System and methods for processing analyte sensor data
Systems and methods for processing sensor analyte data, including initiating calibration, updating calibration, evaluating clinical acceptability of reference and sensor analyte data, and evaluating the quality of sensor calibration. During initial calibration, the analyte sensor data is evaluated over a period of time to determine stability of the sensor. The sensor may be calibrated using a calibration set of one or more matched sensor and reference analyte data pairs. The calibration may be updated after evaluating the calibration set for best calibration based on inclusion criteria with newly received reference analyte data. Fail-safe mechanisms are provided based on clinical acceptability of reference and analyte data and quality of sensor calibration. Algorithms provide for optimized prospective and retrospective analysis of estimated blood analyte data from an analyte sensor.
1. A method for calibrating an analyte sensor, the method comprising:
receiving sensor data from an analyte sensor, including one or more sensor data points;
receiving reference data, including one or more reference data points;
providing one or more matched data pairs by matching a reference data point to a substantially time corresponding sensor data point;
forming a calibration set including one or more matched data pairs;
forming a conversion function based at least in part on the calibration set;
modifying the conversion function using a modified calibration set including one or more matched data pairs, wherein the modified calibration set is not the same as the calibration set; and
converting, using a processor module, the sensor data into calibrated sensor data using the modified conversion function.
2. The method of claim 1 , wherein the modifying the conversion function is responsive to a clinical acceptability analysis.
3. The method of claim 2 , wherein the clinical acceptability analysis comprises using a clinical cost function.
4. The method of claim 3 , wherein the clinical cost function comprises a Clarke Error Grid, a Consensus Grid or a mean absolute relative difference.
5. The method of claim 1 , wherein the modifying the conversion function is responsive to a statistical association of the one or more matched data pairs in the calibration set.
6. The method of claim 5 , wherein the statistical association is determined based on a cost function.
7. The method of claim 5 , wherein the statistical association is determined by evaluating a correlation of the one or more matched data pairs of the calibration set with a regression line formed from the one or more matched data pairs of the calibration set.
8. The method of claim 1 , wherein receiving sensor data comprises receiving the sensor data from a continuous glucose sensor.
9. The method of claim 1 , wherein receiving reference data comprises receiving reference data from an in vitro blood glucose monitor.
10. The method of claim 1 , wherein receiving reference data comprises downloading reference data via a wireless connection.
11. The method of claim 1 , wherein receiving reference data from a reference analyte monitor comprises receiving within a receiver an internal communication from a reference analyte monitor integral with the receiver.
12. The method of claim 1 , wherein the calibration set comprises a single matched data pair.
13. The method of claim 1 , wherein the calibration set comprises a plurality of matched data pairs.