Predicting analyte levels based on intermittent sensor data
Techniques for predicting analyte levels based on intermittent sensor data are disclosed. First sensor data is acquired from an analyte sensor. The first sensor data reflects analyte levels of a user who is wearing the analyte sensor. The first sensor data is collected over a first time period. Later, a determination is made as to whether data is still being acquired from the analyte sensor. As a result of determining that data is no longer being acquired from the analyte sensor, the first sensor data is classified as intermittent analyte data. The intermittent analyte data is then used to generate predicted analyte level data for the user during a second time period that is subsequent to the first time period. The predicted analyte level data is reflective of the intermittent analyte data.
1 . A computer system that predicts analyte levels, the computer system comprising:
a processor system; and
a storage system comprising instructions that are executable by the processor system to cause the computer system to:
acquire first sensor data from a first analyte sensor, wherein the first sensor data reflects analyte levels of a user who is wearing the first analyte sensor, and wherein the first sensor data is collected over a first time period;
determine whether data is still being acquired from the first analyte sensor;
upon a determination that data is no longer being acquired from the first analyte sensor, classify the first sensor data as intermittent analyte data;
use the intermittent analyte data to generate predicted analyte level data for the user during a second time period that is subsequent to the first time period, wherein the predicted analyte level data is reflective of the intermittent analyte data;
determine that the predicted analyte level data is stale, wherein determining that the predicted analyte level data is stale comprises designating the predicted analyte level data as unreliable for continued use without acquisition of new sensor data, and wherein determining that the predicted analyte level data is stale comprises:
(i) determining that a time elapsed since acquisition of most recent sensor data from the analyte sensor exceeds a predetermined staleness threshold; or
(ii) determining that a confidence level or accuracy metric associated with the predicted analyte level data has fallen below a threshold value due to an absence of updated sensor data;
in response to determining that the predicted analyte level data is stale, generate a prompt to the user to facilitate acquisition of new analyte sensor data, wherein the prompt instructs the user to use a new analyte sensor to acquire the new analyte sensor data or to facilitate acquiring the new analyte sensor data from the first analyte sensor; and
acquire the new analyte sensor data from either the first analyte sensor or the new analyte sensor, wherein the second sensor data is acquired after the prompt is provided to the user.
2 . The computer system of claim 1 , wherein the intermittent analyte data is compared against a repository of baseline sensor data for other users and a result of said comparison is included in a user profile, or, alternatively, wherein the instructions further cause the computer system to determine if the user is not wearing the first analyte sensor during the second time period.
3 . The computer system of claim 1 , wherein the predicted analyte level data is compared against metadata associated with the user, wherein the metadata includes one or more of a past glucose level data for the user, location, food consumed, information obtained from a food delivery application, physical activity, or date and time.
4 . The computer system of claim 1 , wherein the first time period is at least 10 contiguous days.
5 . The computer system of claim 1 , wherein the first time period is at least 14 contiguous days.
6 . The computer system of claim 1 , wherein the instructions are further executable to cause the computer system to:
acquire new sensor data from a second analyte sensor, different than the first analyte sensor, wherein the second analyte sensor is detected via an application, and wherein a profile is updated using data obtained from the second analyte sensor.
7 . The computer system of claim 1 , wherein the predicted analyte level data is compared against trending data of other users.
8 . The computer system of claim 1 , wherein the instructions are further executable to cause the computer system to:
determine a standard population trend for multiple other users; and
compare a profile of the user against the standard population trend.
9 . The computer system of claim 1 , wherein a profile of the user includes information relating to meals the user has consumed.
10 . A method for generating an analyte level profile for a user, the method comprising:
acquiring first sensor data from an analyte sensor, wherein the first sensor data reflects analyte levels of a user who is wearing the analyte sensor, and wherein the first sensor data is collected over a first time period;
using the first sensor data to generate a profile for the user, wherein the profile tracks the analyte levels of the user throughout the first time period;
during a second time period that is subsequent to the first time period, determining that sensor data is no longer being acquired from the analyte sensor, resulting in the first sensor data being discontinuous analyte data;
using the discontinuous analyte data to generate predicted analyte level data for the user;
updating the profile by including the predicted analyte level data;
determining that the predicted analyte level data is stale, wherein determining that the predicted analyte level data is stale comprises designating the predicted analyte level data as unreliable for continued use without acquisition of new sensor data, and wherein determining that the predicted analyte level data is stale comprises:
(i) determining that a time elapsed since acquisition of most recent sensor data from the analyte sensor exceeds a predetermined staleness threshold; or
(ii) determining that a confidence level or accuracy metric associated with the predicted analyte level data has fallen below a threshold value due to an absence of updated sensor data;
in response to determining that the predicted analyte level data is stale, generating a prompt to the user to facilitate acquisition of new analyte sensor data, wherein the prompt instructs the user to use a new analyte sensor to acquire the new analyte sensor data or to facilitate acquiring the new analyte sensor data from the first analyte sensor; and
acquiring the new analyte sensor data from either the first analyte sensor or the new analyte sensor, wherein the second sensor data is acquired after the prompt is provided to the user.
11 . The method of claim 10 , wherein the profile includes trend data reflecting how a body of the user is becoming more insulin resistant.
12 . The method of claim 10 , wherein supplemental data is added to the profile, and wherein the supplemental data includes data from a hemoglobin A1C test.
13 . The method of claim 10 , wherein the method further includes receiving user input that supplements the discontinuous analyte data, and wherein the user input provides a context for at least some of the discontinuous analyte data.
14 . The method of claim 10 , wherein the method further includes generating a meal score for a meal the user has consumed.
15 . A method for generating a health model for a user the method comprising:
acquiring first sensor data from an analyte sensor, wherein the first sensor data reflects a physiological response of a user who is wearing the analyte sensor, and wherein the first sensor data is collected over a first time period;
using the first sensor data to generate the health model for the user, wherein the health model tracks the physiological response of the user throughout the first time period;
during a second time period that is subsequent to the first time period, determining that updated sensor data is no longer being acquired from the analyte sensor, resulting in the first sensor data being periodic sensor data;
using the periodic sensor data to generate predicted physiological response data for the user during the second time period;
updating the model by including the predicted physiological response data;
determining that the predicted analyte level data is stale, wherein determining that the predicted analyte level data is stale comprises designating the predicted analyte level data as unreliable for continued use without acquisition of new sensor data, and wherein determining that the predicted analyte level data is stale comprises:
(i) determining that a time elapsed since acquisition of most recent sensor data from the analyte sensor exceeds a predetermined staleness threshold; or
(ii) determining that a confidence level or accuracy metric associated with the predicted analyte level data has fallen below a threshold value due to an absence of updated sensor data;
in response to determining that the predicted analyte level data is stale, generating a prompt to the user to facilitate acquisition of new analyte sensor data, wherein the prompt instructs the user to use a new analyte sensor to acquire the new analyte sensor data or to facilitate acquiring the new analyte sensor data from the first analyte sensor; and
acquiring the new analyte sensor data from either the first analyte sensor or the new analyte sensor, wherein the second sensor data is acquired after the prompt is provided to the user.
16 . The method of claim 15 , wherein the method further includes:
facilitating peer to peer sharing of sensor data; and
identifying users who share physiological response characteristics that are similar to the user.
17 . The method of claim 15 , wherein the method further includes:
accessing one of a calendar or a global positioning system (GPS) data to acquire supplemental data about the user; and
associating supplemental data with periodic sensor data using time information, wherein the supplemental data provides context for at least a portion of the periodic sensor data.
18 . The method of claim 15 , wherein the method further includes providing a coaching prompt to the user during the second time period.
19 . The method of claim 15 , wherein the first time period is at least 10 contiguous days.
20 . The method of claim 15 , wherein the first time period is at least 14 contiguous days.