Database management and graphical user interfaces for measurements collected by analyzing blood
Inventors
McRAITH, Kevin • KESANI, Hari • Iyer, Anand • SUSAI, Gabriel • SHOMALI, Mansur • RAO, Prasad Matti
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-11587681-B2
Publication Date
2023-02-21
Expiration Date
Abstract
Methods and devices include database management and graphical user interfaces for measurements collected by analyzing blood.
Core Innovation
A computer-implemented method and system manage a user’s blood glucose levels by using continuous glucose monitoring values and electronically receiving data at a server, including a length of time that the user has been diagnosed with a blood glucose condition, types and dosages of medications consumed by the user, and the blood glucose values of the user, or types and dosages of medications consumed by the user and A1C values of the user. The method stores the received data in a database, extracts health data or metadata from the stored data, and outputs a generated treatment plan to the user via an electronic device of the user.
The treatment plan and goal are generated using one or more machine learning algorithms that include artificial neural networks, Bayesian statistics, case-based reason, decision trees, inductive logic processing, Gaussian process regression, Gene expression programming, Logistic model trees, stochastic modeling, or statistical modeling. The machine learning algorithms input cohort data from one or more other users corresponding to one or more other health devices, and use the cohort data to generate a treatment plan for improving blood glucose values of the user at an end of a treatment period as compared to blood glucose values at a beginning of the treatment period, or to generate a goal that includes a reduction in A1C value at the end of the treatment period.
The method revises the treatment plan for a subsequent subset of the treatment period based on determined medication compliance, determined exercise compliance, determined diet compliance, and identified patterns, and outputs the revised treatment plan with tasks that include a change in prescribed blood glucose measurement pairs, prescribed timing and dosage of medication, prescribed amount of carbohydrates, or prescribed exercise. Based on the identified patterns, the method determines a trigger event that occurs before an adverse effect, where the trigger event corresponds to deviation from an expected pattern, and sends a notification to the user upon detecting a subsequent instance of the trigger event, where the notification includes an identification of the trigger event and an identification of the adverse effect.
In additional aspects, the method receives GPS data from the electronic device, receives an indication from the user that the user would like to exercise, retrieves GPS data, and generates a route for the user to walk along, wherein a distance of the route corresponds to the prescribed exercise in the treatment plan. The method administers a questionnaire to the user, electronically receives user answers to the questionnaire, and determines a tendency of the user to follow a medication regimen and a diet regimen based on the user answers, and uses this tendency as part of generating the treatment plan. In one recited configuration, the trigger event is a day of the week and the adverse effect includes hyperglycemia.
Claims Coverage
The provided material includes three independent claims: clm-00001, clm-00010, and clm-00018. Across these independent claims, the main inventive features revolve around cohort-based machine learning for generating and revising a blood glucose treatment plan with user tasks, and detecting trigger events before adverse blood glucose effects followed by user notifications.
Cohort-based machine learning treatment plan generation using CGM and patient metadata
Extracting health data comprising cohort data from one or more other health devices corresponding to one or more other users, inputting the cohort data into one or more machine learning algorithms to generate a treatment plan for improving blood glucose values of the user at an end of a treatment period as compared to blood glucose values at a beginning of the treatment period, wherein the treatment plan includes instructions for tasks to be performed by the user during a first subset of the treatment period, and wherein the one or more machine learning algorithms include one or more of artificial neural networks, Bayesian statistics, case-based reason, decision trees, inductive logic processing, Gaussian process regression, Gene expression programming, Logistic model trees, stochastic modeling, or statistical modeling.
Time-subset execution and compliance- and pattern-based plan revision
Electronically receiving data relating to the treatment plan during the first subset of the treatment period, revising the treatment plan for a subsequent subset of the treatment period based on a determined medication compliance, a determined exercise compliance, and identified patterns, and outputting the revised treatment plan including one or more tasks to be performed during the subsequent subset, wherein the tasks include a change in one or more of the prescribed blood glucose measurement pairs to be measured before and after meals, prescribed timing and dosage of medication to be consumed, prescribed amount of carbohydrates, or prescribed exercise to be performed.
Trigger event detection preceding adverse effects and notification
Determining, based on the identified patterns, a trigger event that occurs before an adverse effect, the trigger event corresponding to deviation from an expected pattern, and sending a notification to the user upon detecting an instance of the trigger event, wherein the notification includes an identification of the trigger event to the user and an identification of the adverse effect.
A1C-focused goal generation and adherence/tendency conditioning
Inputting metadata into one or more machine learning algorithms to generate (a) a goal for the user, wherein the goal includes a reduction in the A1C value of the user at an end of a treatment period as compared to an A1C value at a beginning of the treatment period, and wherein the machine learning algorithms generate the goal based at least on cohort data for a cohort associated with the user, and (b) the treatment plan for the user to achieve the goal based on the types and dosages of medications consumed by the user, further including determining a tendency of the user to follow at least one of a medication regimen, a diet regimen, and an exercise regimen, wherein generating the treatment plan also is based on the determined tendency.
GPS-based exercise route generation corresponding to prescribed exercise distance
Receiving GPS data from the electronic device of the user, receiving an indication from the user that the user would like to exercise, retrieving GPS data from the electronic device, and generating a route for the user to walk along, wherein a distance of the route corresponds to the prescribed exercise to the user in the treatment plan.
Questionnaire-based regimen/diet tendency determination
Administering, via a software application downloaded to the electronic device of the user, a questionnaire to the user, electronically receiving user answers to the questionnaire, determining based on the received user answers a tendency of the user to follow a medication regimen and a diet regimen, and wherein generating the treatment plan also is based on the determined tendency.
Across clm-00001, clm-00010, and clm-00018, the independent claims emphasize machine learning algorithms using cohort data to generate a blood glucose treatment plan and/or A1C reduction goal, execution via task prescriptions over treatment-period subsets, revision based on compliance and identified patterns, and trigger-event determination to notify the user about impending adverse blood glucose effects.
Stated Advantages
Documented Applications
No documented applications found
Interested in licensing this patent?