CALIBRATION METHOD FOR THE PROSPECTIVE CALIBRATION OF MEASURING EQUIPMENT

• US 20160070675A1
• Filed: 11/18/2015
• Published: 03/10/2016
• Est. Priority Date: 08/27/2009
• Status: Active Application

• Current Assignee: Roche Diabetes Care Inc. (Roche Holding AG)
• Original Assignee: Roche Diabetes Care Inc. (Roche Holding AG)
• Inventors: Schmitt, Nikolaus, Frisch, Gerhard
• Granted Patent: US 10,289,652 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: A61B 2560/0223 of calibration, e.g. protocols for calibrating sensors
• CPC Class Codes: A61B 5/14532 for measuring glucose, e.g. by tissue impedance measurement
• CPC Class Codes: A61B 5/1495 Calibrating or testing of in-vivo probes
• CPC Class Codes: A61B 5/7207 of noise induced by motion artifacts
• CPC Class Codes: A61B 5/7232 involving compression of the physiological signal, e.g. to extend the signal recording period
• CPC Class Codes: G01N 2035/00702 Curve-fitting; Parameter matching; Calibration constants
• CPC Class Codes: G01N 27/3271 Amperometric enzyme electrodes for analytes in body fluids, e.g. glucose in blood amperometry per se G01N27/49; aspects concerning the enzyme reagent C12Q1/001
• CPC Class Codes: G01N 33/66 involving blood sugars, e.g. galactose
• CPC Class Codes: G01N 35/00693 Calibration
• CPC Class Codes: G06F 17/18 for evaluating statistical data , e.g. average values, frequency distributions, probability functions, regression analysis forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04

Abstract

A method for operating measuring equipment for detecting at least one analyte in a bodily fluid by means of at least one continuously measuring blood glucose sensor. At least one calibration method is carried out for the prospective calibration of the measuring equipment. At least three calibration points are detected in the calibration method, wherein each calibration point comprises at least one measurement signal from the measuring equipment and at least one reference value of an associated reference measurement. A plurality of possible slopes are established between the calibration points. At least one robust estimation method, more particularly using a formation of at least one median, is used to determine at least one probable slope from the plurality of possible slopes. Furthermore, at least one measurement is carried out. During the measurement and using the probable slope, a concentration of the analyte in the bodily fluid is deduced from at least one measurement signal from the measuring equipment and the probable slope.

Claims

1. A method for operating measuring equipment for detecting at least one analyte in a bodily fluid with a continuously monitoring sensor, the method comprising:

detecting at least three calibration points, each calibration point comprising a measurement signal from the measuring equipment and a reference value of an associated reference measurement, wherein the reference value corresponds to an actual present value of concentration of the analyte in the bodily fluid;

establishing a plurality of possible slopes between the calibration points;

using a robust estimation method to determine a probable slope from the plurality of possible slopes, the robust estimation method comprising a statistical estimation method which supplies stable statistical estimators even if outliers occur or if distribution assumptions are only approximately valid, the robust estimation being based on one or more permutation algorithms and/or on one or more sorting algorithms; and

performing at least one measurement in which the probable slope is used with the measurement equipment to determine a concentration of the analyte in the bodily fluid.

2. The method of claim 1, further comprising:

determining through a plurality of the calibration points a plurality of straight lines having the probable slope;

determining the axis intercepts of the straight lines;

forming a median to determine a probable axis intercept from the axis intercepts; and

using the probable axis intercept during the measurement.

3. The method of claim 2, wherein the calibration method is carried out repeatedly at different times, the method further comprising determining a new probable slope and a new probable axis intercept after the calibration method has been repeated and then using the new probable slope and the new probable axis intercept in at least one subsequent measurement.

4. The method of claim 1, further comprising preforming plausibility analysis during the calibration and using the plausibility analysis to discard unrealistic calibration points and/or unrealistic slopes and/or unrealistic axis intercepts.

5. The method of claim 1, further comprising using a sorting algorithm and/or a permutation algorithm during the robust estimation to form a median.

6. The method of claim 1, further comprising determining at least one current calibration quality having at least one confidence interval.

7. The method of claim 1, further comprising automatically inviting a user to carry out the calibration method.

9. The method of claim 1, further comprising informing a user if a current measurement signal lies in a region of insufficient calibration quality.

10. The method of claim 1, further comprising discarding at regular or irregular intervals calibration points that are older than a predetermined time threshold.

11. The method of claim 1, wherein the calibration points and/or the measurement signals are subjected to at least one smoothing method, using an exponential filter.

12. The method of claim 1, wherein the calibration points and/or the measurement signals are subjected to at least one linearization step, wherein the linearization step produces an approximately linear relationship between the measurement signals and the reference values.

13. The method of claim 1, further comprising detecting a plurality of the measurement signals from the measuring equipment over a measurement time period and applying a first data reduction method, comprising:

combining the measurement signals with associated measurement times to form measurement value pairs, wherein a plurality of possible slopes are established between the measurement value pairs;

using a median to establish a probable slope from the plurality of possible slopes, wherein a plurality of straight lines with the probable slope through the measurement value pairs are formed;

selecting a representative time for the measurement time period and establishing functional values of the straight lines for the representative time;

establishing a probable functional value from the functional values; and

using the at least one probable functional value and the at least one representative time to replace the plurality of the measurement value pairs with a representative measurement value pair;

the method further comprising applying a second data reduction method, comprising;

forming a median and using the median to determine at least one probable measurement signal for the measurement time from the plurality of measurement signals, wherein the plurality of measurement signals are replaced by the probable measurement signal.

14. The method of claim 13, further comprising determining a standard deviation and/or a measure of the quantile.

15. The method of claim 13, further comprising comparing robust estimators of the measurement signals and parametric estimators of the measurement signals.

16. The method of claim 13, further comprising determining a trend from the plurality of measurement signals.

17. The method of claim 1, wherein, during the calibration method, a calibration of measurement signals from the measuring equipment is carried out against reference values of associated reference measurements and the calibration strategy comprises one or more of the following steps:

a region of measurement signals from the measuring equipment that can be used for a measurement is restricted depending on a current calibration quality, wherein information is provided to a user when detecting measurement signals outside of the region that is usable or covered by calibration values or outside of a tolerance region based on the calibration values;

a user is invited to carry out the calibration method again;

a user is informed that now is a good time for carrying out the calibration method; and

a user is invited to carry out the calibration method again at regular or irregular intervals or at predetermined times.

8. The method of claim 8, wherein the user is invited to carry out the calibration method if at least one current measurement signal lies in a region in which the calibration method was not yet carried out with sufficient calibration quality.

18. A computer readable storage medium storing one or more programs, the one or more programs comprising instructions, which when executed by a computer of measuring equipment for detecting an analyte in a bodily fluid, cause the computer to perform a method comprising the following steps:

detecting at least three calibration points, each calibration point comprising a measurement signal from the measuring equipment and a reference value of an associated reference measurement, wherein the reference value corresponds to an actual present value of concentration of the analyte in the bodily fluid;

establishing a plurality of possible slopes between the calibration points;

using a robust estimation method to determine a probable slope from the plurality of possible slopes, the robust estimation method comprising a statistical estimation method which supplies stable statistical estimators even if outliers occur or if distribution assumptions are only approximately valid, the robust estimation being based on one or more permutation algorithms and/or on one or more sorting algorithms; and

performing at least one measurement in which the probable slope is used with the measurement equipment to determine a concentration of the analyte in the bodily fluid.

19. Measuring equipment for detecting at least one analyte in a bodily fluid, comprising:

a continuously measuring bodily fluid analyte sensor;

a processor having a program stored in memory and configured for execution by the processor, the program including instructions for;

detecting at least three calibration points, each calibration point comprising a measurement signal from the measuring equipment and a reference value of an associated reference measurement, wherein the reference value corresponds to an actual present value of concentration of the analyte in the bodily fluid;

establishing a plurality of possible slopes between the calibration points;

using a robust estimation method to determine a probable slope from the plurality of possible slopes, the robust estimation method comprising a statistical estimation method which supplies stable statistical estimators even if outliers occur or if distribution assumptions are only approximately valid, the robust estimation being based on one or more permutation algorithms and/or on one or more sorting algorithms; and

performing at least one measurement in which the probable slope is used with the measurement equipment to determine a concentration of the analyte in the bodily fluid.

20. The measuring equipment of claim 19, wherein the bodily fluid analyte sensor comprises a glucose sensor.