured to indicate... At least one of the following: a pattern of at least one of the user's blood glucose measurements and insulin resistance during a sub-phase or phase of the user's menstrual cycle, and a pattern of the physiological effect of treatment on the user's blood glucose measurements during a sub-phase or phase of the user's menstrual cycle, the pattern indicating the effectiveness of treatment in achieving target blood glucose; historical data associated with the user's stratification group is structured to indicate at least one of the following: a pattern of at least one of the user's blood glucose measurements and insulin resistance during a sub-phase or phase of the menstrual cycle, and a pattern of the physiological effect of treatment on the user's blood glucose measurements during a sub-phase or phase of the menstrual cycle, the pattern indicating the effectiveness of treatment in achieving target blood glucose.

[0019] In the aforementioned medium, the method further includes providing treatment. In the aforementioned medium, the treatment comprises a dose of insulin. In the aforementioned medium, the insulin dose is higher than the average insulin dose given to the user during non-luteal phases or menstrual cycles.

[0020] Any feature of one aspect applies to all aspects defined herein. Furthermore, any feature of one aspect may be combined independently, in any way, partly or wholly, with other aspects described herein; for example, one, two, or three or more aspects may be combined wholly or wholly. Additionally, any feature of one aspect may be optional for other aspects. Any aspect of the method may include another aspect of a system for personalizing diabetes treatment based on information related to a user's menstrual cycle, and any aspect of a system for personalizing diabetes treatment based on information related to a user's menstrual cycle may be configured to perform the method of the other aspect. Attached Figure Description

[0021] Figure 1A An exemplary decision support and treatment management system (“DSTA”) is described based on some of the embodiments disclosed herein.

[0022] Figure 1B Based on some of the implementation methods disclosed herein, the following is described Figure 1AExemplary glucose monitoring systems, more specifically, and some mobile devices.

[0023] Figure 2 Based on some implementations disclosed herein, exemplary inputs and exemplary metrics calculated based on those inputs are described for the purpose of... Figure 1A DSTA is used.

[0024] Figure 3 It is based on some embodiments disclosed herein, such as Figure 1A A flowchart illustrating an exemplary operation performed by the DSTA system.

[0025] Figure 4 This is a diagram illustrating one embodiment of how to determine one or more treatments for a user, based on descriptions of certain implementation methods.

[0026] Figure 5 Based on certain embodiments disclosed herein, a description is provided of a configuration for performing Figure 3 A block diagram of a computing device for one or more steps of the operation described. Detailed Implementation

[0027] In some implementations, an application as described herein provides guidance and treatment that can assist patients, caregivers, healthcare providers, or other users in improving lifestyle or clinical / patient outcomes by addressing various challenges such as nighttime glucose control (e.g., reducing the incidence of hypoglycemic events or hyperglycemic fluctuations), intraprandial and postprandial glycemic control (e.g., using historical information and trends to increase glycemic control), hyperglycemic correction (e.g., increasing the time to target zones while avoiding hypoglycemic events due to overcorrection), hypoglycemic treatment (e.g., resolving hypoglycemia while avoiding “rebound” hyperglycemia), exercise, and / or other health factors. In some implementations, the application may be further configured with optimization tools that learn the patient’s physiological functions and behaviors and calculate guidance to help the patient identify optimal or ideal treatment parameters, such as basal insulin requirements, insulin-to-carbohydrate ratio, correction factors, and / or changes in insulin sensitivity due to exercise.

[0028] For example, the application can help patients respond to problems in real time by predicting hypoglycemic or hyperglycemic events or trends, providing treatment to address ongoing or potential hypoglycemic or hyperglycemic events or trends, and / or monitoring patients' blood glucose, physiological, and / or behavioral responses to different events. This type of computationally-driven guidance and support can reduce the cognitive burden on users.

[0029] Physiological sensors such as continuous glucose monitors can provide useful data that users can use to manage glucose levels, but this data may require extensive processing to develop effective strategies for glucose management. The sheer volume of data, and the understanding of the correlations between data types, trends, events, and outcomes, can far exceed human processing capabilities. This is especially impactful when making real-time decisions about treatment or responses to physiological conditions. Combining real-time or recent data with historical data and patterns can provide useful guidance for making real-time decisions about treatment. Technological tools can process this information to provide decision support guidance that, after calculation, is useful for a specific patient at a specific time under specific conditions or circumstances.

[0030] As mentioned above, a woman's menstrual cycle can significantly affect her insulin resistance depending on which stage of her cycle she is in. However, some existing decision support systems do not take into account this variation in user insulin resistance. For example, it has been shown that patients have different levels of insulin resistance at different stages. While not limiting, one example is the luteal phase. Therefore, existing decision support systems may continue to provide the same guidance (e.g., the same insulin dosage, exercise, or dietary recommendations) at different stages of the menstrual cycle, without considering the stage of the menstrual cycle. As a result, female users may become frustrated when the same treatment that was effective a few days ago is no longer as effective.

[0031] Adding to the complexity is the fact that the amount of insulin resistance fluctuation varies among different users. For example, during the luteal phase, insulin resistance increased by 20% in the first female user, while it increased by 40% in the second female user. This difference can be due to one or more factors, including but not limited to age, weight, race, ethnicity, dietary type, and other types of medical conditions the user may have. Furthermore, the type and amount of treatment that helps the aforementioned first user cope with increased insulin resistance may differ from the type and amount of treatment that helps a third user who experiences a 20% increase in insulin resistance at different stages of the menstrual cycle. Additionally, the percentage change in insulin resistance during the menstrual cycle for the first user (e.g., an increase during the luteal phase) can vary over time due to a variety of reasons, including but not limited to age, weight changes, stress, pregnancy, other types of medical conditions, and diet.

[0032] Therefore, certain embodiments described herein provide technical solutions to the aforementioned technical problems in the field of diabetes intervention management. In some embodiments, the technical solutions provided herein improve existing decision support systems by configuring a system to provide more precise guidance and treatment based on at least one of the following: the user's stage in the menstrual cycle, and further based on records of the user's own historical data (e.g., how the user's physiological functions have responded in the same time / period of the user's menstrual cycle in the past) and records of historical data associated with one or more other users (e.g., how the physiological functions of other similar users (e.g., based on similarity of one or more factors) have responded in the same time / period of the user's menstrual cycle). The main part of the improved DSTA system described herein is a glucose monitoring system 104, which includes a sensor electronics module and a continuous analyte sensor 140 (see [link to documentation]). Figure 1B This allows for continuous measurement of a user's glucose levels and the transmission of glucose measurements to one or more processors within the DSTA system in real-time or near real-time. Considering the user's menstrual cycle, it would be difficult, even if not possible, to provide the user with treatment relevant to their real-time glucose status without receiving a continuous stream of glucose measurements (e.g., in real-time or near real-time) from the glucose monitoring system 104. In other words, without the glucose monitoring system 104, which improves upon the DSTA system described herein, the user would be limited to certain existing techniques for measuring their glucose levels using finger-prick blood sampling. However, it is extremely difficult for the user to continuously (e.g., every five minutes) measure their glucose levels using finger-prick blood sampling, even if it is not impossible or impractical. Therefore, without the glucose monitoring system 104 described herein, the user's glucose measurement results would at best be fragmented, thus making the operation of the DSTA system described herein difficult, or causing them to provide potentially inaccurate or irrelevant treatment based on discontinuous and fragmented glucose measurements. As a result, the operation of the DSTA system described herein is improved by or dependent on the glucose monitoring system 104, which is capable of providing a continuous and true or near real-time stream of glucose measurements.

[0033] In some implementations, records of how a user's physiological function responds at the same time / period of their menstrual cycle may, among other things, include how resistant the user has become to insulin at that time / period in a previous cycle, changes in the user's glucose levels or other indicators in response to previously proposed treatment, etc. In some implementations, records of how other similar users' physiological function responds at the same time / period may, among other things, include similar data regarding the physiological function of such users. Precision treatment may include various treatment recommendations to help users manage their glucose levels, for example, by keeping their glucose levels within a desired range despite changes in insulin resistance during different phases of the menstrual cycle. Precision treatment may include a specific insulin dose calculated based on the factors described above to help users maintain their glucose levels within the range even considering changes in insulin resistance (e.g., increases). In some implementations, an insulin dose may be provided to the user in the form of a treatment recommendation, upon which the user may manually administer (e.g., by injection or oral intake) the insulin dose. In some implementations, the insulin dose can be signaled to a drug delivery device (e.g., an insulin pump or pen or other insulin administration device), based on which the drug delivery device automatically administers a recommended dose of insulin (e.g., after the user approves the recommended dose).

[0034] Precision treatment may additionally or alternatively include treatment recommendations for the user to engage in a certain amount and / or type of exercise to lower her glucose levels back into the range. Furthermore, precision treatment may additionally or alternatively include treatment recommendations for consuming and / or avoiding certain amounts or types of food and / or for lowering her glucose levels back into the range over a certain period of time. Providing precise treatment recommendations (e.g., insulin or other medication dosages, activity levels, carbohydrate intake, etc.) regardless of the user's stage of the menstrual cycle is an improvement over existing decision support systems, leading to significant health improvements and user outcomes.

[0035] As described above, in some embodiments, the application utilizes input from one or more physiological sensors (such as one or more analyte sensors) to provide relevant and effective guidance and treatment. An embodiment of the analyte sensor described herein is a glucose monitoring sensor that measures glucose and / or indicates the concentration of glucose and / or another analyte or the concentration of a substance present in the user's body. In some embodiments, the glucose monitoring sensor is a continuous glucose monitoring device, such as a subcutaneous, transdermal, percutaneous, non-invasive, intraocular, and / or intravascular (e.g., intravenous) device. In some embodiments, the device can analyze multiple intermittent blood samples. The glucose monitoring sensor can use any glucose measurement method, including enzymatic, chemical, physical, electrochemical, optical, photochemical, fluorescence-based methods, spectrophotometry, spectroscopy (e.g., optical absorption spectroscopy, Raman spectroscopy, etc.), optical rotation, calorimetry, iontophoresis, radiometry, etc.

[0036] Glucose monitoring sensors can use any known detection method, including invasive, minimally invasive, and non-invasive sensing technologies, to provide a data stream indicating the concentration of the analyte in the host. This data stream is typically a raw data signal used to provide useful information about the analyte to patients or healthcare professionals (such as doctors, physicians, nurses, and caregivers) who may be using the sensor.

[0037] In some embodiments, the glucose monitoring sensor is an implantable sensor, such as those described with reference to U.S. Patent No. 6,001,067 and U.S. Patent Publication No. US-2011-0027127-A1. In some embodiments, the glucose monitoring sensor is a percutaneous sensor, such as those described with reference to U.S. Patent Publication No. US-2006-0020187-A1. In other embodiments, the glucose monitoring sensor is a dual-electrode analyte sensor, such as those described with reference to U.S. Patent Publication No. US-2009-0137887-A1. In still other embodiments, the glucose monitoring sensor is configured to be implanted in a host blood vessel or externally, such as the sensor described in U.S. Patent Publication No. US-2007-0027385-A1. These patents and publications are incorporated herein by reference in their entirety.

[0038] Exemplary System

[0039] Figure 1AAn exemplary DSTA 100 is described, which in some embodiments includes a diabetes intervention application (“application”) 106 that provides decision support guidance to a user 102 (hereinafter referred to as “user”) and determines / administers one or more treatments. In some embodiments, the user may be a patient or a patient’s caregiver. In the embodiments described herein, it is assumed that the user is a patient for simplicity only, but this is not a limitation. In some embodiments, DSTA 100 includes a user, a glucose monitoring system 104, a mobile device 107 executing the application 106, a decision support engine 112, an optional insulin delivery device (not shown), and a user database 110.

[0040] In some embodiments, the glucose monitoring system 104 includes a sensor electronics module and a glucose sensor that measures blood glucose and / or indicates the concentration of glucose and / or another analyte or the concentration of a substance present in the user's body. In some embodiments, the glucose sensor is configured to perform measurements on a continuous basis. The sensor electronics module transmits the blood glucose measurement values ​​to a mobile device 107 for use by an application 106. In some embodiments, the sensor electronics module transmits the glucose measurement values ​​to the mobile device 107 via a wireless connection (e.g., Bluetooth connection). In some embodiments, the mobile device 107 is a smartphone. However, in some embodiments, the mobile device 107 may alternatively be any other type of computing device, such as a laptop, smartwatch, tablet, or any other computing device capable of executing the application 106.

[0041] In some implementations, decision support engine 112 refers to a set of software instructions having one or more software modules, including a data analysis module (DAM) 113. In some implementations, decision support engine 112 executes entirely on one or more computing devices in a private or public cloud. In such implementations, application 106 communicates with decision support engine 112 via a network (e.g., the Internet). In some other implementations, decision support engine 112 executes partly on one or more local devices, such as mobile device 107, and partly on one or more computing devices in a private or public cloud. In some other implementations, decision support engine 112 executes entirely on one or more local devices, such as mobile device 107.

[0042] In some implementations, DAM 113 is configured to process a set of inputs received from application 106 and compute multiple metrics 130, which can then be stored in user profile 116. Application 106 can use the inputs 127 and metrics 130, such as different characteristics of application 106, to provide real-time guidance and treatment to the user. The various data points of user profile 116 are described in more detail below. In some implementations, the user profile, which includes user profile 116, is stored in user database 110, which is accessible by application 106 and decision support engine 112 via one or more networks (not shown). In some implementations, user database 110 refers to a storage server that can operate in a public or private cloud.

[0043] The real-time guidance and treatment provided to users helps improve their physiological condition and / or enables them to make more informed decisions. To provide effective, relevant, and timely guidance and treatment to users, in some embodiments, features of application 106 may take as input information about the user stored in user profile 116 and / or information related to a pool of similar users stored in user profiles of these users in user database 110. In some embodiments, features of application 106 may interact with users in various ways, such as text, email, notifications (e.g., push notifications), telephone, and / or other forms of communication such as displaying content (e.g., charts, trends, graphs, etc.) on the user interface of application 106.

[0044] As described above, in some embodiments, application 106 is configured to take user-related information as input and store that information in the user's user profile 116. For example, application 106 may obtain and record the user's demographic information 118, disease progression information 120, and / or medication information 122 in the user profile 116. In some embodiments, demographic information 118 may include one or more of the user's age, BMI (body mass index), ethnicity, sex, etc. In some embodiments, disease progression information 120 may include information about the user's disease, such as whether the user has type I, type II, or prediabetes, or whether the user has gestational diabetes. In some embodiments, information about the user's disease may also include the length of time since diagnosis, the level of diabetes control, adherence to diabetes management treatment, predicted pancreatic function, other types of diagnoses (e.g., heart disease, obesity) or health measurements (e.g., heart rate, exercise, stress, sleep, etc.) and / or similar content. In some embodiments, medication regimen information 122 may include information about the amount and type of insulin or non-insulin diabetes medication and / or non-diabetes medication taken by the user. In some implementations, application 106 may obtain demographic information 118, disease progression information 120, and / or medication information 122 from the user in the form of user input or from other sources. In some implementations, application 106 may receive updates from the user or other sources as some of this information changes.

[0045] In some implementations, in addition to the user's demographic information 118, disease progression information 120, and / or medication information 122, application 106 receives an additional set of inputs 127, which are also utilized by various features of application 106 to provide guidance to the user. In some implementations, this input 127 is obtained on a continuous basis. In some implementations, application 106 receives input 127 through user input and / or multiple other sources, including glucose monitoring system 104, other applications running on mobile device 107, such as menstrual cycle management applications, and / or one or more other sensors and devices. In some implementations, such sensors and devices include one or more of the following: insulin management devices, other types of analyte sensors, sensors or devices provided by mobile device 107 (e.g., accelerometers, cameras, GPS, heart rate monitors, etc.) or other user accessories (e.g., smartwatches), or any other sensors or devices that provide relevant information about the user.

[0046] In some implementations, application 106 further uses at least a portion of input 127 to obtain multiple metrics, such as metric 130, which are also stored in user profile 116. (As per the...) Figure 2 As further described, in some embodiments, application 106 transmits at least a portion of input 127 to DAM 113 for processing, based on which DAM 113 generates indicator 130. In some embodiments, indicator 130 can then be used by application 106 as input for providing guidance to the user. Note that in some embodiments, user profiles 116 and user profiles in the user pool of user database 110 are dynamic because the information in the user profiles (including user profile 116) can change when new input 127 is received periodically or when existing information in the user profiles is changed (e.g., user medication information changes, etc.).

[0047] As about Figure 2 As further described herein, in some embodiments, at least in certain circumstances, indicator 130 may generally indicate one or more of the user's current or future health status or state, such as the user's physiological state (e.g., glucose levels, insulin resistance, etc.) or psychological state (e.g., stress levels, well-being, etc.), trends associated with the user's health or state, etc. For example, indicator 130 may include one or more indicators associated with metabolic rate, glucose levels and trends, user health or disease, etc. Indicator 130 may also include behavioral indicators that may indicate user behavior and habits, such as dietary habits, exercise programs, etc. In some embodiments, indicator 130 may include real-time indicators, past indicators, and / or trends.

[0048] While not limited to this list, some exemplary features of application 106 may include one or more of the following: reporting features, intervention features, medication reminder features, blood glucose effect estimator features, educational features, etc.

[0049] In some implementations, the reporting feature can provide reports to the user in various forms. For example, the reporting feature can be configured to isolate blood glucose fluctuations and report them to the user so that the user can reflect on the cause of such fluctuations. In some implementations, the reporting feature can provide reports that incorporate the user's glucose, insulin, and menstrual cycle information. For example, the report can indicate how the user's glucose levels and insulin resistance change relative to different phases of the user's menstrual cycle. In some implementations, the report (e.g., in real-time) can remind the user of the phase in their menstrual cycle and indicate that the blood glucose fluctuations the user is currently experiencing are due to this factor. In some implementations, the report can remind the user about the phase of their menstrual cycle and further warn or notify the user, for example, that they will experience blood glucose fluctuations within 2 days because they will enter the luteal phase within 2 days.

[0050] Furthermore, as those skilled in the art will understand, reports can be provided to users in various forms. For example, a report feature could display a graph of the user's blood glucose measurement trends, with two blood glucose fluctuations highlighted by text such as, "You had two hyperglycemic events today; the first lasted 55 minutes, and the second lasted 30 minutes." In one embodiment, the graph of the user's glucose measurement trends or predictions could be supplemented with a timeline of the user's menstrual cycle to show, for example, why the user is experiencing higher than usual glucose levels, why the user will experience higher than usual blood glucose levels on some future days (e.g., if the user does not take or participate in any treatment), and so on. Combining the user's past, current, and / or predicted physiological statistics with information about the user's menstrual cycle enables the user to make informed decisions about the available treatment options that the DSTA 100 may offer.

[0051] In one embodiment, the reporting feature may provide an afternoon report that provides information on the user's average blood glucose levels so far that day, indicates the average for the remaining time of day needed to track the user's target glucose range, and suggests low glycemic loads or physical activity. In another embodiment, the reporting feature may provide an evening summary that includes daily, weekly, and monthly blood glucose averages and an estimated A1c. For example, this summary informs the user what their average blood glucose level for the following day must be to reach their A1c target.

[0052] In some implementations, reporting features may focus on providing teachable moments to users. In some implementations, teachable moments identify the impact of behaviors (e.g., physical activity, diet, medication adherence, and / or sleep) on blood glucose levels. Teachable moments may be pushed to users as notifications and / or recorded on glucose monitoring curves for timely viewing. For example, teachable moments may be visually displayed on the curve and describe the behavior and glucose response to inform users what behavior led to what glucose response.

[0053] An example of a positive teaching moment can be provided when a user eats a high-glycemic load breakfast around 10 a.m., which would raise their blood sugar. In some implementations, based on accelerometer input from the user, the reporting feature can then determine that the user has walked for 30 minutes, causing their blood sugar to return to their target range. In some implementations, the reporting feature marks this event as a teaching moment (e.g., by displaying an asterisk on the CGM curve) and can send a notification to the user saying, “Great job, Sharon! Your blood sugar is now back to your desired level thanks to your 30-minute walk.”

[0054] In some implementations, intervention features include any features designed to alter user actions, such as by encouraging or preventing the user from engaging in an action. As an example, intervention features may be configured to send push notifications to the user to encourage them to engage in an action. Intervention features may also be able to determine, for example, based on the user's past actions, that the user is about to engage in an action and send a push notification to the user requesting them not to engage in such an action. In some implementations, pushing notifications to the user may be based on user-related information (e.g., input 127, indicator 130, etc.) and / or information related to the user's hierarchical group.

[0055] For example, one type of intervention feature might involve exercise management. Examples include exercise management features that encourage users to exercise, for instance, by sending push notifications, when it is determined that a user's glucose levels are high, will become high, or are at a level where lack of exercise might lead to increased glucose levels. As an example, based on the user's stage in their menstrual cycle and information associated with the user and / or with a group of similar users, the exercise management feature could suggest that the user exercise for an additional two hours over the next few days as they are beginning to enter the luteal phase.

[0056] In some implementations, exercise management features can be configured to receive and analyze data from accelerometers, GPS, heart rate monitoring sensors, glucose monitoring systems 104, and / or other types of sensors and devices to provide users with more effective and personalized guidance. For example, by receiving information from one or more of these sensors and devices, exercise management features may be able to determine whether a user is actually engaging in exercise, how long the user should exercise to ensure that the user's blood sugar returns to a normal range, what walking route the user should take, etc.

[0057] Another type of intervention feature may involve dietary management. For example, a dietary management feature could act as a virtual nutritionist, providing users with guidance on one or more of when to eat, what to eat, and how much to eat. In some implementations, a dietary management feature could provide personalized dietary recommendations based on one or more of the user's real-time conditions (e.g., real-time blood glucose measurements), the user's body's response to certain diets, etc. In some implementations, a dietary management feature could also assist users with meal preparation and / or shopping and / or allow users to input information about the nutritional value of the meals they consume. In some implementations, a dietary management feature could further suggest menus and ingredient alternatives at restaurants, or recommend healthy restaurants and grocery stores within a specific geographic area. In some implementations, based on the user's stage in their menstrual cycle and information associated with the user and / or with a group of similar users, for example, a dietary management feature could advise the user to change their diet over the next few days as the user begins to enter the luteal phase. For example, a dietary management feature could calculate the portion or amount of different types of food (e.g., carbohydrates, sweets, etc.) that the user can consume and that are still within the target glucose range. Dietary management features can also remind users to limit or exceed certain types of food intake, as users are or will be insulin resistant at a particular time.

[0058] In some implementations, dietary management features can also provide notifications based on information about the user's diet. For example, if a user has eaten and their blood glucose has not subsequently returned to the target range (e.g., the next pre-meal spike (2 hours after the initial meal-related glucose spike) exceeds 180 mg / dL), an emergency alert can then be issued to the user to immediately begin exercise. However, if the user's most recent pre-meal spike was below 180 mg / dL and their pre-meal glucose was within the range (80-130), then the dietary management feature might be randomized (e.g., reducing the likelihood of sending an alert by 33%) if the user receives an alert after the next meal. Note that the exercise and dietary management features described above are merely two examples of intervention features.

[0059] In some implementations, medication management features can provide users with notifications about when they need to take medication, what type of medication they should take (e.g., oral medications for type 2 diabetes and insulin injections for type 1 diabetes), and at what dose or amount. Medication management features can provide such notifications based on specific user information, such as the user's disease (e.g., type 1 or type 2), current and predicted indicators (e.g., current blood levels and indicators), the user's menstrual cycle information, information related to similar user groups, etc. For example, medication management features can be configured to understand the user's insulin resistance level on certain days of their menstrual cycle over a certain period of time (e.g., several months) and how much basal insulin needs to be administered to the user on or in preparation for those days. As a more specific embodiment, medication management features can determine that it is currently two days before the user enters the luteal phase and instruct the user to either administer twice the amount of basal insulin to prepare for insulin resistance or ask the user if they can signal their insulin pump to automatically begin administering twice the amount of basal insulin. In another embodiment, the medication management feature can detect when a user begins to develop insulin resistance and recommend administering twice the amount of basal insulin that the user would normally administer on days when the user does not have insulin resistance.

[0060] In some implementations, based on a user's historical information, medication management features can determine that an additional dose of basal insulin administered during a phase of the user's menstrual cycle (e.g., the luteal phase) has successfully maintained the user's glucose levels within a range. In such embodiments, medication management features can recommend additional doses to the user during periods of increased insulin resistance due to hormonal changes during the menstrual cycle or when an increase in insulin resistance is anticipated, or automatically instruct the insulin pump to administer additional doses. In some implementations, medication management features can determine how much additional insulin should be administered to the user based on effectiveness on a group of similar users.

[0061] In some implementations, the learning process involved in determining the appropriate amount of additional insulin to be administered to counteract changes (e.g., increases) in insulin resistance due to the user's menstrual cycle involves examining the user's historical information, such as patterns related to the user's insulin resistance levels on certain dates, the amount of insulin typically required to restore the user's glucose levels to within range, the time the user remained within range based on the amount of insulin they took on certain dates, and / or similar information associated with one or more similar users. For example, in some implementations, the medication management feature might learn that, in month X, on certain cycle-related dates, it is recommended that the user take 1.5 times the amount of basal insulin the user usually takes, but the user's glucose levels have not returned to within range, or the user has been outside the range for a longer period than expected. Based on this, in some implementations, the application can recalculate the dosage and recommend that the user take 1.7 times the amount of basal insulin the user usually takes during the same period in month X+1. Based on the pattern of insulin dosage recommendations and the user's physiological response (e.g., time within the range), medication management characteristics can ultimately learn (e.g., via DAM 113, as described below) to recommend an appropriate amount of insulin to the user, depending on the stage of the user's menstrual cycle.

[0062] In some implementations, medication management features can also track whether a user adheres to their medication schedule, automatically order medication before the user runs out, and / or provide information about the medication itself (e.g., educating the user about the medication's effects and efficacy). For example, a medication management feature might ask a patient if they have taken their medication. If the user answers "yes" three times consecutively, the feature might randomize the process and allocate a 33% chance to ask the user again the following day. If the user does not answer "yes" three times consecutively, the feature might send a reminder to the user to take their medication the next day.

[0063] In some embodiments, as described above, the medication management feature can automatically communicate with the insulin delivery device (e.g., a pump or pen) based on the user's current or predicted indicators 130 to enable the device to administer the correct dose of insulin. For example, the medication management feature can take into account the user's current glucose level and indicators as well as the user's predicted glucose level and indicators, and determine the precise amount of long-acting insulin to be administered. In some embodiments, the medication management feature can signal the drug delivery device to administer the amount of insulin. As described above, in some embodiments, the medication management feature takes into account user-related information and / or information related to the user's stratification group to calculate the amount of insulin to be administered. Exemplary details relating to how the DSTA 100 (including application 106) is able to set the insulin rate of the insulin delivery device (also referred to as a drug delivery device) are described in paragraphs

[0425] -

[0426] of U.S. Patent Application Publication 2019 / 0246973, the entire contents of which are incorporated herein by reference.

[0064] In some implementations, the glycemic impact estimator feature can use the camera of mobile device 107 to scan the menu and convert each meal item into an estimated glycemic impact metric. In some implementations, the glycemic impact estimator feature displays the glycemic impact metric overlaid on the menu item. In some implementations, the glycemic impact metric is based on data from user profiles from hierarchical groups of users. In some implementations, the glycemic impact estimator feature can highlight different menu items based on their health level using different colors (e.g., green for healthier items, red for unhealthier items).

[0065] In some implementations, the educational feature educates the user about their condition and how they can improve their health. In one example, the educational feature informs the user about the potential impacts they might see if they adopt a certain lifestyle. In some implementations, to determine potential impacts, the educational feature may consider the effects experienced by other users in the stratification group who adopt the same lifestyle. For example, the educational feature might tell the user, “By following this procedure, patients like you are able to reduce A1C by 5% during the luteal phase.” In some implementations, the educational feature may also educate the user about the reasons for and effects of potential behaviors, such as based on the effects experienced by users in the stratification group.

[0066] In some implementations, the educational feature provides alternatives to administering insulin or higher-than-usual amounts of insulin during certain periods of a user's menstrual cycle. For example, the educational feature might examine information related to a user's stratification group and discover that some users in that group are able to maintain their glucose levels within the target range by exercising for an extra hour during the luteal phase without taking any higher-than-usual amounts of insulin. In such an implementation, the educational feature can inform the user of this finding. Note that based on this finding, the exercise management feature may also suggest an extra hour of exercise during the luteal phase and recommend additional exercise types and intensities based on the user's habits. In another implementation, the educational feature might examine information related to a user's stratification group and discover that some users in that group are able to maintain their glucose levels within the target range by consuming less of certain foods during the luteal phase without taking any higher-than-usual amounts of insulin. Similarly, in such an implementation, the educational feature can inform the user of this finding, and the diet management function can specifically recommend how the user can change their diet, what parts of their diet to eat, etc., based on the same finding.

[0067] Figure 1B The glucose monitoring system 104 is illustrated in more detail. Figure 1B Several mobile devices 107a, 107b, 107c, and 107d are also illustrated. Note Figure 1A The mobile device 107 can be any one of mobile devices 107a, 107b, 107c, or 107d. In other words, any one of mobile devices 107a, 107b, 107c, or 107d can be configured to execute application 106. The glucose monitoring system 104 can be communicatively coupled to mobile devices 107a, 107b, 107c, and / or 107d. The glucose monitoring system 104 can also be communicatively coupled to an insulin delivery device (not shown), which can also be placed on a user's body to administer insulin to the user's body.

[0068] By way of overview and embodiments, glucose monitoring system 104 can be implemented as a packaged microcontroller that generates sensor measurements, generates analyte data (e.g., by calculating values ​​from continuous glucose monitoring data), and participates in wireless communication (e.g., via Bluetooth and / or other wireless protocols) to transmit such data to remote devices, such as mobile devices 107a, 107b, 107c, and / or 107d. Paragraphs

[0137] -

[0140] of U.S. Patent No. 2019 / 0336053 and Figure 3 A, 3B, and 4 further describe a skin sensor assembly that, in some embodiments, can be used in conjunction with a glucose monitoring system 104. Paragraphs

[0137] -

[0140] of U.S. Patent No. 2019 / 0336053 and Figure 3A, 3B and 4 are incorporated into this paper by reference.

[0069] In some embodiments, the glucose monitoring system 104 includes an analyte sensor electronics module 138 and a glucose sensor 140 associated with the analyte sensor electronics module 138. In some embodiments, the analyte sensor electronics module 138 includes electronic circuitry associated with measuring and processing analyte sensor data or information, including algorithms associated with the processing and / or calibration of the analyte sensor data / information. The analyte sensor electronics module 138 may be physically / mechanically connected to the glucose sensor 140 and may be integrated into (i.e., non-releasably attached to) or releasably attached to the glucose sensor 140.

[0070] The analyte sensor electronics module 138 can also be electrically coupled to the glucose sensor 140, allowing the components to be electrically coupled to each other. The analyte sensor electronics module 138 may include hardware, firmware, and / or software capable of measuring and / or estimating analyte levels in the user's body via the glucose sensor 140 (e.g., may be / include a glucose sensor). For example, the analyte sensor electronics module 138 may include one or more potentiostats, a power supply for providing power to the glucose sensor 140, other components for signal processing and data storage, and a telemetry module for transmitting data from the sensor electronics module to one or more display devices. The electronic equipment may be mounted on a printed circuit board (PCB) within the glucose monitoring system 104 or platform, and can take various forms. For example, the electronic equipment may take the form of integrated circuits (ICs), such as application-specific integrated circuits (ASICs), microcontrollers, processors, and / or state machines.

[0071] The analyte sensor electronics module 138 may include sensor electronics configured to process sensor information, such as sensor data, and generate converted sensor data and displayable sensor information. Embodiments of systems and methods for processing sensor analyte data are described in detail herein and in U.S. Patent Nos. 7,310,544 and 6,931,327 and U.S. Patent Publications 2005 / 0043598, 2007 / 0032706, 2007 / 0016381, 2008 / 0033254, 2005 / 0203360, 2005 / 0154271, 2005 / 0192557, 2006 / 0222566, 2007 / 0203966, and 2007 / 0208245, all of which are incorporated herein by reference in their entirety.

[0072] Glucose sensor 140 is configured to measure the concentration or level of an analyte in user 102. The term analyte is further defined by paragraph

[0117] of U.S. Application No. 2019 / 0336053. Paragraph

[0117] of U.S. Application No. 2019 / 0336053 is incorporated herein by reference. In some embodiments, glucose sensor 140 includes a continuous glucose sensor, such as a subcutaneous, transdermal (e.g., percutaneous), or intravascular device. In some embodiments, glucose sensor 140 can analyze multiple intermittent blood samples. Glucose sensor 140 can be used with any glucose measurement method, including enzymatic, chemical, physical, electrochemical, spectrophotometric, optical rotation, calorimetric, iontophoresis, radiation, immunochemistry, and the like. Additional details relating to continuous glucose sensors are provided in paragraphs

[0072] -

[0076] of U.S. Application No. 13 / 827,577. U.S. Patent Application No. 13 / 827,577, pp.

[0072] -

[0076] , are incorporated herein by reference.

[0073] Further reference Figure 1B Mobile devices 107a, 107b, 107c, and / or 107d may be configured to display (and / or alarm) displayable sensor information (e.g., transmitted in custom data packets to the display device according to respective preferences) that can be transmitted by the sensor electronics module 138. Each of mobile devices 107a, 107b, 107c, and / or 107d may respectively include a display such as touchscreen displays 109a, 109b, 109c, and / or 109d for displaying the graphical user interface of application 106 to present sensor information and / or analyte data to user 102 and / or receive input from user 102. In some embodiments, the mobile device may include other types of user interfaces, such as a voice user interface, alternative to or in addition to a touchscreen display, for transmitting sensor information to user 102 of the mobile device and / or receiving user input. In some implementations, one, some, or all of the mobile devices 107a, 107b, 107c, and / or 107d may be configured to display or otherwise transmit sensor information (e.g., in a data packet transmitted to a corresponding display device) as sensor information is transmitted from the sensor electronics module 138, without requiring any additional processing as is expected for calibration and / or real-time display of sensor data.

[0074] Figure 1BThe multiple mobile devices 107a, 107b, 107c and / or 107d depicted may include custom or proprietary display devices, such as analyte display device 107b, specifically designed to display a particular type of displayable sensor information (e.g., numerical values ​​and / or arrows in some embodiments) associated with analyte data received from sensor electronics module 138. In some embodiments, one of the multiple mobile devices 107a, 107b, 107c and / or 107d includes a smartphone, such as mobile phone 107c, based on Android, iOS, or another operating system configured to display a graphical representation of continuous sensor data (e.g., including current and / or historical data).

[0075] Figure 2 According to certain implementations, exemplary inputs and exemplary metrics determined based on inputs are described in more detail. Figure 2 An exemplary input 127 on the left, an application 106 and a DAM 113 in the middle, and an indicator 130 on the right are shown. In some implementations, each indicator may correspond to one or more values, such as discrete numerical values, ranges, or qualitative values ​​(high / medium / low or stable / unstable). The application 106 obtains the input 127 through one or more channels (e.g., manual user input, sensors, other applications executing on the mobile device 107, etc.). In some implementations, features of the application 106 may use the input 127 to provide guidance and treatment to the user (e.g., it may include signaling an insulin pump to administer a dose of insulin). The input 127 may also be further processed by the DAM 113 to output multiple indicators, such as indicator 130, which the features of the application 106 may similarly use to provide guidance and treatment to the user.

[0076] As shown in the figure, input 127 includes, but is not limited to, food intake information, activity information, patient statistics, insulin information, information from sensors, blood glucose information, time, calendar, user input, menstrual cycle information, etc.

[0077] Food intake information may include information about one or more of meals, snacks, and / or beverages, such as size, content (carbohydrates, fat, protein, etc.), order of intake, and time of intake. In some embodiments, food intake information may be provided by manual input by the user, by providing photos through an application configured to identify food types and quantities, and / or by scanning barcodes or menus. In various embodiments, meal size may be manually entered as one or more of calories, quantity ('three cookies'), menu item ('royal cheese'), and / or food exchange portion (1 piece of fruit, 1 dairy product). In some embodiments, meals typical of the user at that time or in that environment may also be entered (e.g., breakfast at home on weekdays, brunch at a restaurant on weekends). In some embodiments, meal information may be received through a convenient user interface provided by application 106.

[0078] In some implementations, activity information is also provided as input. For example, activity information may be provided by an accelerometer sensor on a wearable device (e.g., a watch, fitness tracker, and / or patch). In some implementations, activity information may also be provided via manual user input.

[0079] In some implementations, patient statistics may also be provided, such as age, height, weight, body mass index, body composition (e.g., body fat percentage), height, body type, or one or more other information. In some implementations, patient statistics are provided via a user interface, an interface with an electronic source such as an electronic medical record, and / or from the measuring device. In some implementations, and in some embodiments, the measuring device includes one or more wireless devices, such as those supporting Bluetooth, a weighing scale, and / or a camera, which may communicate with the mobile device 107 to provide patient data.

[0080] In some implementations, input related to insulin delivery to the patient can be received via a wireless connection on a smart pen, through user input, and / or from an insulin pump (an insulin delivery device). Insulin delivery information may include one or more of the following: insulin volume, delivery time, etc. Other parameters, such as insulin action time or duration of insulin action, may also be received as input.

[0081] In some implementations, input can also be received from sensors, such as physiological sensors, which can detect one or more of heart rate, respiration, oxygen saturation, or body temperature (e.g., detecting disease). In some implementations, electromagnetic sensors can also detect low-power radio frequency fields emitted from objects or tools that are in contact with or near the patient, which can provide information about the patient's activity or location. One embodiment of information that can be received from the sensors is the user's blood glucose level.

[0082] In some embodiments, blood glucose information may also be provided as input, for example, via a glucose monitoring system 104. Blood glucose information may include any glucose-related measurements known in the art. In some embodiments, blood glucose information may be received from one or more smart medication dispensers that track when a user takes medication, a blood ketometer, a laboratory measurement or estimated ALC, other long-term controlled measurements, or sensors that use tactile responses to measure peripheral neuropathy (e.g., using the tactile features of a smartphone or specialized device).

[0083] In some implementations, time may also be provided as input, such as the time of day or time from a real-time clock. Time may also include date, month, and year.

[0084] User input via a user interface (such as the user interface of mobile device 107) can include any other type of input that the user can provide to application 106, such as the other types of input described above. For example, in some embodiments, user input may include one or more types of food intake, delivery of treatment (such as using glucagon to stimulate the liver to release glycogen in response to hypoglycemia), recommended basal rate or insulin-to-carbohydrate ratio (e.g., received from a clinician), recorded activities (e.g., intensity, duration, and time of completion or start), etc. In some embodiments, user input may also indicate medication intake (e.g., type and dosage of medication and time of administration).

[0085] In some implementations, input 127 may also include information about the user's menstrual cycle. As mentioned above, this information may be received from a third-party application, such as a cycle tracking application. In some implementations, the third-party application may interact with application 106 periodically via, for example, an application programming interface (API). The information about the user's menstrual cycle may include one or more of the following: different phases of the user's cycle, the duration of each phase, the corresponding dates, the predicted dates when the user will enter each phase, which phase the user is currently in, and when the current phase is expected to end, etc. In some implementations, the user may supplement the information provided by the third-party application with user input. For example, the user may provide input such as "my menstrual phase has just begun." In some implementations, application 106 may rely solely on user input, without depending on any input from the third-party application. In such implementations, for example, application 106 may calculate information about the user's menstrual cycle based on user input and / or some science-based logic.

[0086] As described above, in some implementations, DAM 113 determines or calculates the user's metrics 130 based on input 127. Figure 2An exemplary list of indicators 130 is shown in the figure.

[0087] In some embodiments, metabolic rate is an indicator that may indicate or include basal metabolic rate (e.g., energy consumed at rest) and / or activity metabolism (e.g., energy consumed during activities such as exercise or exertion). In some embodiments, basal metabolic rate and activity metabolism may be tracked as separate indicators. In some embodiments, metabolic rate may be calculated by DAM 113 based on one or more of inputs 127, such as activity information, sensor inputs, time, user inputs, etc.

[0088] In some implementations, the activity level index can indicate the user's activity level. In some implementations, the activity level index is determined, for example, based on input from an activity sensor or other physiological sensor. In some implementations, the activity level index can be calculated by DAM 113 based on one or more of inputs 127, such as activity information, sensor input, time, user input, etc.

[0089] In some implementations, historical data, real-time data, or a combination thereof may be used, and insulin sensitivity metrics may be determined, for example, based on one or more of one or more inputs 127, such as food consumption information, blood glucose information, insulin delivery information, obtained glucose levels, etc. In some implementations, insulin delivery information and / or known or acquired (e.g., from patient data) insulin time-of-action curves may be used to determine active insulin board metrics, which can explain basal metabolic rate (e.g., insulin renewal to maintain bodily functions) and insulin use driven by activity or food intake.

[0090] In some implementations, the meal status indicator can indicate the user's state in terms of food consumption. For example, meal status can indicate whether the user is in a fasting state, a pre-meal state, an eating state, a post-meal response state, or a steady state. In some implementations, meal status can also indicate nourishment on board, such as ingested meals, snacks, or beverages, and can be determined, for example, based on food intake information, meal timing information, and / or digestibility information, which can be related to food type, quantity, and / or order (e.g., which food / beverage was eaten first).

[0091] In some implementations, health and disease indicators can be determined, for example, based on one or more user inputs (e.g., pregnancy information or known disease information) from physiological sensors (such as temperature), activity sensors, or combinations thereof. In some implementations, based on the values ​​of health and disease indicators, for example, a user's state can be defined as one or more of healthy, sick, resting, or fatigued.

[0092] In some implementations, glucose level indicators can be determined from sensor information (e.g., blood glucose information obtained from glucose monitoring system 104). In some implementations, glucose level indicators can also be determined, for example, based on historical information about glucose levels under specific circumstances, such as given a combination of food intake, insulin, and / or activity.

[0093] In some embodiments, indicator 130 may also include a disease stage, such as a patient with type 2 diabetes. Exemplary disease stages for patients with type 2 diabetes may include a prediabetes stage, an oral treatment stage, and a basal insulin treatment stage. In some embodiments, the degree of glycemic control (not shown) may also be used as an indicator and may be based on one or more of, for example, glucose levels, changes in glucose levels, or insulin administration patterns.

[0094] In some implementations, clinical indicators typically indicate a user's clinical status relative to one or more medical conditions (such as diabetes). For example, in the case of diabetes, clinical indicators may be determined based on blood glucose measurements and include A1c, A1c trends, time within range, time spent below a threshold level, time spent above a threshold level, and / or one or more other indicators derived from blood glucose values. In some implementations, clinical indicators may also include one or more of estimated A1c, blood glucose variability, hypoglycemia, and / or health indicators (the amount of time spent outside a target region).

[0095] In some embodiments, indicator 130 may also include behavioral indicators such as dietary habits, disease treatment adherence, medication type and adherence, exercise programs, etc. As further described below, in some embodiments, DAM 113 may use historical records of user behavioral indicators to develop trends, based on which future user behavior can be predicted. In some embodiments, eating habits are measured by one or more indicators based on the content and timing of a user's meals. For example, in an embodiment, if the eating habit indicator is in the range of 0 to 1, then the better / healthier the user eats, the higher their eating habit indicator value will be, which will be 1. Furthermore, in an embodiment, the more adherent a user's food intake is to a specific schedule, the closer their eating habit indicator will be to 1. In some embodiments, disease treatment and adherence are measured by one or more indicators that indicate the degree to which a user follows the treatment of their disease.

[0096] In some implementations, disease treatment and adherence indicators are calculated based on one or more of the following: the user's diet or food intake, exercise program, medication adherence, etc. In some implementations, medication adherence is measured by one or more indicators that demonstrate the degree to which the user follows their medication regimen. In some implementations, medication adherence indicators are calculated based on one or more of the following: the time the user takes the medication (e.g., whether the user takes it on time or as scheduled), the type of medication (e.g., whether the user takes the correct medication), and the dosage of medication (e.g., whether the user takes the correct dosage).

[0097] In some implementations, the exercise program is measured by one or more metrics that indicate one or more of the following: the type of activity the user participates in, the intensity of the activity, the frequency of the user's participation in such activities, etc. In some implementations, the exercise program metrics may be calculated based on one or more activity sensors, calendar input, user input, etc.

[0098] In some implementations, input 127 and indicator 130 are timestamped to create a historical data record for the user. Based on this historical data record, DAM 113 is able to determine correlations between different inputs 127, between different indicators 130, and / or between input 127 and indicator 130. For example, based on historical data, the correlation between the user's menstrual cycle and the user's blood glucose levels at different stages of the menstrual cycle can be determined. As those skilled in the art will understand, correlations between various inputs 127 and / or indicators 130 can be similarly determined. For example, the correlation between administered insulin dose, blood glucose levels, and different stages of the menstrual cycle can be determined. In some implementations, based on the historical data record and / or the determined correlations, DAM 113 is then able to predict changes in insulin resistance and blood glucose levels during different stages of the user's menstrual cycle, and the predictive effects of different doses of basal and bolus insulin, exercise, and food intake on glucose levels at different stages of the menstrual cycle. In some implementations, based on such predictions, different features of application 106 can recommend different treatments, such as different doses of basal and / or bolus insulin, certain types and / or durations of exercise, and certain types and / or portions of food.

[0099] Figure 3 This is a flowchart illustrating an exemplary operation 300 performed by a system (e.g., DSTA 100) for providing one or more treatments to a user based on a phase of her menstrual cycle. One or more treatments are determined based on information about the user and / or information about a pool of users, such as a stratified group of users similar to the user in one or more ways. Operation 300 is referenced below. Figure 1A-1B The steps of operation 300 and its components are described below. Note that the steps of operation 300 do not necessarily have to be performed in the order described here. In addition, some steps may be omitted here, and / or additional states may be added.

[0100] In step 302, operation 300 begins with measuring the user's blood glucose. In some embodiments, step 302 is performed by a glucose monitoring system 104. In some embodiments, the glucose monitoring system 104 is a continuous glucose monitoring system, enabling it to periodically (e.g., every 5 minutes) measure the user's blood glucose levels. In some embodiments, glucose measurements are received and recorded over time by an application 106, thereby creating a record of glucose measurements. In some embodiments, the record indicates all past glucose measurements, including the most recent glucose measurement, which is considered the user's real-time glucose measurement. Note that in some embodiments, the application 106 is able to receive glucose measurements from the glucose monitoring system 104 because the user, prior to using the glucose monitoring system 104, completes a setup process with the application 106 to ensure that the application 106 and the glucose monitoring system 104 can recognize each other and communicate securely. In some embodiments, during the setup process or early stages of using the application 106, the user provides the application 106 with at least one of her demographic information 118, disease progression information 120, and medication information 122.

[0101] In some embodiments, at step 304, operation 300 then receives information about the user's menstrual cycle. In some embodiments, step 302 is performed by DAM 113. For example, as part of input 127, DAM 113 may receive information about the user's menstrual cycle. As described above, in some embodiments, this information may be received from a third-party software application that tracks the user's phases in her menstrual cycle in real time and is able to determine how long each phase lasts and when each phase begins, etc. Such a software application may be executed on mobile device 107 or on some other computing device communicating with mobile device 107 and / or DAM 113 via a network. In some embodiments, as described above, DAM 113 may receive information about the user's menstrual cycle in the form of manual user input. For example, the user may input information about when their menstruation begins and ends, as well as any additional information the user may have about their menstrual cycle.

[0102] In step 306, operation 300 then determines one or more treatments for the user to maintain or restore the user's glucose levels to a target range based on information about the user's menstrual cycle, the user's historical data, and historical data associated with a stratified group of similar users. In some embodiments, step 306 may be performed by DAM 113.

[0103] In some implementations, DAM 113 determines one or more treatments based on data points indicated or calculated (e.g., predicted) by the user's historical data and / or historical data associated with a stratified group of similar users. More specifically, in some implementations, DAM 113 may determine one or more treatments based on the user's current and / or future glucose levels and / or insulin resistance (e.g., different phases of the menstrual cycle). Furthermore, in some implementations, DAM 113 may determine one or more treatments based on past treatments (e.g., treatments that have effectively maintained or restored the user's glucose levels to a target range over the past (e.g., past days, months, years) and / or past treatments that have effectively maintained or restored the glucose levels of one or more users in a stratified group to a target range. Note that a target range may refer to or include a target glucose level (e.g., a specific measurement).

[0104] For example, a user's current glucose level may be higher than a target glucose level. In such an embodiment, DAM113 can determine one or more treatments based at least on the user's current glucose level, the difference between the current glucose level and the target glucose level, and / or the user's current insulin resistance. In this case, the user's current insulin resistance is either calculated in real time or based on the user's own historical data and / or historical data associated with a stratification group. For example, in some implementations, the user's current insulin resistance may be based on the user's insulin resistance at a stage in their menstrual cycle and / or the user's insulin resistance at the same time period in the user's past menstrual cycles.

[0105] In another embodiment, DAM 113 can identify one or more treatments to help a user maintain or reach a certain glucose level at a future point in time. In such an embodiment, DAM 113 can identify one or more treatments based at least on the user's future glucose levels and / or insulin resistance, which can be predicted based on the user's historical data records and / or the historical data records of one or more users in a stratified group. For example, the user may be two days away from entering the luteal phase. In such an embodiment, DAM 113 can predict this event, predict the user's level of insulin resistance during the luteal phase, and identify one or more treatments that have helped the user reach a certain glucose level during the luteal phase in the past. For example, in a simplified case, DAM 113 can determine that administering a specific dose of basal insulin two days before the user enters the luteal phase can effectively maintain the user's glucose levels within the target range, for example, assuming the user continues to follow the same diet and exercise regimen.

[0106] In some implementations, when a user begins using application 106, there may not be sufficient inputs 127 and indicators 130 available regarding them. Therefore, in some implementations, conclusions regarding how a user's glucose levels fluctuate and / or the degree of insulin resistance in response to hormonal changes at different stages of the menstrual cycle cannot be drawn with high confidence. Similarly, in these implementations, conclusions regarding which treatments can effectively maintain or restore a user's glucose levels at different stages of the menstrual cycle cannot be drawn with high confidence. In some implementations, in order to draw reliable conclusions based on the user's own historical data, inputs 127 and indicators 130 for at least several months (e.g., one or more months) should be available to the user, although not required. Therefore, at least initially, DAM 113 can use historical data from one or more users in user database 110 to predict how a user's glucose levels fluctuate and / or the degree of insulin resistance at different stages of the menstrual cycle, and which treatments are effective or potentially effective at these stages.

[0107] In some implementations, DAM 113 may utilize historical data from all users in user database 110 to make the aforementioned predictions. However, in other implementations, DAM 113 may first stratify user database 110 based on one or more similarity or stratification factors. In some implementations, depending on the data model and analysis used, predictions based on datasets of stratified groups of similar users may be more accurate than predictions based on datasets associated with all users in user database 110.

[0108] A user stratification group refers to a group of users in user database 110 who are similar to the user in one or more aspects. For example, when a user first starts using application 106, certain information about the user can be determined. Such information may include the user's demographic information 118, disease progression information 120, and / or medication information 122. As part of input 127, application 106 may also initially receive input about the user's menstrual information. As mentioned above, all of these inputs can be recorded in user profile 116. Therefore, DAM 113 can retrieve user profile 116 from user database 110 and select user stratification groups from user database 110 based on one or more similarities between user profile 116 and information in user profiles of the user pool in user database 110. For example, DAM 113 may use one or more similarity or stratification factors for stratification, which include additional data (e.g., input 127 and indicator 130, if available) available in user profile 116, including at least one of the user's disease progression information, medication information, demographic information, menstrual cycle information, and / or any other information.

[0109] One of several methods and approaches can be used to stratify the user database 110 based on one or more stratification factors (e.g., disease progression, medication information, demographic information, goals, menstrual cycle, input 127, indicator 130, or combinations thereof). In some implementations, DAM 113 may use one of various data filtering techniques to filter a broader user database 110 based on one or more stratification factors. For example, if a user has type 1 diabetes, DAM 113 can filter all user profiles in user database 110 that also have type 1 diabetes. In this case, the stratification group of users would include all such users. However, in some implementations, if additional stratification factors are used for stratification, additional filtering can be performed to further narrow down the user groups within the stratification groups. For example, if the stratification factors include disease progression and demographic information and the user is a woman with type 1 diabetes, DAM 113 can filter all user profiles of all female users in user database 110 who also have type 1 diabetes.

[0110] In some implementations, DAM 113 can use machine learning algorithms to hierarchically stratify the user database 110. For example, an unsupervised learning algorithm can be used to cluster all user profiles in the user database 110 and determine which cluster a user profile 116 belongs to. Unsupervised learning is a type of machine learning algorithm used to make inferences from a dataset consisting of input data with unlabeled responses. As will be understood by those skilled in the art, other types of unsupervised learning algorithms besides those focused on cluster analysis can also be used.

[0111] In some implementations, supervised learning algorithms may be used instead. Supervised learning is a machine learning task of learning a function, for example, mapping an input to an output based on an exemplary input-output pair. In some implementations, using a supervised learning algorithm, DAM 113 may be configured to classify user profile 116 by determining which category or stratification group a user belongs to, based on a machine learning model already trained using a labeled dataset. In some implementations, the labeled data already includes different categories of users classified based on one or more features (e.g., disease progression). For example, in some implementations, one category of users includes all user profiles of women aged 25-27, while another category includes all women aged 27-29. In such an embodiment, if the demographic information 118 of user profile 116 indicates that the user is 26 years old, DAM 113 selects the first category as the stratification group of the user (if the only stratification factor is age).

[0112] In some implementations, when selecting user stratification groups, DAM 113 can be configured to define a range around each stratification factor. For example, if a user is 25 years old, a range of four years can be defined to include all users in the 23-27 age range in the stratification group.

[0113] Once a group of users is selected from user database 110, whether the group corresponds to all users in user database 110 or users in a stratified group, DAM 113 uses historical data records associated with that group to predict the user's glucose level and / or insulin resistance based on the user's menstrual cycle (e.g., at different stages of the menstrual cycle).

[0114] As an example, when a user initially uses application 106, user profile 116 may indicate that the user is a 25-year-old Caucasian woman with type 1 diabetes and injects insulin. User profile 116 may also indicate details about the user's menstrual cycle. In such an embodiment, DAM 113 may stratify user database 110 based on one or more of the aforementioned stratification factors. For example, DAM 113 may stratify user database 110 to find a dataset associated with all Caucasian female users aged 24-26 who have type 1 diabetes and inject insulin. Note that DAM 113 may or may not use users' menstrual cycle information as a stratification factor.

[0115] In some implementations, DAM 113 may then examine the dataset to predict the user's glucose levels and insulin resistance at a particular point in the menstrual cycle. For example, the user may be two days away from entering her luteal phase. In such an implementation, DAM 113 makes this prediction based on the user's menstrual cycle information. Since information about how the user's insulin resistance might become during her luteal phase or certain sub-phases of the luteal phase (or other phases and / or sub-phases of the menstrual cycle) may still be unavailable, in some implementations, DAM 113 subsequently examines the dataset associated with the user's stratification group to determine the level of insulin resistance in the stratified group users at various phases and / or sub-phases of the menstrual cycle (e.g., the luteal phase).

[0116] In some implementations, a user's luteal phase (or any other phase in the menstrual cycle) can be divided into various sub-phases, each of which may correspond to a specific time range of the luteal phase. As an example, a user's luteal phase may include sub-phases such as a start sub-phase, a middle sub-phase, and a finish sub-phase. In another implementation, sub-phases may correspond to days; for example, if a user's luteal phase is typically seven days, then seven sub-phases may be defined. Sub-phases may also correspond to hours and minutes, such that, for example, the first sub-phase of the luteal phase may correspond to the first hour of the luteal phase. Because insulin resistance varies by different amounts at different times during the luteal phase, in some implementations, dividing the luteal phase into sub-phases helps to more accurately determine the status of insulin-resistant users in stratified groups within specific time periods, on which basis, the user's insulin resistance and / or glucose levels can be predicted.

[0117] When the luteal phase for all users is divided into sub-phases, DAM 113 can determine the average change in insulin resistance across the user's stratified group within the same sub-phase that the user is currently entering or will enter. For example, if a user is 2 days away from entering the luteal phase, this could mean that the user is 2 days away from entering the first sub-phase of the luteal phase, where the first sub-phase corresponds to the start phase, day one, or other agreed-upon time that DAM 113 may configure.

[0118] Therefore, in some implementations, DAM 113 examines datasets associated with stratified groups to determine the average level of insulin resistance and / or the extent of change in glucose levels for such users during the first luteal phase. Based on this information, in some implementations, DAM 113 is able to predict how much insulin resistance and / or glucose levels will change in the first subphase if the user does not receive any additional treatment (e.g., additional exercise, a stricter diet, additional doses of basal and / or bolus insulin).

[0119] Note that while some embodiments described herein depict changes in insulin resistance during the luteal phase, a user's insulin resistance may change during other phases of the menstrual cycle. Therefore, the embodiments described herein are equally applicable to other phases or sub-phases of the menstrual cycle. In other words, the luteal phase is used as an example only.

[0120] Furthermore, note that in some implementations, instead of or in addition to examining datasets associated with stratified groups, DAM113 may consider scientific data points that indicate, on average, how much a user's insulin resistance and / or glucose levels might change during certain phases and / or subphases of the user's menstrual cycle. For example, in some implementations, such scientific data points may be supported by scientific research and provided to DAM113 as datasets indicating that different characteristics (e.g., demographic information, disease progression, medication information, menstrual cycle information, etc.) might experience changes in insulin resistance and / or glucose levels during different subphases and / or phases of the menstrual cycle. For example, in this case, a rule-based approach could be used; for instance, if the user is 25 years old, the rule could indicate, based on the scientific data points, that the user's insulin resistance will increase by 30% in the first subphase of a certain phase of the user's menstrual cycle (e.g., the luteal phase), because women of this age typically experience a significant increase in insulin resistance. The scientific data points can indicate changes in insulin resistance in each subphase and / or phase of the menstrual cycle. For example, the data might indicate that the user's insulin resistance will increase by 40% in the second subphase.

[0121] In some implementations, predicting the user's level of insulin resistance and / or how much the user's glucose levels might change in the first sub-phase can help DAM 113 predict one or more treatments to help the user maintain glucose levels within a target range upon entering the first sub-phase. For example, DAM 113 may be configured with scientific data points that indicate a specific type or number of treatments that can help maintain the user's glucose levels within a specific range, even if insulin resistance changes during a specific phase or sub-phase. For example, some scientific data points might indicate that, on average, for a 25-year-old user, administering an additional 20% of basal insulin before or at some point during the first sub-phase has been shown to counteract a 30% increase in insulin resistance in a sub-phase or phase of the menstrual cycle. In another embodiment, scientific data points might indicate that, in a sub-phase or phase of the menstrual cycle, an additional hour of exercise per day has been shown to effectively counteract a 10% increase in insulin resistance. In yet another embodiment, scientific data points might indicate that following a stricter diet daily (e.g., fewer carbohydrates, etc.) has been shown to effectively counteract a 15% increase in insulin resistance in a sub-phase or phase of the menstrual cycle.

[0122] Please note that these are simplified embodiments used to illustrate how data points based on scientific research can be used to recommend treatment to a user. Therefore, these simplified embodiments are not intended to limit the scope of this disclosure. In some embodiments, when using scientific data points, DAM 113 may be configured with a rule-based approach; for example, if a user's insulin resistance increases by 20% during a first sub-phase, and the scientific data points indicate that an additional dose of insulin can offset a 10% increase in insulin resistance, DAM 113 may be configured to recommend a double dose to the user. Conversely, in another embodiment, DAM 113 may be configured to recommend the same additional dose to offset a 10% increase and an additional hour of daily exercise to offset another 10% increase in insulin resistance, etc. Therefore, combinations can also be used to recommend treatments (e.g., insulin, dietary restrictions, exercise).

[0123] In some implementations, instead of determining one or more treatments based on examining the aforementioned scientific data points, DAM 113 can predict, based on a dataset associated with a stratification group, that one or more treatments may be effective in counteracting changes in insulin resistance during a sub-phase or phase of a user's menstrual cycle. Note that, as stated above, in some implementations, the dataset may correspond to the entire user pool rather than a stratified pool of users. In other words, predictions of effective treatments can be made similarly based on data associated with all users; therefore, it is not necessary to stratify the user database 110. Those skilled in the art will understand the various methods and operations that can be used to make such predictions based on datasets associated with user stratification groups.

[0124] For example, in some implementations, one or more data models, machine learning models, regression models, functions, and algorithms may be used to predict one or more treatments, or combinations thereof, to counteract a certain amount of change in insulin resistance during a sub-phase or phase of the menstrual cycle. Figure 4 Examples of some different algorithms that can be used are described in the document.

[0125] In some implementations, these data models, machine learning models, regression models, functions, or algorithms can be used to discover different variables (e.g., data features, such as those related to...). Figure 4 (More detailed description). In one embodiment, a variable may be defined based on a menstrual subphase or phase. Another variable may be glucose levels or changes therein. Many other variables may also be defined, including variables associated with insulin dosage, variables associated with exercise intensity and / or type, and variables associated with the type and / or quantity of food ingested.

[0126] In some implementations, based on these correlations, DAM 113 may be able to identify consistent patterns, based on which DAM 113 can be configured to predict effective treatment for the user. In some implementations, these patterns show the effects of different types / quantities of medications (e.g., insulin) and types / quantities of user behaviors (e.g., exercise, food) on the user's glucose levels / menstrual cycle at different sub-phases.

[0127] For example, a dataset might show that, on average, for female users aged 24–26, an additional 20% of basal insulin (e.g., an amount at some point before the first subphase or at some point during the first subphase) offsets a 30% increase in insulin resistance during the first subphase of the luteal phase. In another embodiment, a dataset might show that for Caucasian female users with type 1 diabetes, an additional hour of exercise per day during a subphase or phase of the menstrual cycle offsets a 10% increase in insulin resistance. In yet another embodiment, a dataset might show that for female users with the same menstrual cycle as the user, following a stricter diet daily (e.g., fewer carbohydrates, etc.) can offset a 15% increase in insulin resistance during a subphase or phase of the menstrual cycle. Note that these are simplified embodiments used to illustrate how to recommend treatment to users using a dataset associated with a group of users. Therefore, these simplified embodiments are not intended to limit the scope of this disclosure. Based on these consistent patterns, DAM 113 can predict that one or more treatments may be helpful to the user.

[0128] In some implementations, as additional inputs 127 and indicators 130 are received and stored in the user profile 116 over time, a dataset associated with the user's own historical data can be developed. Based on this dataset, DAM 113 can determine one or more treatments to help the user maintain or achieve a certain glucose range depending on the user's position in the menstrual cycle. For example, additional inputs 127 and indicators 130 may include information about different treatments offered to the user based on datasets associated with the user's stratified group (or the entire user group in user database 110) and / or scientific data points. Additional inputs 127 and indicators 130 may also include information about the effects of such treatments on the user, such as effects on the user's physiology (e.g., glucose levels, etc.).

[0129] In some implementations, a user's own historical data can more accurately predict what a user's glucose levels and / or insulin resistance are or will be during a specific sub-phase or phase of the menstrual cycle, and what one or more treatments will be effective during that sub-phase or phase, even with more precise predictions. Those skilled in the art will understand the various methods and operations that can be used to make such predictions based on datasets associated with the user themselves.

[0130] For example, in some implementations, one or more data models, machine learning models, regression models, functions, and algorithms may be used to predict one or more treatments, or combinations thereof, to counteract a certain amount of change in insulin resistance during a sub-phase or phase of the menstrual cycle. Figure 4 Examples of different algorithms that can be used are described herein. In some implementations, these data models, machine learning models, regression models, functions, or algorithms can be used to discover correlations between different variables (e.g., features). As mentioned above, variables may include glucose levels and / or changes therein, insulin resistance and / or changes therein, menstrual subphases or phases, and exercise and food-related variables.

[0131] In some implementations, based on these correlations, DAM 113 may be able to identify consistent patterns, based on which DAM 113 can be configured to determine an effective treatment for the user. In some implementations, these patterns show the effects of different types / amounts of medication (e.g., insulin) and different types / amounts of user behaviors (e.g., exercise, food) on the user's glucose levels at different stages of the menstrual cycle.

[0132] For example, a dataset might show that, on average, administering an additional 40% of basal insulin over the past 12 months offset a 30% increase in insulin resistance during a subphase or phase of the menstrual cycle. In another embodiment, a dataset might show that an extra hour of exercise per day is insufficient to offset a 10% increase in insulin resistance during a subphase or phase of the menstrual cycle. Similar embodiments are also within the scope of this disclosure. Based on the user's own historical data and the patterns found therein, DAM 113 can then determine whether one or more treatments are effective or ineffective.

[0133] In step 306, operation 300 then provides the user with one or more treatments. As mentioned above, one or more treatments may have been determined based on historical data associated with the user, historical data associated with one or more users in the user database 110, and / or scientific data points.

[0134] In some implementations, if one or more treatments involve administering a specific dose of insulin, the medication management feature of application 106 may send the user treatment recommendations (e.g., notifications) with delivery dose and / or delivery time once the specific dose of insulin is determined by DAM113, so that the user can administer insulin manually. Alternatively, the medication management feature may (e.g., upon receiving the user's approval) automatically communicate with the insulin delivery device to deliver the determined dose at the delivery time. In some implementations, when the user ingests insulin only in the form of an oral medication, the treatment may indicate an effective dose of such oral insulin. Providing the correct dose of insulin, whether automatically or manually administered, is advantageous because otherwise, the user may (1) overcompensate by administering an excessive amount of insulin to counteract changes in her insulin sensitivity, or (2) administer too little insulin and see her glucose levels spike, which could lead the user to continue injecting additional doses.

[0135] In some implementations, if one or more treatments include more and / or different exercise during a sub-phase or phase, the exercise management feature provides the user with treatment recommendations including the type and / or intensity of exercise. For example, a dataset associated with the user's historical data might show that running for an hour during the user's luteal phase consistently helps the user maintain her glucose levels within the target range. In such embodiments, the exercise management feature might recommend an additional hour of running or exercise of a similar type and / or intensity (swimming) to the user. For example, the exercise management feature might calculate that if the user chooses to swim instead of running, then 45 minutes of swimming would be sufficient.

[0136] In some implementations, if one or more treatments involve dietary changes during a sub-phase or phase, then dietary management features can provide treatment recommendations that allow the user to adjust different ways and forms of their diet accordingly. For example, a dataset associated with a user's historical data could show that consuming less than 15 grams of sugar or 50 grams of carbohydrates during the user's luteal phase consistently helps the user maintain her glucose levels within the target range. In such embodiments, dietary management features could recommend this finding as treatment to the user, and also recommend a diet consistent with previously effective meals and dishes.

[0137] In some implementations, a combination of two or more treatments may be recommended to a user. For example, for various reasons, some users may generally wish to reduce their dependence on insulin, in which case, to counteract their insulin resistance, application 106 may recommend a combination of exercise and diet. Such a combination that excludes insulin administration may be determined based on the user's own historical data and / or historical data associated with one or more users in user database 110. For example, at a certain stage of the menstrual cycle, such as the luteal phase, a user may always only administer additional insulin to counteract changes in her insulin resistance. However, the user could instead decide to do additional exercise and follow a stricter diet during her luteal phase instead of administering additional insulin. In this case, the user's own historical data may not strongly support how the user's glucose levels respond to more exercise and / or a stricter diet during her luteal phase. In such embodiments, inferences from historical data from similar users may suggest that a certain amount of additional exercise and / or certain dietary changes may be effective.

[0138] Figure 4 This is a diagram illustrating one embodiment of how to determine one or more treatments for a user, according to certain implementation methods. As shown in the diagram, in... Figure 4 In this embodiment, data is collected and prepared through three steps, A, B, and C (including steps C1 and C2). In step A, application 106 receives certain initial input from or about the user, such as demographic information 118, disease progression information 120, medication information 122, input 127 (which includes menstrual cycle information), and / or indicators 130. As mentioned above, because initially there may not be enough input 127 and indicators 130 to be received and / or available to the user, therefore... Figure 4 In one embodiment, in step C1, DAM 113 initially determines one or more treatments for a user based on dataset 440 associated with one or more users in user database 110.

[0139] Dataset 440 may correspond to records of historical data associated with all users or hierarchical groups of users in user database 110. These alternatives are shown with different types of dashed lines. For example, in some embodiments, in step B, as described above, DAM 113 hierarchizes user database 110 based on one or more hierarchical factors, resulting in a hierarchical group of users, shown as user group 2. In such an embodiment, dataset 440 corresponds to records of historical data associated with user group 2. In some other embodiments, in step B, DAM 113 may decide not to hierarchize user database 110, in which case dataset 440 corresponds to records of historical data associated with all users in user database 110.

[0140] In step C1, in some embodiments, dataset 440 is used as a training dataset, which is fed into machine learning (ML) algorithm 470 to output hypothesis function 450 (i.e., h). As described above, dataset 440 is developed and documented such that it includes a variety of data features (DFs), such as: one or more DFs corresponding to the time and / or location (e.g., which or which sub-phases) of a user's menstrual cycle, as understood by those skilled in the art; one or more DFs corresponding to blood glucose levels and / or variations therein; one or more DFs corresponding to insulin resistance and / or variations therein; one or more exercise-related DFs corresponding to the type and amount of exercise performed; one or more diet-related DFs corresponding to dietary restrictions and / or the type and amount of food consumed; one or more DFs corresponding to the type and / or amount of insulin administered; and / or additional DFs that can be used in the model. For example, any inputs and metrics (e.g., input 127 and metric 130) can be used to define additional DFs. Note that in machine learning and pattern recognition, a DF is a single measurable attribute or feature of the observed phenomenon. In some embodiments, ML algorithm 470 is a supervised learning algorithm. Supervised learning is a machine learning task that learns a function, for example, a function that maps inputs to outputs based on an exemplary input-output pair. Figure 4 In one embodiment, the ML algorithm 470 may include a supervised learning algorithm for solving multivariate or multivariate regression problems.

[0141] As described above, feeding dataset 440 into ML algorithm 470 yields hypothesis function 450, which takes one or more x as input and maps them to y. Hypothesis function 450 is developed based on the correlation between different DFs in dataset 440. For example, hypothesis function 450 might correspond to the following function, where Y is the dependent variable and X1, X2, ..., Xp are various independent variables.

[0142] Y=β0+β1X1+β2X2+β3X3+…+βpXp+∈.

[0143] In some implementations, each independent variable corresponds to a different hypothesis function (DF). For example, Y may correspond to glucose level (e.g., a decrease in glucose level, or a target glucose level or range), X1 may correspond to the time and / or phase of the user's menstrual cycle (e.g., which sub-phase or stage), X2 may be related to diet, X3 may correspond to the type and / or amount of exercise performed, X4 may correspond to the type and / or amount of insulin administered, X5 may be related to insulin resistance, and so on. In some implementations, B0, B1, ..., Bp are coefficients representing the correlation between corresponding X and Y. Note that choosing glucose level as Y or the output is merely one way of arranging hypothesis function 450. For the sake of brevity, this document does not describe in detail how hypothesis function 450 is output by ML algorithm 470 based on dataset 440 and continuously adjusted (e.g., potentially using cost functions, gradient descent algorithms, etc.), but this is known to those skilled in the art.

[0144] Using hypothesis 450, in some implementations, DAM 113 is able to predict a user's glucose levels, given a specific sub-phase, stage, or time point during the user's menstrual cycle, the amount / type of exercise, the quantity / type of food, and / or the amount / type of insulin used. Using hypothesis 450, DAM 113 is also able to determine one or more treatments for the user. To illustrate this, a user may have two days remaining before entering the first sub-phase of the luteal phase. The user may have indicated that they do not wish to engage in any additional exercise or follow any additional restrictive diet. In this embodiment, DAM 113 may use the target glucose range or level in hypothesis function 450 as "Y" and determine the amount of insulin "X4" that needs to be administered during that sub-phase, given that the value of "X1" represents the first sub-phase of the luteal phase, and the same values ​​are given for X2 and X3 (e.g., values ​​representing the same exercise and food regimen). In another embodiment, the amount of insulin administered may remain constant, while DAM 113 determines how much additional exercise the user needs to engage in during the first sub-phase to achieve that target glucose level. As will be appreciated by those skilled in the art, there are various ways to arrange and use hypothesis function 450. For example, a single treatment or a combination of treatments can be determined based on hypothesis function 450. In another embodiment, "Y" can represent a decrease in glucose levels. In such an embodiment, DAM 113 can use hypothesis function 450 to predict one or more treatments that can lead to a certain amount of decrease in glucose levels.

[0145] Please note that, although it is assumed that the development of 450 is in Figure 4As shown in step C1, but assumption 450 can even be developed by DAM 113 based on dataset 440 before the user uses application 106 (e.g., before step A). ​​In other words, DAM 113 can develop features using datasets associated with different stratified groups, and then use these features when a new user starts using application 106 and needs treatment advice. For example, different hypothesis functions can be developed for user group 3 such that if a new user starts using application 106 and belongs to user group 3, then DAM 113 can use the hypothesis function to predict which one or more treatments might be effective for the user at a specific sub-phase, stage, or time point in the user's menstrual cycle.

[0146] Using supervised learning algorithms to develop hypothesis functions such as 450 is merely one example of how DAM 113 can predict a user's glucose levels during a specific sub-phase or stage of the menstrual cycle and / or predict one or more effective treatments for that sub-phase or stage to help the user maintain or reach a certain glucose level or range. Alternatively, other machine learning algorithms, such as neural network algorithms, can be used.

[0147] Output 460 represents one or more treatments that can be recommended or administered to the user. The application 106 then receives and records the physiological effects of output 460 on the user (e.g., displayed as user 102) in the form of input / metric 461. For example, output 460 could include administering a dose of basal insulin one day before the user enters the first sub-phase of the luteal phase. In such an embodiment, when the user enters the first sub-phase, her glucose level may be higher than expected. In this embodiment, the user's glucose level and other metrics are then received, recorded, and analyzed as input / metric 461. Input / metric 461 is not only recorded in the user profile 116, and its information is subsequently fed back to dataset 440, but it is also used to develop a training dataset 472 specifically for the user. Therefore, output 460 and input / metric 461 are used to develop datasets 440 and 472.

[0148] When additional inputs and metrics are fed back into dataset 440, the updated dataset 440 is again fed back into ML algorithm 470, resulting in a constantly changing and dynamic hypothesis function 450 that reflects the updates to the data in dataset 440. The constantly changing dynamic hypothesis function 450 specifically refers to a set of dynamic coefficients (e.g., B0, B1, ..., Bp).

[0149] After a period of time, dataset 472 may be sufficiently developed so that, based on dataset 472, the user's glucose levels and / or insulin resistance, as well as the effectiveness of different treatments in specific sub-phases or stages of the user's menstrual cycle, can be predicted with high confidence. At such a point in time, in step C2, in some embodiments, DAM 113 may utilize the predictions provided by dataset 472 to provide treatment to the user, or at least give such predictions more weight, compared to predictions provided by dataset 440. Similar to dataset 440, dataset 472 can be used as a training dataset, which can be fed into machine learning algorithm 475 to develop hypothesis function 480. Machine learning algorithm 475, similar to or different from dataset 440, can be used, as understood by those skilled in the art or of ordinary skill. For example, ML algorithm 475 may be a supervised learning regression algorithm. Thus, in some embodiments, the resulting hypothesis function 480, based on the user's own historical data (dataset 472), is used to predict the user's glucose levels and / or insulin resistance, as well as one or more treatments that may effectively help the user maintain or achieve certain glucose levels in a certain sub-phase or stage of the user's menstrual cycle.

[0150] In some implementations, the hypothesis function 480 outputs its content as treatment recommendations and / or treatment administration as output 460 to the user. The physiological effects of this being performed or managed output 460 are then received and / or recorded as inputs / indicators 461 to further develop the dataset 472. As additional inputs / indicators 461 are fed back into the dataset 472, the updated dataset 472 is again fed back into the ML algorithm 475, resulting in a continuously evolving, dynamic hypothesis function 480 that reflects the dynamic updates of the data in the dataset 472.

[0151] In some implementations, DAM 113 may determine one or more treatments for a user based on the outputs provided by hypothesis functions 480 and 450. For example, in some implementations, DAM 113 may not have high confidence in the outputs provided by hypothesis function 480. Therefore, in such embodiments, DAM 113 may supplement the outputs provided by hypothesis function 480 with the outputs provided by hypothesis function 450. For example, DAM 113 may use a function that provides a final output based on weights assigned to the outputs provided by hypothesis function 480 and hypothesis function 450.

[0152] Note that using a combination of historical data associated with a user's stratification group and historical data associated with the user is just one example of how to predict and deliver effective treatment to the user. In some other implementations, DAM 113 may initially predict and deliver treatment to the user based solely on scientific data points, and then use the user's own historical data when sufficient is available. In still other implementations, only the user's own data may be used, since sufficient information about the user is available from the outset. As an example, a user may use application 106 and a third-party period tracking application separately for several years. In such an implementation, at some point, the user may grant application 106 access to the user's historical menstrual cycle information over the past few years by enabling application 106 to communicate with a third-party period tracking application that has that record. Thus, in this implementation, DAM 113 may correlate the user's menstrual cycle information with the user's glucose and / or insulin resistance trends and create a set of timestamped data, which can then be fed into a machine learning algorithm to create a hypothesis function for prediction. In still other implementations, DAM 113 may rely solely on historical data associated with the user's stratification group to deliver effective treatment to the user.

[0153] Figure 5 This is a block diagram depicting a computing device 500 configured to predict a user's glucose levels and / or insulin resistance, and one or more treatments that can effectively help the user maintain or achieve a certain glucose level during a sub-phase or phase of the user's menstrual cycle. In some embodiments, according to certain embodiments disclosed herein, an application provides one or more treatments to the user, the application executing on the computing device 500 or another computing device communicating with the computing device 500. Although depicted as a single physical device, in embodiments, the computing device 500 may be implemented using a virtual device and / or across multiple devices (e.g., in a cloud environment). As shown, the computing device 500 includes a processor 505, memory 510, storage 515, a network interface 525, and one or more I / O interfaces 520. In the illustrated embodiment, the processor 505 retrieves and executes programming instructions stored in memory 510 and stores and retrieves application data residing in storage 515. The processor 505 typically represents a single CPU and / or GPU, multiple CPUs and / or GPUs, a single CPU and / or GPU with multiple processing cores, etc. Typically includes memory 510 to represent random access memory. Memory 515 can be any combination of disk drives, flash-based storage devices, etc., and can include fixed and / or removable storage devices such as fixed disk drives, removable memory cards, caches, optical storage, network-attached storage (NAS), or storage area networks (SANs).

[0154] In some embodiments, input and output (I / O) devices 535 (e.g., keyboard, monitor, speaker, etc.) can be connected via I / O interface 520. Furthermore, via network interface 525, computing device 500 can be communicatively coupled to one or more other devices and components (e.g., user database 110). In some embodiments, computing device 500 is communicatively coupled to other devices via a network, which may include the Internet, a local network, etc. The network may include wired connections, wireless connections, or a combination of wired and wireless connections. As shown, processor 505, memory 510, storage 515, network interface 525, and I / O interface 520 are communicatively coupled via one or more interconnects 530. In some embodiments, computing device 500 represents a mobile device 107 associated with a user. In some embodiments, as described above, mobile device 107 may include a user's laptop, computer, smartphone, etc. In another embodiment, computing device 500 is a server operating in a cloud environment.

[0155] In the illustrated embodiment, memory 515 includes user profile 116. In some embodiments, memory 515 also includes features for use with... Figure 3 and 4 Any dataset relating to one or more operations. In some other embodiments, such a dataset may be stored alternatively or additionally in user database 110. Memory 510 includes decision support engine 112, which itself includes DAM 113. Decision support engine 112 is executed by computing device 500 to perform... Figure 3 One or more steps of operation 300 in the process.

[0156] Each of these non-limiting examples may exist independently or may be combined with one or more other embodiments in various permutations or combinations. The above detailed description includes references to the accompanying drawings, which form part of the specification. The drawings illustrate, by way of illustration, specific embodiments in which the invention may be practiced. These embodiments are also referred to herein as “examples.” Such embodiments may include elements other than those shown or described. However, the inventors also contemplate embodiments in which only those elements shown or described are provided. Furthermore, the inventors contemplate examples of any combination or permutation of the elements (or one or more aspects thereof) shown or described, whether for a particular example (or one or more aspects thereof) or for other examples (or one or more aspects thereof) shown or described herein.

[0157] In the event of any inconsistency between the usage of this document and any other document incorporated by reference, the usage in this document shall prevail.

[0158] In this document, as is common in patent documents, the terms “a” or “an” are used to include one or more, regardless of any other instance or use of “at least one” or “one or more.” In this document, the term “or” is used to indicate a non-exclusive “or,” thus “A or B” includes “A but not B,” “B but not A,” and “A and B,” unless otherwise stated. In this document, the terms “comprising” and “wherein” are used as their simple English equivalents. Furthermore, in the following claims, the terms “comprising” and “including” are open-ended, meaning that a system, apparatus, article, composition, formulation, or process that includes those elements in addition to those listed after such terms in the claim is still considered to fall within the scope of that claim. Furthermore, in the following claims, the terms “first,” “second,” and “third,” etc., are used merely as labels and are not intended to impose numerical requirements on their objects.

[0159] Geometric terms, such as “parallel,” “perpendicular,” “circular,” or “square,” are not intended to demand absolute mathematical precision unless the context otherwise requires. Instead, such geometric terms allow for variation due to manufacturing or equivalent functions. For example, if an element is described as “circular” or “generally circular,” components that are not precisely circular (e.g., slightly elliptical or polygonal components) are still included in that description.

[0160] The method embodiments described herein can be implemented, at least in part, by a machine or computer. Some embodiments may include a computer-readable or machine-readable medium encoded with instructions operable to configure an electronic device to perform the methods described in the embodiments above. Implementations of this method may include code, such as microcode, assembly language code, high-level language code, etc. Such code may include computer-readable instructions for performing various methods. This code may form part of a computer program product. Furthermore, in embodiments, the code may be physically stored on one or more volatile, non-transient, or non-volatile tangible computer-readable media, for example, during execution or at other times. Examples of such tangible computer-readable media may include, but are not limited to, hard disks, removable disks, removable optical disks (e.g., optical discs and digital video disks), magnetic tapes, memory cards or memory sticks, random access memory (RAM), read-only memory (ROM), etc.

[0161] The above description is intended to be illustrative and not restrictive. For example, the embodiments described above (or one or more aspects thereof) may be used in combination with each other. Other embodiments may be used, for example, by those skilled in the art after reading the above description. An abstract is provided to comply with the requirements of 37 CFR § 1.72(b) to enable the reader to quickly determine the nature of the technical disclosure. It should be understood that it is not intended to interpret or limit the scope or meaning of the claims. Furthermore, in the above detailed description, various features may be combined to simplify this disclosure. This should not be construed as meaning that any disclosed feature not claimed is essential to any claim. Rather, the subject matter of the invention may not be limited to all features of a particular disclosed embodiment. Therefore, the following claims are incorporated herein as embodiments or implementations, each claim existing independently as a separate implementation, and these implementations are contemplated to be combined with each other in various combinations or arrangements. The scope of the invention should be determined by reference to the appended claims and the full scope of their equivalents.

Claims

1. A system comprising: A glucose monitoring system, comprising: processor; A glucose sensor, configured to measure a user's glucose levels to generate a glucose measurement. A sensor electronics module configured to transmit sensor data corresponding to the glucose measurement value provided by the glucose sensor to the processor; A storage circuit that communicates with the processor; The processor is configured to execute instructions stored thereon to: receive information related to the user's menstrual cycle, which includes multiple menstrual phases. Identify the menstrual phase the user is experiencing from the multiple menstrual phases. Based on historical data associated with user stratification groups, it is determined that the user achieved targeted glucose treatment during an identified menstrual phase of the menstrual cycle, wherein the user stratification groups are grouped at least based on menstrual cycle information. The historical data associated with user hierarchical groups is structured to indicate: Historically, changes in insulin resistance were associated with stratified groups of users in the identified menstrual phases, and The physiological effect patterns of multiple treatments, including treatments that determine glucose measurements of a user's stratified group during the identified menstrual phase, indicate the effectiveness of the multiple treatments in counteracting changes in insulin resistance and achieving the target glucose level.

2. The system according to claim 1, wherein, The processor is configured to provide the treatment.

3. The system according to claim 2, wherein, The treatment includes a dose of insulin.

4. The system according to claim 3, wherein, The dose of insulin is higher than the average dose of insulin administered to the user during the non-luteal phase of the user's menstrual cycle.

5. The system according to claim 3, wherein, The processor is configured to transmit signals to a drug delivery device to administer the dose of insulin to the user.

6. The system according to claim 3, wherein, The processor is configured to provide a treatment recommendation to the user or another individual, the treatment recommendation indicating the dose of insulin.

7. The system according to claim 3, wherein, The processor is configured to provide treatment recommendations to the user or another individual, the treatment recommendations indicating at least one of the amount, type, duration, and intensity of exercise.

8. The system according to claim 3, wherein, The processor is configured to provide a treatment recommendation to the user or another individual, the treatment recommendation indicating at least one of the quantity and type of food.

9. A method for personalized guidance based on information related to a user's menstrual cycle, comprising: Use a glucose monitoring system to measure the user's glucose levels; The processor, which communicates with the glucose monitoring system, receives information related to the user's menstrual cycle, which includes multiple menstrual phases. Identify the menstrual phase the user is experiencing from the plurality of menstrual phases; Based on historical data associated with a user's stratification group, guidance for that user is determined at the processor to achieve target glucose levels during the user's menstrual phase, wherein the user's stratification group is grouped at least based on menstrual cycle information. The historical data associated with user hierarchical groups is structured to indicate: Historically, changes in insulin resistance were associated with stratified groups of users in identified menstrual phases, and Multiple guidance patterns of physiological effects, including guidance on determining glucose measurements for a stratified group of users during the identified menstrual phase, the patterns indicating the effectiveness of the multiple guidance patterns in counteracting changes in insulin resistance and achieving the target glucose level.

10. The method according to claim 9, wherein, Providing the guidance also includes providing guidance to the user or another individual, the guidance indicating at least one of the amount, type, length, and intensity of exercise.

11. The method according to claim 9, wherein, Providing the guidance also includes providing guidance to the user or another individual, the guidance indicating at least one of the quantity and type of food.

12. A non-transitory computer-readable medium having instructions stored thereon, which, when executed by a system, causes the system to perform a method comprising: Use a glucose monitoring system to measure the user's glucose levels; The processor, which communicates with the glucose monitoring system, receives information related to the user's menstrual cycle, which includes multiple menstrual phases. Identify the menstrual phase the user is experiencing from the multiple menstrual phases. Based on historical data associated with a user's stratification group, the processor determines the user's treatment to achieve target glucose levels during the user's identified menstrual phase, wherein the user's stratification group is grouped at least based on menstrual cycle information. The historical data associated with user hierarchical groups is structured to indicate: Historically, changes in insulin resistance were associated with stratified groups of users in the identified menstrual phases, and A pattern of physiological effects of multiple treatments, including treatments that determine glucose measurements of a stratified group of users during the menstrual phase, the pattern indicating the effectiveness of the multiple treatments in counteracting changes in insulin resistance and achieving the target glucose level.

13. The non-transitory computer-readable medium according to claim 12, wherein, The method also includes providing the treatment.

14. The non-transitory computer-readable medium according to claim 13, wherein, The treatment includes a dose of insulin.

15. The non-transitory computer-readable medium according to claim 14, wherein, The dose of insulin is higher than the average dose of insulin administered to the user during the non-luteal phase of the user's menstrual cycle.

Images