Accuracy of continuous glucose sensors

1 . A continuous glucose monitoring system comprising:

a continuous glucose device comprising:

a glucose sensor configured to generate raw output comprising a plurality of data sets;

a communication connection configured to communicate with a portable device;

at least one processing unit; and

a non-transitory memory coupled to the at least one processing unit;

wherein the memory is configured to store instructions that, when executed by the at least one processing unit, cause the at least one processing unit to:

preprocess the plurality of data sets, and

determine an estimated blood glucose value from the preprocessed plurality of data sets based at least in part on one or more interstitial fluid glucose proportionality coefficients and a time derivative of the raw output; and

wirelessly send the estimated blood glucose value to the portable device via the communication connection;

a glucose monitoring center configured to further process the estimated blood glucose value,

wherein the estimated blood glucose value is indicative of a real-time estimated blood glucose value; and

wherein the one or more interstitial fluid glucose proportionality coefficients account at least in part for a difference between a blood glucose value and an interstitial fluid glucose value caused by a physiological time lag and wherein the interstitial glucose proportionality coefficients are determined prior to a first calibration,

wherein the further processed estimated blood glucose value is indicative of a retrospective estimated blood glucose value.

2 . The continuous glucose monitoring system of claim 1 , wherein the instructions, when executed by the at least one processing unit, cause the at least one processing unit to preprocess the plurality of data sets by regressing or filtering the plurality of data sets to remove noise and random spikes.

3 . The continuous glucose monitoring system of claim 1 , wherein the instructions, when executed by the at least one processing unit, cause the at least one processing unit to determine the time derivative of the raw output.

4 . The continuous glucose monitoring system of claim 1 , wherein the plurality of data sets is indicative of a glucose value in an interstitial fluid at a plurality of sample times, and wherein the estimated blood glucose value is indicative of a blood glucose value at the plurality of sample times.

5 . The continuous glucose monitoring system of claim 1 , wherein the glucose sensor is coupled to the at least one processing unit and the memory.

6 . The continuous glucose monitoring system of claim 1 , wherein the at least one processing unit and the memory are disposed in a portable computing device in wired or wireless communication with the glucose sensor.

7 . A method of improving accuracy of a continuous glucose sensor by accounting for a difference caused by a physiological time lag between a blood glucose value and an interstitial fluid glucose value, the method comprising:

receiving raw output from the continuous glucose sensor, wherein the raw output comprises a plurality of data sets;

preprocessing the plurality of data sets; and

determining an estimated blood glucose value from the preprocessed plurality of data sets based at least in part on one or more interstitial fluid glucose proportionality coefficients and a time derivative of the raw output;

wherein the estimated blood glucose value is indicative of a real-time estimated blood glucose value; and

wherein the one or more interstitial fluid glucose proportionality coefficients account at least in part for a difference between a blood glucose value and an interstitial fluid glucose value caused by a physiological time lag and wherein the interstitial glucose proportionality coefficients are determined prior to a first calibration.

8 . The method of claim 7 , wherein preprocessing the plurality of data sets comprises regressing or filtering the plurality of data sets to remove noise and random spikes.

9 . The method of claim 8 , wherein preprocessing the plurality of data sets further comprises applying a filter based upon a clinically-observed blood glucose rate-of-change and Bayesian estimation techniques.

10 . The method of claim 8 , wherein preprocessing the plurality of data sets further comprises applying Kalman filter-based methods.

11 . The method of claim 7 , further comprising determining the time derivative of the raw output.

12 . The method of claim 11 , wherein determining the time derivative of the raw output comprises applying a short-interval polynomial smoothing technique using exponential weights to the plurality of data sets.

13 . The method of claim 7 , wherein the estimated blood glucose value is determined retrospectively.

14 . The method of claim 7 , wherein the plurality of data sets is indicative of a glucose value in an interstitial fluid at a plurality of sample times, and wherein the estimated blood glucose value is indicative of a blood glucose value at the plurality of sample times.

15 . The method of claim 7 , wherein the estimated blood glucose value is determined by a portable computing device in wired or wireless communication with the glucose sensor.