Blood glucose states based on sensed brain activity
Blood glucose states based at least on sensed brain activity data. Blood glucose states may be predicted states using prediction models, and may be real-time or future glucose states. One or more wearable sensors, optionally adapted for use in either single channel or dual channel sensing, can record brain activity signals and communicate brain activity signal data to a remote device.
1. A system adapted for predicting a glucose state of a subject, comprising:
a wearable non-invasive brain activity signal sensor (“wearable sensor”) that includes sensing electrodes consisting of first and second sensing electrodes and a third reference electrode and adapted for single channel or dual channel sensing, the wearable sensor sized and configured to be secured at a behind the ear location and on skin of a subject, the wearable sensor including a wireless communication component, a power source and one or more processing components;
a subject device configured to receive information wirelessly from the wearable sensor, the subject device including:
one or more processors;
a memory coupled to the one or more processors, the memory storing computer-program instructions, that, when executed by the one or more processors, cause the performance of:
receiving or generating sensed brain activity signals or processed brain activity signals sensed by the wearable non-invasive brain activity sensor;
predicting, using a trained prediction model, a predicted glucose state of the subject that is based at least partially on the sensed brain activity signals or the processed brain activity signals,
wherein predicting the predicted glucose state comprises predicting based on at least brain activity data from only a subset of frequencies within a larger range of frequencies that includes a delta band, a theta band, an alpha band, a beta band, and a gamma band, and
wherein predicting the predicted glucose state further comprises predicting the glucose state at one or more future epochs of time, wherein the one or more future epochs of time comprise a personal future epoch of time for the subject that has greater predictive accuracy of glucose state than other future epochs of time for the subject; and
outputting instructions to initiate a communication indicative of the predicted glucose state based on and in response to the predicted glucose state.
2. The system of claim 1 , wherein the system is further configured to filter out sensed brain activity signals that are within one or more of the delta band, the theta band, the alpha band, the beta band, or the gamma band, that are outside of the subset of frequencies.
3. The system of claim 2 , wherein the wearable sensor includes one or more filtering components to filter out frequencies outside of the subset of frequencies.
4. The system of claim 2 , wherein the subject device includes one or more filtering components to filter out frequencies outside of the subset of frequencies.
5. The system of claim 1 , wherein predicting the predicted glucose state further comprises predicting real-time blood glucose values.
6. The system of claim 5 , wherein predicting real-time blood glucose values comprises predicting the real-time blood glucose values while the subject is asleep.
7. The system of claim 6 , wherein the computer-program instructions further cause the performance of comparing the predicted real time glucose values to sensed glucose values sensed from a continuous glucose monitor adapted to sense interstitial glucose and worn by the subject at a location different than the wearable sensor while the subject is asleep, and initiate instructions to cause the continuous glucose monitor to be calibrated based on the predicted real time glucose values.
8. The system of claim 1 , wherein predicting glucose values comprises predicting glucose values from 15 minutes to 13 hours from when the non-invasively sensed brain activity signals were sensed by the wearable sensor.
9. The system of claim 1 , wherein a personalized lag time has an accuracy of at least 90% accuracy for the subject.
10. The system of claim 1 , wherein the subset of frequencies includes one or more frequencies within the beta and gamma bands, and wherein the subset excludes frequencies within the delta, theta, and alpha bands.
11. The method of claim 1 , wherein the subset of frequencies includes one or more frequencies within the delta, theta, and alpha bands, and wherein the subset excludes frequencies within the gamma and beta bands.
12. The system of claim 1 , wherein predicting the predicted glucose state comprises predicting one or more aspects of a post-prandial event.
13. The system of claim 12 , wherein the one or more aspects of the post-prandial event comprises at least one of a glucose value at a post-prandial glucose peak, or an epoch of time until a post-prandial peak in glucose value is predicted to occur.
14. The system of claim 12 , wherein predicting the one or more aspects of the post-prandial event comprises predicting the one or more aspects of the post-prandial event from 5 min-90 minutes before the post-prandial event.
15. The system of claim 1 , wherein predicting the predicted glucose state further comprises,
predicting a real-time glucose state while the subject is asleep, and
while the subject is awake, predicting the glucose state at the one or more future epochs of time.
16. The system of claim 1 , wherein the predicted glucose state is one of a predicted blood glucose level or a predicted interstitial glucose level.
17. The system of claim 1 , wherein predicting the predicted glucose state comprises detecting a change in power over time in the subset of frequencies.
18. The system of claim 1 , wherein the trained model is a model trained on personal non-invasively sensed brain activity signals from a behind the ear location and on glucose values estimated using one or more of a continuous glucose monitor adapted to sense interstitial fluid glucose, an implantable glucose sensor, finger prick blood measurements, or a non-invasive wearable glucose monitor adapted to monitor glucose.
19. The system of claim 1 , wherein the computer-program instructions are adapted to train the trained prediction model, comprising:
receiving or generating the sensed brain activity signals or processed brain activity signals sensed by the wearable sensor;
providing or receiving personal sensed glucose data from a glucose monitor worn by the subject; and
based on the sensed brain activity signals or processed brain activity signals, and the personal sensed glucose data from the glucose monitor,
identifying or determining the one or more future epochs of time for the subject that have greater predictive accuracy of glucose state than other future epochs of time for the subject.
20. A system adapted for predicting a glucose state of a subject, comprising:
a wearable non-invasive brain activity signal sensor (“wearable sensor”) that includes sensing electrodes consisting of first and second sensing electrodes and a third reference electrode and adapted for single channel or dual channel sensing, the wearable sensor sized and configured to be secured at a behind the ear location and on skin of a subject, the wearable sensor including a wireless communication component, a power source and one or more processing components;
a subject device configured to receive information wirelessly from the wearable sensor, the subject device including:
one or more processors;
a memory coupled to the one or more processors, the memory storing computer-program instructions, that, when executed by the one or more processors, causes the performance of:
receiving or generating sensed brain activity signals or processed brain activity signals sensed by the wearable sensor;
predicting, using a trained prediction model, a predicted glucose state of the subject that is based at least partially on the sensed brain activity signals or the processed brain activity signals,
wherein predicting the predicted glucose state comprises predicting based on at least brain activity data from only a subset of frequencies within a larger range of frequencies that includes a delta band, a theta band, an alpha band, a beta band, and a gamma band, and
wherein predicting the predicted glucose state comprises at least one of,
predicting the glucose state at one or more future epochs of time that comprise a personal future epoch of time that has have greater predictive accuracy of glucose state for the subject than other future epochs of time for the subject,
predicting the glucose state with brain activity signals or processed brain activity signals from one or more personal wearable sensor locations on the subject's head that have greater predictive accuracy of the glucose state for the subject when the wearable sensor is properly worn than other wearable sensor locations on the subject's head when the wearable sensor is properly worn, or
predicting the glucose state with one or more brain signal frequencies that comprise personal brain signal frequencies that have greater predictive accuracy of glucose state for the subject than other brain signal frequencies; and
outputting instructions to initiate a communication indicative of the predicted glucose state based on and in response to the predicted glucose state.
21. The system of claim 20 , wherein the computer-program instructions are adapted to further cause the performance of,
in response to an on-demand glucose state prediction request that follows a period that is free of predicting a glucose state based on the sensed brain activity signals or the processed brain activity signals, predicting the predicted glucose state of the subject, wherein the on-demand glucose state prediction request is communicated by the subject to the subject device.
22. The system of claim 20 , wherein the computer-program instructions are adapted to train the trained prediction model, comprising:
receiving or generating the sensed brain activity signals or processed brain activity signals sensed by the wearable sensor;
providing or receiving personal sensed glucose data from a glucose monitor worn by the subject; and
based on the sensed brain activity signals or processed brain activity signals, and the personal sensed glucose data from the glucose monitor, performing at least one of
identifying or determining the one or more personal future epochs of time that have greater predictive accuracy of glucose state than other future epochs of time for the subject,
identifying or determining the one or more personal wearable sensor locations on the subject's head that have greater predictive accuracy of the glucose state for the subject than other wearable sensor locations on the subject's head when the wearable sensor is properly worn, or
identifying or determining the personal brain signal frequencies that have greater predictive accuracy of glucose state for the subject than other brain signal frequencies.
23. The system of claim 20 , wherein the computer-program instructions further cause the performance of comparing the predicted glucose state to sensed glucose values sensed from a continuous glucose monitor adapted to sense interstitial glucose and worn by the subject at a location different than the wearable sensor, and initiate instructions to cause the continuous glucose monitor to be calibrated based on the predicted glucose state.
24. A system adapted for predicting a glucose state of a subject, comprising:
a wearable non-invasive brain activity signal sensor (“wearable sensor”) that includes a plurality of sensing electrodes and a reference electrode, the wearable sensor sized and configured to be secured at a behind the ear location and on skin of a subject, the wearable sensor including a wireless communication component, a power source and one or more processing components;
a subject device configured to receive information wirelessly from the wearable sensor, the subject device including:
one or more processors;
a memory coupled to the one or more processors, the memory storing computer-program instructions, that, when executed by the one or more processors, cause the performance of:
receiving or generating sensed brain activity signals or processed brain activity signals sensed by the wearable sensor;
predicting, using a trained prediction model, a predicted glucose state of the subject that is based at least partially on the sensed brain activity signals or the processed brain activity signals,
wherein predicting the predicted glucose state comprises predicting based on at least brain activity data from only a subset of frequencies within a larger range of frequencies that includes a delta band, a theta band, an alpha band, a beta band, and a gamma band, and
wherein predicting the predicted glucose state further comprises predicting the glucose state at one or more future epochs of time, wherein the one or more future epochs of time comprise a personal future epoch of time for the subject that has greater predictive accuracy of glucose state than other future epochs of time for the subject; and
outputting instructions to initiate a communication indicative of the predicted glucose state based on and in response to the predicted glucose state.
25. The system of claim 24 , wherein the trained model is a model trained on at least personal non-invasively sensed brain activity signals from a behind the ear location and on personal glucose values of the subject.
26. The system of claim 24 , wherein predicting the predicted glucose state further comprises predicting real-time blood glucose values of the subject.