Personalized health coaching system
Systems and methods for a comprehensive and personalized approach to health and lifestyle coaching are described. The system may determine health metrics of a user based on detected physiological parameters. The health metrics may be used to determine health recommendations and transmit feedback to the user based on user compliance with the recommendations.
1. A method of generating a recommendation based on a classification of a glucose response of a user, the method comprising:
receiving an input comprising at least three blood glucose values of the user measured by at least one first sensor over a first period;
assigning a first classification to the input using a first classifier trained using machine learning using historical glucose values of the user measured by the at least one first sensor and which classifies the input according to a variability of the at least three blood glucose values measured by the at least one first sensor, the first classification associated with a glucose sensitivity due to lifestyle of the user;
generating a modified input by applying a bias correction to the input;
assigning a second classification to the modified input using a second classifier, the second classification associated with glycemic load of a recent meal, a recent food, or a recent activity;
determining a score for the user's glucose response during the first period based on the first classification and the second classification;
determining that the score passes a recommendation threshold; and
outputting to a display a user recommendation relating to the recent meal, the recent food, or the recent activity based on the score.
2. The method of claim 1 , wherein generating the modified input is further based on lifestyle factors.
3. The method of claim 1 , wherein generating the modified input is further based on an absolute value of blood glucose measurements during the first period.
4. The method of claim 1 , wherein the second classification is associated with glucose variations during the first period.
5. The method of claim 1 , wherein the score is a grade on an alphanumeric grading scale.
6. The method of claim 1 , further comprising generating at least one glucose prediction based on the score.
7. The method of claim 1 , wherein the classifier comprises a neural network having at least three layers.
8. The method of claim 1 , wherein the first period is 2.5 hours.
9. The method of claim 1 , wherein the at least three blood glucose values are sampled at a frequency of 5-7 minutes.
10. The method of claim 1 , wherein the first and second sensors are the same.
11. The method of claim 1 , further comprising outputting the score to a meal log of the display.
12. A health coaching system comprising:
a non-transitory computer storage medium configured to at least store computer-readable instructions; and
one or more hardware processors in communication with the non-transitory computer storage medium, the one or more hardware processors configured to execute the computer-readable instructions to at least:
receive a user input comprising at least three blood glucose values of a user measured by at least one first sensor over a first period;
assign a first classification to the user input using a first classifier trained using machine learning using historical glucose values of the user measured by the at least one first sensor and which classifies the user input according to a variability of the at least three blood glucose values measured by the at least one first sensor, the first classification associated with a glucose sensitivity due to lifestyle of the user;
generate a modified input by applying a bias correction to the user input;
assign a second classification to the modified input using a second classifier, the second classification associated with glycemic load of a recent meal, a recent food, or a recent activity;
determine a score associated with a glucose response during the first period based on the first classification and the second classification;
determine that the score passes a recommendation threshold;
generate in a window on a display, a recommendation relating to the recent meal, the recent food, or the recent activity based on the determined score.
13. The health coaching system of claim 12 , wherein the recommendation comprises at least one similar food or beverage to avoid.
14. The health coaching system of claim 12 , wherein the recommendation comprises at least one food or beverage to substitute.
15. The health coaching system of claim 12 , wherein the recommendation comprises at least one food or beverage to ingest in conjunction with the user input.
16. The health coaching system of claim 12 , wherein the recommendation comprises an activity recommendation.
17. The health coaching system of claim 16 , wherein the activity recommendation comprises a length and type of activity.
18. The health coaching system of claim 12 , wherein the at least one recent food comprises a plurality of food or beverage and wherein the score comprises an overall score associated with the plurality of food or beverage.
19. The health coaching system of claim 12 , wherein the first classifier is a neural network having at least three layers.
20. The health coaching system of claim 12 , wherein the one or more hardware processors are further configured to output the score to a meal log of the display.