Transcutaneous analyte sensors and monitors, calibration thereof, and associated methods

Abstract

To provide systems and methods for calibrating an analyte concentration sensor within a biological system, generally using only a signal from the analyte concentration sensor.SOLUTION: For example, at a steady state, an analyte concentration value within the biological system is known, and it may provide a source for calibration. Similar techniques may be employed with slow-moving averages. Variation forms are disclosed.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] A system and method for processing sensor data from a continuous analyte sensor and for calibration of the sensor. [Background technology]

[0002] Diabetes mellitus is a disease in which the pancreas cannot make enough insulin (Type I or insulin-dependent) and / or insulin is not effective (Type 2 or non-insulin-dependent). In the diabetic state, the victim suffers from hyperglycemia, which can cause many physiological disorders associated with the deterioration of small blood vessels, such as kidney failure, skin ulcers, or bleeding into the vitreous humor of the eye. Hypoglycemic reactions (low blood sugar) can be induced by inadvertent overdosing of insulin or after normal administration of insulin or glucose-lowering drugs accompanied by abnormal exercise or inadequate food intake.

[0003] Traditionally, people with diabetes carry self-monitoring blood glucose (SMBG) monitors, which typically require an uncomfortable finger-prick method. Due to lack of comfort and convenience, people with diabetes typically only measure their glucose levels two to four times per day. Unfortunately, such time intervals are spread so far apart that people with diabetes may not know about a hyperglycemic or hypoglycemic condition until it is too late, sometimes resulting in dangerous side effects. Glucose levels may alternatively be continuously monitored by a sensor system that includes an on-skin sensor assembly. The sensor system may have a wireless transmitter that transmits measurement data to a receiver, which can process and display information based on the measurements.

[0004] To date, various glucose sensors have been developed to continuously measure glucose levels. Many implantable glucose sensors suffer from internal complications and provide only short-term and inaccurate detection of blood glucose. Similarly, transcutaneous sensors have faced problems in accurately detecting and reporting back continuous glucose levels over extended periods of time. Efforts have been made to obtain blood glucose data from implantable devices and retrospectively determine blood glucose trends for analysis, but these efforts do not aid the diabetic patient in determining real-time blood glucose information. Efforts have also been made to obtain blood glucose data from transcutaneous devices for predictive data analysis, but similar problems have arisen.

[0005] In a continuous glucose monitor (CGM), after the sensor is implanted, it is calibrated, after which it provides substantially continuous sensor data to the sensor electronics. The sensor electronics converts the sensor data so that an estimated analyte value can be continuously provided to the user. As used herein, the terms "substantially continuous," "continuously," and the like refer to a data stream of individual measurements taken at spaced intervals, which may range from less than a second up to, for example, 1, 2, or 5 minutes or more. As the sensor electronics continue to receive sensor data, the sensor may be recalibrated from time to time to account for possible changes (drift) in sensor sensitivity and / or baseline. Sensor sensitivity may refer to the amount of current generated in the sensor by a given amount of measured analyte.

[0006] The sensor baseline refers to the signal output by the sensor when no analyte is detected. Over time, the sensitivity and baseline can change due to various factors, including cellular attack or migration to the sensor, which can affect the ability of the analyte to reach the sensor.

[0007] This Background is provided to introduce a brief context for the Summary and Detailed Description that follow. This Background is not intended to aid in determining the scope of the claimed subject matter, nor is it to be considered as limiting the claimed subject matter to implementations that solve any or all of the disadvantages or problems discussed above. Summary of the Invention [Means for solving the problem]

[0008] Without limiting the scope of the present embodiments as expressed by the following claims, salient features of systems and methods in accordance with the present principles will now be briefly discussed. After considering this discussion, and particularly after reading the section entitled "Description of the Preferred Embodiments," one will understand how the features of the present embodiments provide the advantages described herein.

[0009] In a first aspect, a method is provided for calibrating an analyte concentration sensor in a biological system using only a signal from the analyte concentration sensor, where upon the occurrence of a repeatable event, an analyte concentration value in the biological system is known, the method comprising detecting, on a monitoring device, when an analyte concentration value measured by an analyte concentration sensor placed in the biological system constitutes a first repeatable event, and correlating, on the monitoring device or on a device or server operatively coupled to the monitoring device, a measurement of the analyte concentration value when the biological system is at the detected first repeatable event with the known analyte concentration value.

[0010] Implementations of the embodiments and aspects may include one or more of the following: The correlation may include determining a functional relationship between the sensor readings and the known analyte concentration values. The functional relationship may include a multiplicative constant. The detection may include waiting a predetermined time following entry of an event, such as a meal or exercise, on the monitoring device. The method may further include detecting an occurrence of a second repeatable event following the correlation, the second repeatable event being different from the first repeatable event, and recalibrating the analyte concentration sensor by correlating the sensor readings when the biological system is in the detected second repeatable event to the known analyte concentration values. The sensor readings may have a first raw value at the initial calibration and a second raw value at the recalibration, the first and second raw values ​​being different. The method may further include, following the correlation, displaying a graph or table showing the currently measured and historical values ​​of the analyte concentration calibrated at least in part based on the correlation, and, following the recalibration, updating the display of the graph or table showing the currently measured and historical values ​​of the analyte concentration according to the recalibration. The updating may change the display of the historical values ​​of the analyte concentration. The method may further include determining a difference between the first raw value and the second raw value, comparing a quantity based on the difference to a predetermined standard, and determining whether the sensor calibration is drifting based on the comparison. The method may further include determining a quantitative amount by which the sensor calibration is drifting. The method may further include adjusting the sensor calibration based on the determined quantitative amount. The quantity may be a slope between the first raw value and the second raw value. The method may further include, if the slope exceeds a predetermined threshold, inhibiting future calibrations based on a steady state until the slope no longer exceeds the predetermined threshold. The method may further include prompting a user to input a measurement. The sensor may be a glucose sensor. The method may further include receiving a signal from the sensor following correlation and displaying a value corresponding to the received signal, the displayed value being based on the received signal and the known analyte concentration value.The method may further include determining the known analyte concentration value by prompting a user to input a measurement. The method may further include determining the known analyte concentration value by accessing a population average. Recalibration may be configured to occur when the sensor readings are substantially stable or within a predetermined range of readings for a threshold period of time, thereby reducing the occurrence of unexpected spikes in readings, and such recalibration at such times may be triggered or configured to occur in any of the described embodiments or aspects.

[0011] In a second aspect, a method of compensating for drift of an analyte concentration sensor in a biological system using only a signal from the analyte concentration sensor is provided, comprising measuring an analyte value using an in-dwelling analyte concentration sensor; determining a first slow moving average of the analyte measurement over a first period of time; referencing a calibration of the sensor based at least in part on the first slow moving average; following the first determination, determining a second slow moving average of the analyte measurement over a second period of time; and adjusting the calibration of the sensor based at least in part on a difference between the first slow moving average and the second slow moving average.

[0012] Implementations of aspects and embodiments may include one or more of the following: The duration of the first time period may be greater than about 12 hours or greater than about 24 hours. The duration of the first time period may be the same as the duration of the second time period. The method may further include displaying a graph or table showing at least historical values ​​of the analyte concentration calibrated at least in part based on the first slow moving average following referencing the calibration of the sensor based at least in part on the first slow moving average, and updating the display of the graph or table showing at least historical values ​​of the analyte concentration according to the adjusted calibration following the adjustment. The updating may change the display of the historical values ​​of the analyte concentration. The displayed graph or table may further show a currently measured value of the analyte concentration. Referencing the calibration of the sensor based at least in part on the first slow moving average may further include basing the calibration on a seed value, such as a seed value received from a population average or from a previous session. The method may further include changing the seed value following the adjustment, at least in part based on the adjustment. The method may further include varying the seed value based on a difference between the first slow moving average and the second slow moving average.

[0013] In a third aspect, a method of compensating for drift of an analyte concentration sensor in a biological system using only a signal from the analyte concentration sensor is provided, comprising: measuring a value of an analyte using an in-dwelling analyte concentration sensor; determining a first slow moving average of the analyte measurement over a first period of time; basing a first apparent sensitivity of the sensor at least in part on the first slow moving average; following the first determination, determining a second slow moving average of the analyte measurement over a second period of time; referencing a second apparent sensitivity of the sensor based at least in part on the second slow moving average; determining whether a change in apparent sensitivity of the sensor between the first apparent sensitivity and the second apparent sensitivity matches a predetermined criterion; if the change in apparent sensitivity matches the predetermined criterion, adjusting the actual sensitivity of the sensor to an adjusted value based on the difference between the first apparent sensitivity and the second apparent sensitivity; and if the change in apparent sensitivity does not match the predetermined criterion, prompting a user to input data so that a change in apparent sensitivity may be determined.

[0014] Implementations of aspects and embodiments may include one or more of the following: The determining may include determining whether the change in apparent sensitivity is due to a sensitivity drift or a change in the slow moving average. If the change in apparent sensitivity is due to a change in the slow moving average, the method may further include prompting the user to input data related to the change. The predetermined criteria may include known behavior regarding sensitivity change over time for the sensor. The known behavior regarding sensitivity change may constitute an envelope of acceptable sensitivity change over time. The adjusted value may be based at least in part on the second slow moving average. The predetermined criteria may further include known values ​​regarding physiologically likely analyte changes. The prompting of the user may include prompting the user to input a calibration value. The prompting of the user may include prompting the user to input meal or exercise information. Upon receiving the calibration value or meal or exercise information from the user, the method may further include determining whether the change in apparent sensitivity is due to a sensitivity drift or a change in the slow moving average. The method may further include adjusting the actual sensitivity of the sensor based on the received calibration value or meal or exercise information.

[0015] In a fourth aspect, a method is provided for checking the calibration of an analyte concentration sensor system in a biological system using only signals from the analyte concentration sensor, comprising: measuring analyte values ​​over time using an indwelling analyte concentration sensor after an initial calibration; calculating a clinical value of the analyte concentration based on the measurements and the initial calibration; adjusting the initial calibration to an updated calibration, where the updated calibration is based on only the analyte measurements over time or a subset thereof; and calculating the clinical value of the analyte concentration based on the measurements and the updated calibration.

[0016] Implementations of aspects and embodiments may include one or more of the following: The initial calibration may be based on a population average or data entered by a user. The adjustment may be based on a slow moving average of measurements of the analyte over time. The adjustment may be based on a steady state value of the analyte. The initial calibration may be based on data determined prior to a session associated with the in-dwelling analyte concentration sensor. The data may be determined a priori, on the bench, or in vitro.

[0017] In a fifth aspect, there is provided a method of calibrating an analyte concentration sensor in a biological system using a signal from the analyte concentration sensor, where at a steady state an analyte concentration value in the biological system is known, the method comprising: receiving on a monitoring device a seed value for calibration parameters; detecting on the monitoring device when an analyte concentration value measured by an analyte concentration sensor placed in the biological system is at the steady state; correlating on the monitoring device or on a device or server operatively coupled to the monitoring device the measured analyte concentration value when the biological system is at the detected steady state with the known analyte concentration value; following the correlation, receiving a signal from the sensor; and calculating and displaying a value corresponding to the received signal, the calculated value being based on the received signal, the known analyte concentration value, and the seed value.

[0018] Implementations of aspects and embodiments may include one or more of the following: The received seed value may be received from a source including factory calibration information. The method may further include detecting behavior in the received signal other than a predetermined parameter and prompting the user to input external calibration information. The displayed value may be further based on the external calibration information. The external calibration information may be received from an SMBG or fingerstick calibration. The method may further include resetting the known calibration value to a new known calibration value, the resetting being based at least in part on the external calibration information. The method may further include resetting the seed value to a new seed value, the resetting being based at least in part on the external calibration information. The method may further include modifying the display based on the determined accuracy of the value. The modifying the display may include displaying a range rather than a value, or vice versa. The received seed value of the calibration parameter may be a user-entered medical condition characteristic. The user-entered medical condition characteristic may include an indication of type I diabetes, type II diabetes, non-diabetes, or pre-diabetes. The received seed value of the calibration parameter may be a value based on one or more user-entered blood glucose values. Displaying the values ​​corresponding to the received signals may include displaying a graph or table showing currently measured and historical values ​​of the analyte concentration, and further includes detecting that a calibration change is occurring, adjusting one or more calibration parameters of the analyte concentration sensor in accordance with the calibration change, and updating the display of the graph or table showing currently measured and historical values ​​of the analyte concentration in accordance with the adjusted calibration parameters following the adjustment. Detecting that a calibration change is occurring may include detecting a change in a slow moving average or detecting a change in a steady state value.

[0019] In a sixth aspect, a method of calibrating an analyte concentration sensor is provided, using only parameters based on or derivable from the sensor signal following insertion of the sensor into a patient, comprising receiving at least an initial value of analyte concentration and an initial value or initial value distribution of sensor sensitivity following insertion of the analyte concentration sensor, monitoring a signal from the sensor over a duration of time following insertion of the analyte concentration sensor, calculating a plurality of analyte concentration values ​​based on the initial value or value distribution of the monitored sensor signal and the sensor sensitivity over the duration of time, determining a value distribution of the monitored signal over the duration of time, optimizing the initial value or value distribution of the sensor sensitivity and the plurality of analyte concentration values ​​to match the value distribution of the monitored signal, and determining an updated sensitivity based on the optimization.

[0020] Implementations of aspects and embodiments may include one or more of the following: The receiving may be receiving an initial value distribution of sensor sensitivity, and the calculation of the plurality of analyte concentration values ​​may be based on a representative value from the monitored sensor signal and the initial value distribution of sensor sensitivity. The representative value may be selected from a mean or a midpoint or median. The determining of the updated sensitivity may further include dividing the representative value by the initial value of the analyte concentration, and updating the value of the sensitivity to be equal to a result of the division. The initial value of the analyte population may be a population mean, may be entered by a user, or may be transferred from a previous session. The optimizing may include optimizing a product of the initial value or value distribution of the sensor sensitivity and the plurality of analyte concentration values. Optimizing the product may include optimizing the product to match the value distribution of the monitored signal, while adjusting parameters of the value distribution of the sensor sensitivity and the plurality of analyte concentration values ​​to most closely match the corresponding population means. The receiving may further include receiving an initial value distribution of the baseline, and the optimizing may further include optimizing the value distribution of the baseline, along with the value distribution of the sensor sensitivity and the plurality of analyte concentration values, to match the value distribution of the monitored signal. The initial value distribution of the baseline may follow a normal distribution. At least the initial value of the analyte concentration may be used as part of a seed value input to the slow moving average filter. The initial value distribution of the sensor sensitivity may be defined by a normal distribution. The determined value distribution of the monitored signal may follow a log-normal distribution. The duration may be one day. The method may further include continuing to determine an updated sensitivity based on a previous updated sensitivity and the received analyte concentration values. The method may further include detecting a slow moving average of the monitored analyte concentration values. If the absolute value of the change in the slow moving average is greater than a predetermined threshold value over a predetermined unit time, the method may include prompting a user to input data. If the absolute value of the change in the slow moving average is greater than a predetermined threshold value over a predetermined unit time, the method may include determining whether the change is due to a system error or due to a change in the actual sensitivity of the sensor.Determining whether the change is due to a systematic error or due to a change in the actual sensitivity of the sensor may include determining whether the subsequent behavior of the sensitivity is consistent with a known sensitivity profile including an envelope of a sensitivity curve. If the absolute value of the change in the slow moving average is determined to be due to a change in the actual sensitivity of the sensor, the method may include updating the sensitivity based at least in part on the value of the change in the slow moving average. Determining whether the change is due to a systematic error or due to a change in the actual sensitivity of the sensor may include determining whether the subsequent behavior of the analyte concentration values ​​is consistent with a known envelope of physiological feasibility. If the absolute value of the change in the slow moving average is determined to be due to a systematic error, the method may include prompting a user to input data.

[0021] In a seventh aspect, a method of calibrating an analyte concentration sensor in a biological system using a signal from the analyte concentration sensor is provided, comprising receiving or determining a seed value for a calibration parameter associated with the analyte concentration sensor, determining at least in part a calibration of the analyte concentration sensor using the seed value, measuring an analyte concentration using the analyte concentration sensor, and displaying a measurement value calibrated at least in part using the seed value.

[0022] Implementations of aspects and embodiments may include one or more of the following: The receiving or determining may be performed on a monitoring device in operative signal communication with the analyte concentration sensor. The displaying may be performed on the monitoring device or on a mobile device in signal communication with the monitoring device. The displaying of the measurements may include displaying a graph or table showing at least the historical values ​​of the analyte concentration and may further include detecting that a change in calibration is occurring, adjusting one or more calibration parameters of the analyte concentration sensor in accordance with the detected change in calibration, and updating the display of the graph or table showing at least the historical values ​​of the analyte concentration in accordance with the adjusted calibration parameters following the adjustment. The updating may change the display of the historical values ​​of the analyte concentration. The seed value may be based at least in part on a code. The code may be entered into the monitoring device by a user. The monitoring device may be configured to receive the code without substantial user involvement. The seed value may be based at least in part on an impedance measurement. The seed value may be based at least in part on information associated with a manufacturing lot of the sensor. The seed value may be based at least in part on a population average. The seed value may be based at least in part on a previous analyte value of the user.

[0023] In an eighth aspect, a method is provided for calibrating and compensating for drift of an in-dwelling analyte concentration sensor in a biological system using only a signal from the analyte concentration sensor, where at a steady state an analyte concentration value in the biological system is known, the method comprising: detecting on a monitoring device when an analyte concentration value measured by an analyte concentration sensor in-dwelling in the biological system is at steady state; correlating on the monitoring device or on a device or server operatively coupled to the monitoring device a measurement of the analyte concentration value when the biological system is at the detected steady state with said known analyte concentration value; determining a first slow moving average of the analyte measurement value over a first period of time; referencing a calibration of the sensor based at least in part on the first slow moving average and the known analyte concentration value; following the first determination, determining a second slow moving average of the analyte measurement value over a second period of time; and adjusting the calibration of the sensor based at least in part on a difference between the first slow moving average and the second slow moving average.

[0024] In a ninth aspect, a method of calibrating a first portion of a number of sensors, a second portion being in use, is provided, comprising receiving calibration data from some of the second portion of the sensors, and updating one or more calibration parameters of the first portion based on the received data.

[0025] Implementations of aspects and embodiments may include one or more of the following: The update may be performed before the first portion is incorporated into the user. The update may be performed after the first portion is incorporated into the user. The update may be performed by transmitting new or updated calibration parameters over a network to a sensor electronics module associated with the monitoring device or sensor. The second portion of the sensor may be configured to be calibrated using an a priori calibration. The second portion of the sensor may be configured to be calibrated using user data. The second portion of the sensor may be configured to be calibrated using an ex vivo bench calibration. The second portion of the sensor may be configured to be calibrated using blood measurements.

[0026] In a tenth aspect, a method of compensating for drift of an analyte concentration sensor in a biological system using only a signal from the analyte concentration sensor is provided, comprising measuring a value of an analyte as a function of time using an in-dwelling analyte concentration sensor, filtering the measurements using a double exponential smoothing filter, and displaying the filtered measurements versus time following filtering.

[0027] Implementations of aspects and embodiments may include one or more of the following: The double exponential smoothing filter may be governed by the equations described herein. The subsequent glucose signal as a function of time may be provided by the equations described herein.

[0028] In an eleventh aspect, there is provided a method of calibrating an analyte concentration sensor in a biological system using only a signal from the analyte concentration sensor, where analyte concentration values ​​in the biological system are known at or during the occurrence of a repeatable event, the method comprising detecting, on a monitoring device, when a set of analyte concentration values ​​measured by an analyte concentration sensor placed in the biological system constitute a repeatable event, and correlating, on the monitoring device or on a device or server operatively coupled to the monitoring device, the set of analyte concentration values ​​at the time of the repeatable event with the known analyte concentration values.

[0029] Implementations may include being selected from the group consisting of steady state, post-prandial rise, daily high and low glucose ranges, decay rate, or rate of change.

[0030] In a twelfth aspect, a method of compensating for drift in an analyte concentration sensor in a biological system using only a signal from the analyte concentration sensor is provided, comprising: measuring an analyte value using an in-dwelling analyte concentration sensor; determining a first slow moving average of the analyte measurement value over a first set of periods, the first set including event-based periods; referencing a calibration of the sensor based at least in part on the first slow moving average; following the first determination, determining a second slow moving average of the analyte measurement value over a second set of periods, the second set including event-based periods; and adjusting the calibration of the sensor based at least in part on a difference between the first slow moving average and the second slow moving average.

[0031] Implementations may include one or more of the following: The first and second event-based periods may be selected from the group consisting of a post-meal period, a sleep period, and a post-breakfast period.

[0032] In a thirteenth aspect, a method of calibrating an analyte concentration sensor in a biological system is provided, comprising determining a sensitivity profile over time for a set of sensors of a type, measuring the sensitivity profile for each sensor of the type, measuring an electrical characteristic of a transmitter, reading an identifier of the sensor, receiving data corresponding to the sensitivity of the sensor, and storing the identifier and the received data on the transmitter.

[0033] The implementation may include one or more of the following: The set of sensors of a type may correspond to a set of sensors in a lot. The method may further include packaging the individual sensors with the transmitter as a kit. The method may further include coupling the transmitter to a mobile device running a monitoring application. The method may further include calibrating the transmitter and the sensor using the monitoring application. The calibration may be with respect to a measured electrical characteristic of the transmitter. The monitoring application may be configured to initiate a sensor session with a signal from the transmitter, the signal detecting that the transmitter is coupled to the sensor. The transmitter may be configured to initiate the sensor session when the transmitter detects coupling to the sensor. The method may further include coupling the transmitter to a mobile device running the monitoring application. The method may further include receiving a representative set of measured analyte values. The method may further include using the received representative set of measured analyte values, or a subset thereof, to determine seed parameters for a forward filter, a reverse filter, or both. The seed values ​​may be determined using a signal median, a drift value, or both. Both forward and reverse filters may be used, and the method may further include optimizing a seed value to minimize the mean square error between the two signal filters. The method may further include adjusting the sensitivity and baseline for the sensor according to a signal-based calibration algorithm, which uses an average of the signals from the forward and reverse filters along with the raw sensor signal. The method may further include adjusting the sensitivity and baseline based on one or more criteria. The criteria may include that the average glucose value should be consistent with a predicted diabetes average value. The criteria may include that the CGM glucose variability should be consistent with the average glucose level.The method may further include detecting an amount of sensor change, determining that the amount of sensor change exceeds a threshold criterion, and preventing display of the reading, whereby the potentially inaccurate reading is not displayed to the user.

[0034] In a fourteenth aspect, a method of compensating for drift of an analyte concentration sensor in a biological system using only a signal from the analyte concentration sensor is provided, comprising: measuring an analyte value using an in-dwelling analyte concentration sensor; determining a first slow moving average of the analyte measurement value over a first period of time; referencing a calibration of the sensor based at least in part on the first slow moving average; following the first determination, determining a second slow moving average of the analyte measurement value over a second period of time; and adjusting the calibration of the sensor based at least in part on a seed value and a difference between the first slow moving average and the second slow moving average.

[0035] Implementations may include one or more of the following: A seed value may be determined using the signal median, drift value, or both. Both forward and reverse filters may be used, and the method may further include optimizing the seed value to minimize the mean square error between the two signal filters. The method may further include adjusting the sensitivity and baseline for the sensor according to a signal-based calibration algorithm, which uses an average of the signals from the forward and reverse filters along with the raw sensor signal. The method may further include adjusting the sensitivity and baseline based on one or more criteria. The criteria may include that the average glucose value should be consistent with a predicted diabetes average value. The criteria may include that the CGM glucose variability should be consistent with the average glucose level.

[0036] In further aspects and embodiments, the above method features of the various aspects are described in terms of systems as in the various aspects configured to implement the method features. Any of the features of the embodiments of any of the aspects, including but not limited to any of the embodiments of any of the first to fourteenth aspects mentioned above, can be applied to all other aspects and embodiments identified herein, including but not limited to any of the embodiments of any of the first to fourteenth aspects mentioned above. Furthermore, any of the features of the embodiments of the various aspects, including but not limited to any of the embodiments of any of the first to fourteenth aspects mentioned above, can be independently combined in any manner, partially or in whole, with other embodiments described herein, e.g., one, two, three or more embodiments can be combined in whole or in part. Furthermore, any of the features of the embodiments of the various aspects, including but not limited to any of the embodiments of any of the first to fourteenth aspects mentioned above, can be made optional with respect to other aspects or embodiments. Any aspect or embodiment of the method may be performed by a system or apparatus of another aspect or embodiment, and any aspect or embodiment of the system or apparatus may be configured to perform a method of another aspect or embodiment, including, but not limited to, any embodiment of any of the first to fourteenth aspects referred to above.

[0037] This Summary is provided to introduce a selection of concepts in a simplified form. The concepts are further described in the Detailed Description. Elements or steps other than those described in this Summary are contemplated, and no element or step is necessarily required. This Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended for use as an aid in determining the scope of the claimed subject matter. The claimed subject matter is not limited to implementations that solve any or all of the disadvantages noted in any part of this disclosure.

[0038] The present embodiments will now be discussed in detail with an emphasis on highlighting the advantageous features. These embodiments illustrate novel and non-obvious sensor signal processing and calibration systems and methods that are illustrated in the accompanying drawings, which are for illustrative purposes only and are not to scale, but rather to emphasize the principles of the disclosure. These drawings include the following figures, in which like numerals indicate like parts: [Brief description of the drawings]

[0039] [Figure 1] FIG. 1 is a schematic diagram of a continuous analyte sensor system attached to a host and in communication with multiple exemplary devices. [Diagram 2] FIG. 2 is a block diagram showing the electronics associated with the sensor system of FIG. 1. [Diagram 3] 1 shows a graph illustrating the linear relationship between measured sensor counts and analyte concentration. [Figure 4] shows an exemplary change in sensitivity over time. [Diagram 5] 1 illustrates various ways to provide factory information about the sensor to the transmitter electronics. [Figure 6] 13 illustrates a "no code" option for providing factory information about the sensor to the sensor electronics. [Figure 7] 13 illustrates an option to provide factory information about the sensor to the sensor electronics without explicit user input. [Figure 8] 13 illustrates an option to provide factory information about the sensor to the sensor electronics with user input. [Figure 9] 1 illustrates the use of one portion of a number of sensors, acquiring field data, to calibrate another portion of a number of sensors. [Figure 10] 1 illustrates the use of one portion of a number of sensors, acquiring field data, to calibrate another portion of a number of sensors. [Figure 11]11 illustrates an exemplary method in accordance with the present principles, and in particular, a flow chart for implementing the method in accordance with FIGS. [Figure 12] FIG. 1 is a schematic diagram of a sensor and transmitter within a host in communication with a receiver and / or smartphone. [Figure 13] 4 is a flow chart illustrating another exemplary method in accordance with the present principles. [Figure 14] 1 is a graph showing analyte concentration over time before changes in sensitivity. [Figure 15] 1 is a graph showing analyte concentration over time following a change in sensitivity. [Figure 16] 1 is a modular diagram of an analyte concentration measurement system according to the present principles. [Figure 17] 4 is a flow chart illustrating another exemplary method in accordance with the present principles. [Figure 18] 11 is a flow chart illustrating another exemplary method in accordance with the present principles for performing a calibration using steady state. [Figure 19] 1 is a graph showing two calibration lines before and after drift has occurred. [Figure 20] 1 shows a slow moving average of the sensor count over time. [Figure 21] 1 shows a slow moving average of the sensor count over time. [Figure 22] 11 is a flow chart illustrating another exemplary method in accordance with present principles using a slow moving average. [Figure 23A] 11 is a flow chart illustrating another exemplary method in accordance with the present principles illustrating updating of history values. [Figure 23B] 1 is a chart showing sensitivity data over an extended sensor session, showing a characteristic drift. [Figure 23C] 1 is a chart showing sensitivity data over an extended sensor session showing characteristic drift along with failure modes. [Figure 24] 4 is a flow chart illustrating another exemplary method in accordance with the present principles. [Diagram 25]4 is a flow chart illustrating another exemplary method in accordance with the present principles. [Figure 26] 4 is a flow chart illustrating another exemplary method in accordance with the present principles. [Figure 27] 4 is a flow chart illustrating another exemplary method in accordance with the present principles. [Figure 28] 1 is a graph showing the distribution of sensitivity. [Figure 29] 1 is a graph showing the distribution of baselines. [Diagram 30] 1 is a graph showing a distribution of glucose values, for example glucose values ​​over time. [Diagram 31] 1 is a graph showing most likely sensitivity (slope) values ​​given example parameters. [Diagram 32] 1 is a graph showing most likely baseline values ​​given example parameters; [Diagram 33] 1 is a graph showing most likely glucose values ​​given example parameters. [Diagram 34] 1 shows an exemplary glucose trace. [Diagram 35] 1 shows an exemplary glucose trace and the effect of a double exponential filter acting on the glucose trace. [Diagram 36] 3 shows estimated drift curves for the sensors used to determine FIGS. 34 and 35. [Figure 37] 4 is a further chart showing drift correction according to the above principles; [Figure 38A] 4 is a further chart showing drift correction according to the above principles; [Figure 38B] 4 is a further chart showing drift correction according to the above principles; [Figure 39] 4 is a further chart showing drift correction according to the above principles; [Diagram 40] 1 shows the measured relationship between the signal coefficient of variation and the standard deviation of glucose concentration values. [Diagram 41]1 shows a measured glucose concentration signal over a period of time. [Diagram 42] 1 shows the linear relationship between the signal coefficient of variation and the glucose concentration value standard deviation, with exemplary values ​​marked. [Diagram 43] The distribution of the difference between the measured and predicted standard deviations is shown. [Diagram 44] 1 shows the relationship between mean glucose and glucose standard deviation. [Diagram 45] The differences between glucose standard deviations are shown as between patient populations: non-diabetic, type I diabetes, and type II diabetes. [Diagram 46] Data points separated by a time lag are shown, and Δ represents the individual percent change between two adjacent points. [Figure 47] 1 is a flow chart illustrating another implementation of a method according to the present principles. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0040] The following description and examples illustrate in detail some exemplary embodiments of the disclosed invention. Those skilled in the art will recognize that there are many variations and modifications of the invention that are encompassed by the scope of the invention. Therefore, the description of a specific exemplary embodiment should not be considered as limiting the scope of the invention.

[0041] definition To facilitate understanding of the preferred embodiments, a number of terms are defined below.

[0042] As used herein, the term "analyte" is a broad term given its ordinary and customary meaning to those of skill in the art (and is not limited to any special or customized meaning), and further refers to, but is not limited to, a substance or chemical constituent in a bodily fluid (e.g., blood, interstitial fluid, cerebrospinal fluid, lymphatic fluid, or urine) that may be analyzed. Analytes may include naturally occurring substances, man-made substances, metabolic products, and / or reaction products. In some embodiments, the analyte for measurement by the sensor head, devices, and methods is glucose. However, other analytes are contemplated as well, including acarboxyprothrombin, acylcarnitines, adenine phosphoribosyltransferase, adenosine deaminase, albumin, α-fetoprotein, amino acid profile (arginine (Krebs cycle), histidine / urocanic acid, homocysteine, phenylalanine / tyrosine, tryptophan), andrenostenedione, antipyrine, arabinitol enantiomers, arginase, benzoylecgonine (cocaine), biotinidase, biopterin, c-reactive protein, carnitine, carnosinase, CD4, ceruloplasmin, chenodeoxycholic acid, chloroquine, cholesterol, cholinesterase, conjugated 1-β hydroxycholic acid, cortisol, creatine kinase, creatine kinase MM isoenzyme, cyclosporin A, d-penicillin, creatine kinase MM isoenzyme, d-penicillin ... Lamin, de-ethylchloroquine, dehydroepiandrosterone sulfate, DNA (acetylation polymorphisms), alcohol dehydrogenase, alpha 1-antitrypsin, cystic fibrosis, Duchenne / Becker muscular dystrophy, analyte-6-phosphate dehydrogenase, hemoglobin A, hemoglobin S, hemoglobin C, hemoglobin D, hemoglobin E, hemoglobin F, D Punjab, beta-thalassemia, hepatitis B virus , HCMV, HIV-1, HTLV-1, Leber's hereditary optic neuropathy, MCAD, RNA, PKU, Plasmodium vivax, sex differentiation, 21-deoxycortisol), desbutylhalofantrine, dihydropteridine reductase, diphtheria / tetanus antitoxin, erythrocyte arginase, erythrocyte protoporphyrin, esterase D, fatty acids / acylglycines, free β-human chorionic gonadotropin, free erythrocyte porphyrin,Free thyroxine (FT4), free tri-iodothyronine (FT3), fumarylacetoacetase, galactose / gal-1-phosphate, galactose-1-phosphate uridyltransferase, gentamicin, analyte-6-phosphate dehydrogenase, glutathione, glutathione peroxidase, glycocholate, glycosylated hemoglobin, halofantrine, hemoglobin variants, hexosaminidase A, human erythrocyte carbonic anhydrase I, 17-α-hydroxyprogesterone, hypoxanthine phosphoribosyltransferase, immunoreactive trypsin, lactate, Lead, lipoproteins ((a), B / A-1, β), lysozyme, mefloquine, netilmicin, phenobarbitone, phenytoin, phytanic acid / pristanic acid, progesterone, prolactin, prolidase, purine nucleoside phosphorylase, kinin, inverted tri-iodothyronine (rT3), selenium, serum pancreatic lipase, sisomicin, somatomedin C, specific antibodies (adenovirus, antinuclear antibody, anti-zeta antibody, arbovirus, pseudorabies virus, dengue virus, dracunculia, tapeworm, cecum, entamoeba histolytica, enterovirus, Giardia lamblia (giardia duodenalisa, Helicobacter pylori, Hepatitis B virus, Herpes virus, HIV-1, IgE (atopic disease), Influenza virus, Leishmania donovani, Leptospira, Measles / Mumps / Rubella, Mycobacterium leprae, Mycoplasma pneumoniae, Myoglobin, Onchocerca volvulus, Parainfluenza virus, Plasmodium falciparum, Poliovirus, Pseudomonas aeruginosa, Respiratory syncytial virus, Rickettsia (scrub typhus), Schistosoma mansoni, Toxoplasma gondii, Trepenoma pallidium, Trypanosoma cruzi / Langer, Vesicular stomatitis virus virus), Wuchereria bancrofti, Yellow fever virus), specific antigens (Hepatitis B virus, HIV-1), succinylacetone, sulfadoxine, theophylline, thyrotropin (TSH), thyroxine (T4), thyroxine-binding globulin, trace elements, transferrin, UDP-galactose-4-epimerase, urea, uroporphyrinogen I synthase, vitamin A, leukocytes, and zinc protoporphyrin.The analytes may be introduced into the body as, for example, contrast agents for diagnostic imaging, radioisotopes, chemical agents, fluorocarbon-based artificial blood, or drugs or pharmaceutical compositions, including, but not limited to, insulin, ethanol, cannabis (marijuana, tetrahydrocannabinol, hashish), inhalants (nitrous oxide, amyl nitrite, butyl nitrite, chlorohydrocarbons, hydrocarbons), cocaine (crack cocaine), stimulants (amphetamines, methamphetamines, Ritalin, Cylert, Preludin, Didrex, PreState, Voranil, Sandrex, Plegine), depressants (barbiturates, methaqualone, tranquilizers, e.g., , Valium, Librium, Miltown, Serax, Equanil, Tranxene), hallucinogens (phencyclidine, lysergic acid, mescaline, peyote, psilocybin), narcotics (heroin, codeine, morphine, opium, meperidine, Percocet, Percodan, Tussionex, Fentanyl, Darvon, Talwin, Lomotil), designer drugs (fentanyl, meperidine, amphetamine, methamphetamine, and analogs of phencyclidine, e.g., Ecstasy), anabolic steroids, and nicotine. Metabolites of drugs and pharmaceutical compositions may also be contemplated as analytes. For example, analytes such as neurochemicals and other chemicals produced in the body, such as ascorbic acid, uric acid, dopamine, noradrenaline, 3-methoxytyramine (3MT), 3,4-dihydroxyphenylacetic acid (DOPAC), homovanillic acid (HVA), 5-hydroxytryptamine (5HT), and 5-hydroxyindoleacetic acid (FHIAA), may be analyzed.

[0043] As used herein, the terms "microprocessor" and "processor" are used broadly and are given their ordinary and customary meanings to those skilled in the art (and are not limited to any special or customized meanings), and further refer to, but are not limited to, computer systems that perform arithmetic and logical operations using logic circuitry that responds to and processes the basic instructions that run a computer, state machines, and the like.

[0044] As used herein, the terms "raw data stream" and "data stream" are broad terms and are intended to have their ordinary and customary meanings as indicated to those skilled in the art (and are not limited to any special or customized meanings), and further refer to, but are not limited to, an analog or digital signal directly related to glucose measured from a glucose sensor. In one example, the raw data stream is digital data in "counts" (e.g., voltage or amperage) converted by an A / D converter from an analog signal, and includes one or more data points representing glucose concentration. These terms broadly encompass multiple timed data points from a substantially continuous glucose sensor, including individual measurements taken at time intervals ranging from less than one second up to, for example, one, two, or five minutes or more. In another example, the raw data stream includes integrated digital values, and the data includes one or more data points representing a glucose sensor signal averaged over a period of time.

[0045] As used herein, the term "calibration" is a broad term and is intended to have its ordinary and customary meaning as indicated to those skilled in the art (and is not limited to any special or customized meaning), and further refers to, but is not limited to, a process of determining a relationship between sensor data and corresponding reference data, with or without the use of reference data in real time, which can be used to convert the sensor data into a meaningful value substantially equivalent to the reference data. In some embodiments, i.e., in continuous analyte sensors, calibration can be updated or recalibrated over time (at the factory, in real time, and / or retroactively) as changes in the relationship between the sensor data and the reference data occur, for example, due to changes in sensitivity, baseline, transport, metabolism, etc. Calibration can also be achieved by automatically predicting sensor signal parameters through analysis of one or more signal characteristics or features (autocalibration).

[0046] As used herein, the terms "calibrated data" and "calibrated data stream" are broad terms and are intended to have their ordinary and customary meanings as given to those of skill in the art (and are not limited to any special or customized meanings), and further refer to data that has been transformed from its raw state to another state using a function, including through the use of sensitivities, e.g., a transformation function, in order to provide a meaningful value to a user.

[0047] As used herein, the terms "smoothed data" and "filtered data" are broad terms given their ordinary and customary meanings to those skilled in the art (and are not limited to any special or customized meanings), and further refer to, but are not limited to, data that has been modified to make the data smoother and more continuous and / or to remove or reduce outliers, for example, by performing a moving average, including a slow moving average, of the raw data stream. Examples of data filters include, but are not limited to, FIR (finite impulse response), IIR (infinite impulse response), moving average filters, and the like.

[0048] As used herein, the terms "smoothing" and "filtering" are broad terms and are given their ordinary and customary meanings by those of skill in the art (and are not limited to any special or customized meanings), and further refer to, but are not limited to, the modification of a set of data to make it smoother and more continuous, or to remove or reduce outliers, for example, by performing a moving average of a raw data stream.

[0049] As used herein, the term "algorithm" is a broad term having its ordinary and customary meaning given to those of skill in the art (and is not limited to any special or customized meaning), and further refers to, but is not limited to, a computational process (e.g., a program) involved in transforming information from one state to another, for example, by using computational processing.

[0050] As used herein, the term "counts" is a broad term given its ordinary and customary meaning to those skilled in the art (and is not limited to any special or customized meaning), and further refers to, but is not limited to, a unit of measurement for a digital signal. In one example, the raw data stream measured in counts is directly related to a voltage (e.g., converted by an A / D converter), which is directly related to a current from the working electrode.

[0051] As used herein, the term "sensor" is a broad term given its ordinary and customary meaning to those of skill in the art (and is not intended to be limited to any special or customized meaning), and further refers to, but is not limited to, a component or region of a device where an analyte may be quantified.

[0052] As used herein, the terms "glucose sensor" and "component for determining the amount of glucose in a biological sample" are broad terms and are given their ordinary and customary meanings to those of skill in the art (and are not limited to any special or customized meanings), and further refer to, but are not limited to, any mechanism by which glucose may be quantified (e.g., enzymatic or non-enzymatic). For example, some embodiments utilize a membrane containing glucose oxidase, which catalyzes the conversion of oxygen and glucose to hydrogen peroxide and gluconate as exemplified by the following chemical reaction:

[0053] Glucose + O2 → Gluconate + H2O2

[0054] Because there is a proportional change in the co-reactant O2 and product HO2, electrodes can be used to monitor the current change of either the co-reactant or product to determine the glucose concentration.

[0055] As used herein, the terms "operably connected" and "operably linked" are broad terms and are intended to have their ordinary and customary meanings as indicated to those skilled in the art (and are not limited to special or customized meanings), and further refer to, but are not limited to, one or more components that are linked to another component(s) in a manner that allows for the transmission of a signal between the components. For example, one or more electrodes may be used to detect the amount of glucose in a sample and convert that information into a signal, e.g., an electrical or electromagnetic signal, which may then be transmitted to an electronic circuit. In this case, the electrodes are "operably linked" to the electronic circuit. These terms are broad enough to include wireless connections.

[0056] The term "determining" encompasses a wide range of actions. For example, "determining" may include calculating, computing, processing, deriving, investigating, searching (e.g., searching in a table, database, or another data structure), determining, etc. "Determining" may also include receiving (e.g., receiving information), accessing (e.g., accessing data in a memory), etc. "Determining" may also include resolving, selecting, choosing, calculating, deriving, establishing, and / or the like. Determining may also include determining that a parameter is consistent with a predetermined criterion, including meeting, passing, exceeding a threshold, etc.

[0057] As used herein, the term "substantially" is intended to be broad and to have its ordinary and customary meaning given to those of ordinary skill in the art (and is not intended to be limited to a special or customized meaning), and further to refer, without limitation, to what is primarily, but not necessarily entirely, specified.

[0058] As used herein, the term "host" is a broad term having its ordinary and customary meaning given to those of skill in the art (and is not intended to be limited to any special or customized meaning), and further refers to mammals, particularly humans, but is not limited thereto.

[0059] As used herein, the term "continuous analyte (or glucose) sensor" is a broad term given its ordinary and customary meaning to those of skill in the art (and is not limited to any special or customized meaning), and further refers to, but is not limited to, a device that measures the concentration of an analyte continuously or continuously, for example, at time intervals ranging from less than one second up to, for example, one, two, or five minutes or more. In one exemplary embodiment, the continuous analyte sensor is a glucose sensor as described in U.S. Patent No. 6,001,067, which is incorporated herein by reference in its entirety.

[0060] As used herein, the term "continuous analyte (or glucose) sensing" is a broad term that is given its ordinary and customary meaning to those of skill in the art (and is not limited to any special or customized meaning), and further refers to, but is not limited to, periods during which analyte monitoring is performed continuously or continuously, for example, at time intervals ranging from less than one second up to, for example, one, two, or five minutes or more.

[0061] As used herein, the terms "reference analyte monitor," "reference analyte meter," and "reference analyte sensor" are broad terms and are given their ordinary and customary meanings to those of skill in the art (and are not limited to any special or customized meanings), and further refer to a device that measures the concentration of an analyte and may be used as a reference for a continuous analyte sensor, but are not limited thereto, for example, a self-monitoring blood glucose meter (SMBG) may be used as a reference for a continuous glucose sensor for comparison, calibration, etc.

[0062] As used herein, the term "sensing membrane" is a broad term given its ordinary and customary meaning to those skilled in the art (and is not limited to any special or customized meaning), and further refers, but is not limited to, a permeable or semi-permeable membrane that may consist of two or more regions and is typically constructed of a material several microns or more thick that is permeable to oxygen and may or may not be permeable to glucose. In one example, the sensing membrane contains immobilized glucose oxidase enzyme, which allows an electrochemical reaction to occur to measure the concentration of glucose.

[0063] As used herein, the term "physiologically likely" is a broad term, given its ordinary and customary meaning to those skilled in the art (and is not limited to any special or customized meaning), and further refers to, but is not limited to, physiological parameters obtained from continuous studies of glucose data in humans and / or animals. For example, a maximum sustained rate of glucose change of about 4-5 mg / dL / min and a maximum acceleration of the rate of change of about 0.1-0.2 mg / dL / min / min in humans are considered physiologically likely limits. Values ​​outside these limits, for example, are considered non-physiological and likely to be the result of signal errors. As another example, the rate of glucose change is lowest at the maximum and minimum of the daily glucose range, which ranges are areas of greatest risk in patient treatment, and thus the physiologically likely rate of change may be set at the maximum and minimum based on continuous studies of glucose data. As a further example, it has been found that the best solution for the shape of the curve at any point along the glucose signal data stream over a certain period of time (e.g., about 20-30 minutes) is a straight line, which may be used to set the physiological limits.

[0064] As used herein, the term "frequency content" is a broad term having its ordinary and customary meaning given to those skilled in the art (and is not intended to be limited to any special or customized meaning), and further refers to, but is not limited to, the spectral density, including the frequencies contained within a signal and their power.

[0065] As used herein, the term "linear regression" is a broad term given its ordinary and customary meaning to those skilled in the art (and is not limited to any special or customized meaning), and further refers to, but is not limited to, finding a line where a set of data has the least measured deviation or separation from the line. By-products of this algorithm include the slope, y-intercept, and R-squared value, which determine how well the measured data fits the line. In certain cases, robust regression techniques may also be used to handle outliers in the regression.

[0066] As used herein, the term "nonlinear regression" is a broad term given its ordinary and customary meaning to those of skill in the art (and is not intended to be limited to any special or customized meaning), and further refers to, but is not limited to, fitting a set of data to describe the relationship between a response variable and one or more explanatory variables in a nonlinear manner.

[0067] As used herein, the term "average" is intended to be broad and to have its ordinary and accustomed meaning given to those of ordinary skill in the art (and is not intended to be limited to any special or customized meaning), and further includes, but is not limited to, the sum of the observations divided by the number of observations.

[0068] As used herein, the term "non-recursive filter" is a broad term that is intended to have its ordinary and customary meaning given to those skilled in the art (and is not intended to be limited to any special or customized meaning), and further refers to, but is not limited to, an equation that uses moving averages as inputs and outputs.

[0069] As used herein, the terms "recursive filter" and "autoregressive algorithm" are broad terms given their ordinary and customary meanings to those skilled in the art (and are not limited to any special or customized meanings), and further refer to, but are not limited to, equations in which the previous average is part of the next filtered output. More specifically, the generation of a sequence of observations in which the value of each observation depends in part on the value of the observation immediately preceding it. One example is a regression structure in which lagged response values ​​play the role of independent variables.

[0070] As used herein, the term "variation" is a broad term given its ordinary and customary meaning to those of skill in the art (and is not limited to any special or customized meaning), and further refers to, but is not limited to, the amount of difference or change from a point, line, or set of data. In one embodiment, an estimated analyte value can have a variation that includes a range of values ​​other than the estimated analyte value, e.g., that represents a range of possibilities based on known physiological patterns.

[0071] As used herein, the terms "physiological parameters" and "physiological boundaries" are broad terms and are to be construed as having their ordinary and customary meanings given to those of skill in the art (and are not to be limited to any special or customized meanings), and further refer to parameters obtained from serial studies of physiological data in humans and / or animals, including, but not limited to, a maximum sustained rate of glucose change of about 6-8 mg / dL / min and a maximum sustained rate of glucose change of about 0.1-0.2 mg / dL / min in humans. 2The maximum acceleration of the rate of change of is considered the physiologically possible limit, and values ​​outside these limits are considered non-physiological. As another example, the rate of change of glucose is lowest at the maximum and minimum of the daily glucose range, which are areas of greatest risk in patient treatment, and thus the physiologically possible rate of change may be set at the maximum and minimum based on continuous study of glucose data. As a further example, it has been found that the best solution for the shape of the curve at any point along the glucose signal data stream over a certain period of time (e.g., about 20-30 minutes) is a straight line, which may be used to set the physiological limits. These terms are broad enough to include physiological parameters for any analyte.

[0072] As used herein, the term "measured analyte value" is a broad term given its ordinary and customary meaning to those of skill in the art (and is not limited to any special or customized meaning), and further refers to, but is not limited to, an analyte value or set of analyte values ​​for the period during which the analyte data was measured by an analyte sensor. The term is broad enough to include data from an analyte sensor before or after data processing (e.g., data smoothing, calibration, etc.) in the sensor and / or receiver.

[0073] As used herein, the term "estimated analyte value" is a broad term having its ordinary and customary meaning given to those of skill in the art (and is not intended to be limited to any special or customized meaning), and further refers to, but is not limited to, an analyte value or set of analyte values ​​that are algorithmically extrapolated from a measured analyte value.

[0074] As used herein, the term "sensor data" is a broad term and is intended to have its ordinary and customary meaning as indicated to those skilled in the art (and is not limited to any special or customized meaning), and further refers to any data associated with a sensor, such as a continuous analyte sensor, but is not limited thereto. Sensor data includes raw data streams of analog or digital signals (or other signals received from another sensor) directly related to the measured analyte from an analyte sensor, or simply data streams, as well as raw data that has been calibrated and / or filtered. In one example, sensor data includes digital data in "counts" (e.g., voltage or amperage) converted from an analog signal by an A / D converter, and includes one or more data points representing glucose concentration. Thus, the terms "sensor data point" and "data point" generally refer to a digital representation of sensor data at a particular time. These terms broadly encompass multiple timed data points from a sensor, such as from a substantially continuous glucose sensor, including individual measurements taken at time intervals ranging from less than one second up to, for example, one, two, or five minutes or more. In another example, the sensor data includes an integrated digital value representing one or more data points averaged over a period of time. The sensor data may include calibrated data, smoothed data, filtered data, transformed data, and / or any other data associated with the sensor.

[0075] As used herein, the term "matched data pair" or "data pair" is a broad term having its ordinary and customary meaning as indicated to those of skill in the art (and is not limited to any special or customized meaning), and further refers to, but is not limited to, reference data (e.g., one or more reference analyte data points) that are substantially time-matched with sensor data (e.g., one or more sensor data points).

[0076] As used herein, the term "sensor electronics" is a broad term having its ordinary and customary meaning given to those of skill in the art (and is not intended to be limited to any special or customized meaning) and refers to, but is not limited to, the components (e.g., hardware and / or software) of a device configured to process data.

[0077] As used herein, the term "calibration set" is a broad term given its ordinary and customary meaning to those skilled in the art (and is not limited to any special or customized meaning) and refers to a set of data that contains information useful for calibration, including, but not limited to, calibration. In some embodiments, the calibration set is formed from one or more matched data pairs that are used to determine the relationship between the reference data and the sensor data, although other data obtained externally or internally prior to implantation may also be used. As another example, data may also be used from a previous sensor session of the target user.

[0078] As used herein, the terms "sensitivity" or "sensor sensitivity" are broad terms given their ordinary and customary meanings to those of skill in the art (and are not limited to any special or customized meanings) and refer to, but are not limited to, the amount of signal generated by a particular concentration of a measured analyte or a measured species (e.g., HO) associated with the measured analyte (e.g., glucose). For example, in one embodiment, the sensor has a sensitivity of about 1 to about 300 picoamps of current for every mg / dL of glucose analyte.

[0079] As used herein, the terms "sensitivity profile" or "sensitivity curve" are broad terms and are given their ordinary and customary meanings to those of skill in the art (and are not intended to be limited to any special or customized meanings), and refer to, but are not limited to, a display of change in sensitivity over time.

[0080] Other definitions are provided within the description which follows, and possibly from the context of the use of the terms.

[0081] As used herein, the following abbreviations apply: Eq and Eqs (equivalents), mEq (milliequivalents), M (mole), mM (millimolar), μM (micromolar), N (normal), mol (mole), mmol (millimolar), μmol (micromolar), nmol (nanomole), g (gram), mg (milligram), μg (microgram), Kg (kilogram), L (liter), mL (milliliter), dL (deciliter), μL (microliter), cm (centimeter), mm (millimeter), μm (micrometer), nm (nanometer), h and hr (hours), min (minutes), and sec. (seconds), °C (degrees Celsius).

[0082] System Overview / Overview Conventional in vivo continuous analyte sensing techniques have typically relied on a reference measurement performed during a sensor session for calibration of the continuous analyte sensor. The reference measurement is matched with substantially time-corresponding sensor data to produce a matched data pair. Regression is then performed on the matched data pair (e.g., by using least squares regression) to generate a transfer function that defines the relationship between the sensor signal and the estimated glucose concentration.

[0083] In critical care settings, calibration of continuous analyte sensors is often performed by using a calibration solution with a known concentration of analyte as a reference. This calibration procedure can be cumbersome because a calibration bag separate from (and in addition to) the IV (intravenous) bag is typically used. In ambulatory settings, calibration of continuous analyte sensors is traditionally performed by peripheral blood glucose measurements (e.g., fingerstick glucose tests) through which reference data is obtained and entered into the continuous analyte sensor system. This calibration procedure typically involves frequent fingerstick measurements, which can be inconvenient and painful.

[0084] To date, systems and methods for in vitro calibration (e.g., factory calibration) of continuous analyte sensors by manufacturers that do not rely on periodic recalibration have, in large part, been inadequate with respect to the high levels of sensor accuracy required for glycemic management. Part of this can be attributed to changes in sensor characteristics (e.g., sensor sensitivity) that can occur during sensor use. Thus, calibration of continuous analyte sensors has typically involved periodic input of reference data, whether they are associated with calibration fluids or fingerstick measurements. This can be very cumbersome for users in everyday life, and for patients in outpatient settings or hospital staff in critical care settings.

[0085] The following description and examples illustrate the present embodiments with reference to the drawings, in which reference numerals indicate elements of the present embodiments, and which are reproduced below with reference to a discussion of the corresponding drawing features.

[0086] Described herein are systems and methods for calibrating continuous analyte sensors that are capable of achieving high levels of accuracy without (or with reduced reliance on) reference data from a reference analyte monitor (e.g., from a blood glucose meter).

[0087] Sensor Systems 1 illustrates an example system 100 according to some example implementations. The system 100 includes a continuous analyte sensor system 8, which includes sensor electronics 12 and a continuous analyte sensor 10. The system 100 may include other devices and / or sensors, such as a drug delivery pump 2 and a glucose meter 4. The continuous analyte sensor 10 may be physically connected to the sensor electronics 12, may be integral to the continuous analyte sensor 10 (e.g., non-releasably attached) or may be releasably attachable. The sensor electronics 12, drug delivery pump 2, and / or glucose meter 4 may be coupled to one or more devices, such as display devices 14, 16, 18, and / or 20.

[0088] In some example implementations, system 100 may include a cloud-based analyte processor 490 configured to analyze analyte data (and / or other patient-related data) provided over network 406 (e.g., via wired, wireless, or combinations thereof) from sensor system 8 and other devices associated with a host (also referred to as a patient), such as display devices 14-20, and generate reports that provide high level information, such as statistical data about analytes measured over a particular time frame. A thorough discussion of using cloud-based analyte processing systems can be found in U.S. Patent No. 13 / 788,375, filed March 7, 2013, entitled "Cloud-Based Processing of Analyte Data," which is incorporated herein by reference in its entirety.

[0089] In some exemplary implementations, the sensor electronics 12 may include electronic circuitry associated with measuring and processing data generated by the continuous analyte sensor 10. This generated continuous analyte sensor data may also include algorithms that may be used to process and calibrate the continuous analyte sensor data, although these algorithms may be provided in other manners as well. The sensor electronics 12 may include hardware, firmware, software, or a combination thereof to provide a measurement of an analyte level via a continuous analyte sensor, such as a continuous glucose sensor. An exemplary implementation of the sensor electronics 12 is further described below with respect to FIG. 2.

[0090] Sensor electronics 12, as described, may be coupled (e.g., wirelessly, etc.) with one or more devices, such as display devices 14, 16, 18, and / or 20. Display devices 14, 16, 18, and / or 20 may be configured to present (and / or alert) information, such as sensor information, transmitted by sensor electronics 12 for display on display devices 14, 16, 18, and / or 20.

[0091] The display devices may include a relatively small key fob-type display device 14, a relatively large handheld display device 16, a mobile phone 18 (e.g., a smartphone, tablet, etc.), a computer 20, and / or any other user device configured to present at least information (e.g., drug delivery information, separate self-monitoring glucose readings, a heart rate monitor, a food intake monitor, etc.).

[0092] In some example implementations, the relatively small key fob display device 14 may include a wristwatch, a belt, a necklace, a pendant, jewelry, an adhesive patch, a pager, a key fob, a plastic card (e.g., a credit card), an identification (ID) card, and / or the like. This small display device 14 may include a relatively small display (e.g., smaller than the large display device 16) and may be configured to display certain types of displayable sensor information, such as numbers, arrows, or color codes.

[0093] In some example implementations, the relatively large handheld display device 16 may include a handheld receiver device, a palmtop computer, and / or the like. This large display device may include a relatively large display (e.g., larger than the small display device 14) and may be configured to display information such as a graphical representation of continuous sensor data, including current and historical sensor data, output by the sensor system 8.

[0094] In some exemplary implementations, the continuous analyte sensor 10 includes a sensor for detecting and / or measuring an analyte, and the continuous analyte sensor 10 may be configured to continuously detect and / or measure an analyte as a non-invasive, subcutaneous, transcutaneous, and / or intravascular device. In some exemplary implementations, the continuous analyte sensor 10 may analyze multiple intermittent blood samples, although other analytes may be used as well.

[0095] In some exemplary implementations, the continuous analyte sensor 10 may include a glucose sensor configured to measure glucose in blood or interstitial fluid using one or more measurement techniques, such as enzymatic, chemical, physical, electrochemical, spectrophotometric, polarimetric, calorimetric, iontophoretic, radiometric, immunochemical, etc. In implementations in which the continuous analyte sensor 10 includes a glucose sensor, the glucose sensor may include any device capable of measuring the concentration of glucose and may provide data, such as a data stream indicative of the concentration of glucose in a host, using a variety of techniques to measure glucose, including invasive, minimally invasive, and non-invasive sensing techniques (e.g., fluorescence monitoring). The data stream may be sensor data (raw and / or filtered), which may be converted to a calibrated data stream that is used to provide glucose values ​​to a host, such as a user, patient, or caregiver (e.g., a parent, relative, guardian, teacher, doctor, nurse, or any other individual interested in the health of the host). Additionally, the continuous analyte sensor 10 may be embedded as at least one of the following types of sensors: Implantable glucose sensors, transcutaneous glucose sensors implanted in the host vasculature or extracellular, subcutaneous sensors, replaceable subcutaneous sensors, intravascular sensors.

[0096] The disclosure herein refers to some implementations including a continuous analyte sensor 10 including a glucose sensor, however, the continuous analyte sensor 10 may include other types of analyte sensors as well. Additionally, while some implementations refer to the glucose sensor as an implantable glucose sensor, other types of devices capable of detecting the concentration of glucose and providing an output signal representative of the glucose concentration may be used as well. Additionally, while the disclosure herein refers to glucose as the analyte being measured, processed, etc., other analytes may be used as well, including, for example, ketone bodies (e.g., acetone, acetoacetate and beta-hydroxybutyrate, lactate, etc.), glucose, acetyl-CoA, triglycerides, fatty acids, intermediates in the citric acid cycle, choline, insulin, cortisol, testosterone, etc.

[0097] 2 illustrates one example of sensor electronics 12 according to some example implementations. Sensor electronics 12 may include sensor electronics configured to process sensor information, such as sensor data, and generate transformed sensor data and displayable sensor information, for example, via a processor module. For example, the processor module may transform the sensor data into one or more of the following: filtered sensor data (e.g., one or more filtered analyte concentration values), raw sensor data, calibrated sensor data (e.g., one or more calibrated analyte concentration values), rate of change information, trend information, acceleration / deceleration information, sensor diagnostic information, location information, alarm / alert information, calibration information, sensor data smoothing and / or filtering algorithms, and / or the like.

[0098] In some embodiments, the processor module 214 is configured to accomplish a significant portion, if not all, of the data processing. The processor module 214 may be integral to the sensor electronics 12 and / or may be located remotely, such as in one or more of the devices 14, 16, 18, and / or 20, and / or the cloud 490. In some embodiments, the processor module 214 may include multiple smaller subcomponents or submodules. For example, the processor module 214 may include an alert module (not shown), or a prediction module (not shown), or any other suitable module that may be utilized to efficiently process data. When the processor module 214 is comprised of multiple submodules, the submodules may be located within the processor module 214, including within the sensor electronics 12 or other associated devices (e.g., 14, 16, 18, 20, and / or 490). For example, in some embodiments, the processor module 214 may be located at least partially within the cloud-based analyte processor 490 or elsewhere within the network 406.

[0099] In some example implementations, the processor module 214 may be configured to calibrate the sensor data, and the data storage 220 may store the calibrated sensor data points as transformed sensor data. Additionally, the processor module 214 may be configured in some example implementations to wirelessly receive calibration information from a display device, such as devices 14, 16, 18, and / or 20, to enable calibration of the sensor data from the sensor 12. Additionally, the processor module 214 may be configured to perform additional algorithmic processing on the sensor data (e.g., the calibrated and / or filtered data and / or other sensor information), and the data storage 220 may be configured to store the transformed sensor data and / or sensor diagnostic information associated with the algorithm.

[0100] In some example implementations, sensor electronics 12 may include an application specific integrated circuit (ASIC) 205 coupled to a user interface 222. ASIC 205 may further include a potentiostat 210, a telemetry module 232 for transmitting data from sensor electronics 12 to one or more devices, such as devices 14, 16, 18, and / or 20, and / or other components for signal processing and data storage (e.g., processor module 214 and data storage 220). Although FIG. 2 illustrates ASIC 205, other types of circuits may be used as well, including a field programmable gate array (FPGA), one or more microprocessors configured to provide some (if not all) of the processing performed by sensor electronics 12, analog circuits, digital circuits, or combinations thereof.

[0101] 2, a potentiostat 210 is coupled to the continuous analyte sensor 10, such as a glucose sensor, to generate sensor data from the analyte. The potentiostat 210 may also provide a voltage to the continuous analyte sensor 10 via data line 212 to bias the sensor for measurement of a value (e.g., a current value) indicative of the analyte concentration in the host (also referred to as the analog portion of the sensor). The potentiostat 210 may have one or more channels depending on the number of working electrodes in the continuous analyte sensor 10.

[0102] While in some exemplary implementations the potentiostat 210 may include a resistor that converts the current value from the sensor 10 to a voltage value, in some exemplary implementations a current-to-frequency converter (not shown) may also be configured to continuously incorporate the measured current value from the sensor 10, for example, using a charge counting device. In some exemplary implementations an analog-to-digital converter (not shown) may digitize the analog signal from the sensor 10 into so-called "counts" to enable processing by the processor module 214. The resulting counts may be directly related to the current measured by the potentiostat 210, which may in turn be directly related to an analyte level, such as a glucose level, within the host.

[0103] The telemetry module 232 may be operatively connected to the processor module 214 and may provide hardware, firmware, and / or software to enable wireless communication between the sensor electronics 12 and one or more other devices, such as a display device, a processor, a network access device, etc. Various wireless technologies that may be implemented in the telemetry module 232 include Bluetooth, Bluetooth Low-Energy, ANT, ANT+, ZigBee, IEEE 802.11, IEEE 802.16, cellular radio access technology, radio frequency (RF), infrared (IR), paging network communication, magnetic induction, satellite data communication, spread spectrum communication, frequency hopping communication, short range wireless communication, and / or the like. In some exemplary implementations, the telemetry module 232 includes a Bluetooth chip, although Bluetooth technology may also be implemented in a combination of the telemetry module 232 and the processor module 214.

[0104] The processor module 214 may control the processing performed by the sensor electronics 12. For example, the processor module 214 may be configured to process data (e.g., counts) from the sensor, filter the data, calibrate the data, perform fail-safe checks, and / or the like.

[0105] In some exemplary implementations, the processor module 214 may include a digital filter, such as an infinite impulse response (IIR) or finite impulse response (FIR) filter. The digital filter may smooth the raw data stream received from the sensor 10. Generally, the digital filter is programmed to filter data sampled at predetermined time intervals (also referred to as a sampling rate). In some exemplary implementations, such as when the potentiostat 210 is configured to measure an analyte (e.g., glucose and / or the like) at discrete time intervals, these time intervals determine the sampling rate of the digital filter. In some exemplary implementations, the potentiostat 210 may be configured to measure an analyte continuously, for example, using a current-to-frequency converter. In these current-to-frequency converter implementations, the processor module 214 may be programmed to request digital values ​​from an integrator of the current-to-frequency converter at predetermined time intervals (collection time). These digital values ​​obtained by the processor module 214 from the integrator may be averaged over the collection time by the continuity of the current measurements. Thus, the collection time may be determined by the sampling rate of the digital filter. Other uses of FIR filters are described in more detail below.

[0106] The processor module 214 may further include a data generator (not shown) configured to generate data packages for transmission to devices such as the display devices 14, 16, 18, and / or 20. Additionally, the processor module 214 may generate data packets for transmission to these external sources via the telemetry module 232. In some example implementations, the data packages may be customizable for each display device as described, and / or may include any available data, such as timestamps, displayable sensor information, converted sensor data, identifier codes for the sensors and / or sensor electronics 12, raw data, filtered data, calibrated data, rate of change information, trend information, error detection or correction, and / or the like.

[0107] The processor module 214 may also include program memory 216 and other memory 218. The processor module 214 may be coupled to a communications interface, such as a communications port 238, and a power source, such as a battery 234. In addition, the battery 234 may be further coupled to a charger and / or regulator 236 to provide power to the sensor electronics 12 and / or charge the battery 234.

[0108] The program memory 216 may be implemented as a pseudo-static memory for storing data such as identifiers (e.g., sensor identifiers (IDs)) for coupled sensors 10, and for storing code (also referred to as program code) for calibrating the ASIC 205 to perform one or more of the operations / functions described herein. For example, the program code may configure the processor module 214 to process and filter data streams or counts, perform calibration methods described below, perform fail-safe checks, etc.

[0109] The memory 218 may also be used to store information. For example, the processor module 214 including the memory 218 may be used as a cache memory for the system, with temporary storage provided for recent sensor data received from the sensors. In some exemplary implementations, the memory may include storage components such as read-only memory (ROM), random access memory (RAM), dynamic RAM, static RAM, non-static RAM, easily erasable programmable read-only memory (EEPROM), rewriteable ROM, flash memory, and the like.

[0110] The data store 220 may be coupled to the processor module 214 and may be configured to store various sensor information. In some example implementations, the data store 220 stores one or more days of continuous analyte sensor data. For example, the data store may store 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 20, and / or 30 days (or more) of continuous analyte sensor data received from the sensor 10. The stored sensor information may include one or more of the following: timestamp, raw sensor data (one or more raw analyte concentration values), calibrated data, filtered data, converted sensor data, and / or any other displayable sensor information, calibration information (e.g., reference BG values ​​and / or previous calibration information), sensor diagnostic information, etc.

[0111] The user interface 222 may include various interfaces such as one or more buttons 224, a liquid crystal display (LCD) 226, a vibrator 228, an audio transducer (e.g., speaker) 230, a backlight (not shown), and / or the like. The components that comprise the user interface 222 may provide controls for interacting with a user (e.g., host). The one or more buttons 224 may enable, for example, toggles, menu selections, option selections, state selections, yes / no responses to on-screen questions, a "turn off" function (e.g., for alarms), an "acknowledged" function (e.g., for alarms), a reset, and / or the like. The LCD 226 may provide, for example, visual data output to the user. The audio transducer 230 (e.g., speaker) may provide audible signals in response to the triggering of certain alerts, such as current and / or predicted hyperglycemic and hypoglycemic conditions. In some example implementations, the audible signals may be differentiated by tone, volume, duty cycle, pattern, duration, and / or the like. In some example implementations, the audible signal may be configured to be muted (e.g., acknowledged or turned off) by pressing one or more buttons 224 on the sensor electronics 12 and / or by signaling the sensor electronics 12 using a button or selection on a display device (e.g., a key fob, cell phone, and / or the like).

[0112] Although audio and vibration alerts are described with respect to Figure 2, other alert mechanisms may be used as well. For example, in some exemplary implementations, tactile alerts are provided, including a poking mechanism configured to "poke" or make physical contact with the patient in response to one or more alarm conditions.

[0113] Another battery 234 may be operatively connected to the processor module 214 (and possibly other components of the sensor electronics 12) and provide the necessary power for the sensor electronics 12. In some exemplary implementations, the battery is a lithium manganese dioxide battery, but any appropriately sized and powered battery may be used (e.g., AAA, nickel cadmium, zinc carbon, alkaline, lithium, nickel metal hydride, lithium ion, zinc air, zinc mercury oxide, silver zinc, or sealed). In some exemplary implementations, the battery is rechargeable. In some exemplary implementations, multiple batteries may be used to power the system. In still other implementations, the receiver may be powered transcutaneously, for example, via inductive coupling.

[0114] The battery charger and / or balancer 236 may be configured to accept energy from an internal and / or external charger. In some exemplary implementations, the battery conditioner (or balancer) 236 regulates the recharging process by bleeding off excess charge current to allow all cells or batteries in the sensor electronics 12 to be fully charged without overcharging other cells or batteries. In some exemplary implementations, the battery(ies) 234 are configured to be charged via an inductive and / or wireless charging pad, although any other charging and / or power mechanism may be used as well.

[0115] One or more communication ports 238, also referred to as external connector(s), may be provided to enable communication with other devices, for example, a PC communication (com) port may be provided to enable communication with systems separate from or integral to the sensor electronics 12. For example, the communication port may include a serial (e.g., Universal Serial Bus or "USB") communication port to enable communication with another computer system (e.g., a PC, a personal digital assistant or "PDA", a server, etc.). In some example implementations, the sensor electronics 12 can transmit historical data to a PC or other computing device (e.g., an analyte processor as disclosed herein) for retrospective analysis by the patient and / or physician. As another example of data transmission, factory information may also be transmitted to the algorithm from the sensor or from a cloud data source.

[0116] In some continuous analyte sensor systems, the on-skin portion of the sensor electronics may be simplified to minimize the complexity and / or size of the on-skin electronics, for example, providing only raw, calibrated, and / or filtered data to a display device configured to perform calibration and other algorithms required to display the sensor data. However, the sensor electronics 12 (e.g., via the processor module 214) may also be implemented to perform predictive algorithms used to generate the converted sensor data and / or displayable sensor information, including, for example, algorithms to evaluate clinical acceptability of reference and / or sensor data, evaluate calibration data for best calibration based on inclusion criteria, evaluate calibration quality, compare predicted analyte values ​​to time-corresponding estimated analyte values, analyze variability in estimated analyte values, evaluate stability of the sensor and / or sensor data, detect signal artifacts (noise), replace signal artifacts, determine rate of change and / or trend of sensor data, perform dynamic and intelligent analyte value estimation, perform diagnostics on the sensor and / or sensor data, set modes of operation, evaluate data with respect to the above, and / or the like.

[0117] Although separate data storage and program memory are shown in Figure 2, a variety of configurations may be used as well. For example, one or more memories may be used to provide storage space to support data processing and storage requirements in sensor electronics 12.

[0118] proofreading While some continuous glucose sensors rely on (and estimate the accuracy of) BG values ​​and / or factory derived information for calibration, the disclosed embodiments utilize real-time information (e.g., in some implementations including only the sensor data itself) to determine and calibrate aspects based thereon.

[0119] In some cases, calibration of the analyte sensor may use a priori calibration distribution information. For example, in some embodiments, the a priori calibration distribution information or code may be received as information from a previous calibration and / or sensor session (e.g., built into the same sensor system), stored in memory, coded on a barcode on packaging at the factory (e.g., as part of a factory setup), transmitted from a cloud or network of remote servers, coded by a caregiver or user, received from another sensor system or electronic device based on results from laboratory testing, and / or the like.

[0120] As used herein, a priori information includes information obtained prior to a particular calibration, such as from a previous calibration of a particular sensor session (e.g., feedback from previous calibration(s)), information obtained prior to sensor insertion (e.g., factory information from in vitro testing or data obtained from already implanted analyte concentration sensors, e.g., sensors from the same manufacturing lot of sensors and / or sensors from one or more different lots, etc.), previous in vivo testing of a similar sensor on the same host, and / or previous in vivo testing of a similar sensor on a different host. Calibration information includes information useful for calibrating a continuous glucose sensor, such as sensitivity (m), change in sensitivity (Δdm / dt) (which may also be referred to as drift in sensitivity), rate of change in sensitivity (ddm / ddt), baseline / intercept (b), change in baseline (Δdb / dt), rate of change in baseline (ddb / ddt), baseline and / or sensitivity profile (i.e., change over time) associated with the sensor, linearity, response time, relationships between characteristics of the sensor (e.g., the relationship between sensitivity and baseline), or a particular stimulus signal output (e.g., impedance of the sensor), as described in U.S. Patent Publication No. 2012 / 0265035-A1, which is incorporated by reference in its entirety. The calibration information may include, but is not limited to, a relationship between an output indicative of impedance, capacitance, or other electrical or chemical property (e.g., determined from previous in vivo and / or ex vivo studies) and sensor sensitivity or temperature (e.g., determined from previous in vivo and / or ex vivo studies), sensor data obtained from an already implanted analyte concentration sensor, a calibration code(s) associated with a sensor that has been calibrated, a patient-specific relationship between the sensor and sensitivity, baseline, drift, impedance, impedance / temperature relationship (e.g., determined from previous studies of the patient or other patients with characteristics in common with the patient), sensor implantation site (abdomen, arm, etc.), and / or a specific relationship (different sites may have different vascular densities). Distribution information includes a range, distribution function, distribution parameters (e.g., mean, standard deviation, skewness, etc.), a general function, a statistical distribution, a profile, or the like that represents multiple possible values ​​for the calibration information.In summary, a priori calibration distribution information includes range(s) or distribution(s) of values ​​(e.g., describing their associated probabilities, probability density functions, likelihoods, or frequencies of occurrence) provided prior to a particular calibration process that is useful for calibrating a sensor (e.g., sensor data).

[0121] For example, in some embodiments, the a priori calibration distribution information includes a probability distribution for sensitivity (m) or sensitivity-related information, and baseline (b) or baseline-related information, e.g., based on sensor type. As noted above, prior distributions of sensitivity and / or baseline may be obtained from a factory (e.g., from in vitro or in vivo testing of representative sensors) or may be obtained from prior calibration.

[0122] As mentioned above, an analyte sensor generally includes electrodes for monitoring changes in current in either a coreactant or product to determine an analyte concentration, e.g., glucose concentration. In one example, the sensor data includes digital data in "counts" (e.g., voltage or amperage) converted by an A / D converter from an analog signal. Calibration is the process of determining the relationship between the measured sensor signal in counts and the analyte concentration in clinical units. For example, calibration allows a given sensor measurement value in counts to be related to a measured analyte concentration value, e.g., in milligrams per deciliter. With reference to graph 10 in FIG. 3, this relationship is generally a linear relationship of the form y=mx+b, where "y" is the sensor signal in counts (y-axis 12), "x" is the clinical value of the analyte concentration (axis 14), and "m" is the sensor sensitivity, which has units of [counts / (mg / dL)]. A line 16 is shown, the slope of which is referred to as the sensor sensitivity. "b" (see line segment 15) is the baseline sensor signal, which may be taken into account or, for advanced sensors, may generally be reduced to zero or near zero. In either case, in many cases the baseline may be estimated to be small in a predictable manner or may be able to be compensated for. In some implementations, a constant background signal is seen, and such a signal is modeled by y=m(x+c), where c is the glucose offset between the sensor site and blood glucose.

[0123] Once line 16 is determined, the system can convert the measured counts (or amperes as described above, eg, picoamps) to a clinical value of analyte concentration.

[0124] However, the values ​​of m and b vary between sensors and require determination. In addition, the gradient value m is not necessarily constant. For example, and with reference to FIG. 4, the value m may vary over time within a session from an initial sensitivity value m0 to a final sensitivity value m F The rate of change is seen to be greatest during the first few days of use, and this rate of change is mR It is called.

[0125] The slope is a function of in vivo time for many reasons. In particular, with regard to initial changes in calibration, such changes are often due to the sensor membrane "settling" into the in vivo environment and achieving equilibrium with that environment. Sensors are generally calibrated in vitro or on the bench, and although efforts are made to make the in vitro environment as close as possible to the in vivo environment, differences are still evident, and the in vivo environment itself varies between users. In addition, sensors may differ due to differences in sterilization or shelf life / storage conditions. Changes in calibration that occur later in a session are often due to changes in the tissue surrounding the sensor, for example, accumulation of biofilm on the sensor.

[0126] Whatever the cause, certain effects of variability have been measured and determined. For example, the final sensitivity m F It is known that the variability of m is the largest contributor to overall sensor inaccuracy. Similarly, it is known that the variability of initial sensitivity m and physiology are the largest contributors to sensor inaccuracy on the first day.

[0127] Because of the variability discussed above, an initial step in calibration involves determining and using seed values ​​for one or more calibration parameters until additional data is acquired to adjust the seed values ​​to more refined seed values.

[0128] Once calibration is achieved, the sensor and analyte concentration measurement system may be used to accurately determine the clinical value of the analyte concentration in a user. The sensor and system may then provide differentiation of other sensor behaviors, including determination of sensitivity error and drift, as described in more detail below.

[0129] The most common current calibration method is through the use of an external blood glucose meter. Such a calibration method is commonly referred to as "fingerstick calibration" and is a well-known and common part of life for many diabetics. This technique has the advantage of not requiring significant factory information, and furthermore offers a low risk of outliers. The disadvantage is that significant user involvement is required, as well as knowledge of certain other factory calibration information required for proper calibration. Once the meter itself is calibrated, the values ​​from the meter can be used to calibrate indwelling analyte concentration meters, since measurements from such meters are trusted. Although the sensor sensitivity changes over time as seen in FIG. 4, the changing sensitivity is not significant if the user is willing to perform many external calibrations.

[0130] However, users generally do not wish to perform many such calibrations, and in many cases, e.g., for type II diabetics, pre-diabetics, or even non-diabetics, the additional accuracy provided by such calibrations is not strictly required. For example, it may be sufficient for a user to know what ranges analyte concentration values ​​lie within rather than the exact values. In another implementation, the data may be provided with associated confidence intervals to inform the user how much confidence to place in the displayed data.

[0131] Thus, efforts have been made to reduce the number of calibrations. Nevertheless, many current CGM systems still require that blood glucose levels be used for at least the initial calibration, which is often also required at the time of administration. The present systems and methods according to the present principles relate, in part, to methods of reducing or eliminating such required calibrations.

[0132] One simple and convenient way to provide a level of calibration is through the use of calibration information for the associated sensor. Even if the calibration information is approximate, such information may still be sufficient for use with a particular patient population. For example, and referring to flow chart 18 of FIG. 5, if the factory calibration is known for one or more sensors in a manufacturing lot, this information may be provided to other sensors of the manufacturing lot that have not yet been used in a patient (step 20). This step is often referred to as providing a "code" to the transmitter, since the code is often used to identify the manufacturing lot (and therefore the details) of the sensor when the sensors are linked together during insertion into a patient as part of a CGM system. The transmitter may then identify the manufacturing lot from the code and apply the appropriate calibration parameters according to a look-up table or other technique. However, it should be understood that the code may be provided not only to the transmitter, but also to any device where counts can be converted to clinical units, such as a dedicated receiver, an off-the-shelf device that can be used to receive and display analyte concentrations, such as a smartphone, tablet computer, etc. Additionally, the code need not be a code in the typical sense, but simply provide any identifier to any device that needs it for calibration purposes.

[0133] Referring again to Figure 5, methods are described for accomplishing the step of providing factory information about the sensor to a transmitter or other electronic device (step 20). Some of these methods provide an option for providing calibration information without the use of a code (step 22). Other methods include providing the code using a step of user data entry (step 28). In some cases, the code may be entered without user input (step 24). Still other methods of providing the code may be used (step 26), and such additional methods are also described below. Details of these methods will now be described.

[0134] With reference to flow chart 30 of FIG. 6, an embodiment of providing factory calibration information in the absence of a code (step 22) is described. A first method is to use a representative value, e.g., the mean or median of a production lot, or other measurement, e.g., a range (step 34). That is, if the mean value of a production lot is known, or even if the mean value of different production lots manufactured using the same technology is known, it can be presumed that the sensor calibration parameters will be similar and therefore can be used as part of the calibration of a new sensor. Alternatively, predictable relationships can also be used to interpolate the sensor calibration parameters, e.g., to bracket sensor lots.

[0135] As another example, impedance measurements may be used in determining the calibration parameters (step 40).

[0136] Calibration may also be performed using information about the previous calibration (step 38). For example, if a user has just switched sensors that were calibrated and were measuring the user's glucose concentration at, for example, 120 mg / dL, it may be presumed that the proper measurement of the newly inserted sensor should be such that the user's glucose is again 120 mg / dL. In some cases, if a predicted glucose value has been determined, the predicted value may be used for the newly inserted sensor. Even if a short time has elapsed between the last reading of the old sensor and the new sensor reading, recognition of possible physiological glucose changes results in limits on what the new measurement value and therefore the calibration parameters of the new sensor may be.

[0137] In yet another variation, various self-calibration algorithms may be used to self-calibrate the CGM system (step 36). In this sense, the CGM system may be referred to as becoming "self-aware". For example, the CGM system may be seeded with an average glucose value that is known, for example, from a previous session, including using a steady-state value from the previous session or a slow moving average from the previous session, as described in more detail below. The CGM system may also be seeded with an A1C value, if available. Various estimates may also be made as needed. The seeded average value may be represented by a distribution, techniques for which are also described in more detail below.

[0138] 7 describes a system in which a code may be provided from a sensor to a transmitter without a user input step (step 24). Again, it is noted that while language is used herein regarding providing a code to a transmitter, it will be understood that the code may likewise be provided to a variety of devices in signal communication with the transmitter, including dedicated receivers, smartphones, tablet computers, follower devices, or other computing environments.

[0139] In the implementation of diagram 42 of FIG. 7, a code or the like is provided to the transmitter, but without significant user involvement. For example, a degree of coding may be achieved by sending manufacturing lots of sensors to different markets (step 46). In one implementation, sensors with similar codes may be sent to different geographic locations (step 48). For example, sensors sent to a particular geographic area may be from a similar or same manufacturing lot, and known calibration parameters for that lot may be provided to the transmitter when it is inserted and initially interacts with the network, thus providing a degree of immediate calibration based on geographic location. Geolocation information may be used to identify a location, which may then be used to identify or classify the sensor.

[0140] In the same manner, sensors of similar lots may be grouped by product, so that different codes are then associated with different products (step 54). For example, a first product may have a first code associated with it, and all of the sensors for that product may be manufactured in the same or similar manner, resulting in little manufacturing variability between sensors associated with a particular product. In this case, once a product is identified, the associated sensor calibration parameters may be uniquely determined, at least as an average.

[0141] In another implementation, without particular regard to geographic location or product, a particular group of sensors having a particular code may be shipped with the code specifically associated with the corresponding user's transmitter (step 56). In this case, once the calibration parameters are known for one member of the group that has been transmitted, and such parameters may be known well before the group is transmitted, the calibration parameters for the remainder of the group are also known.

[0142] RFID technology may also be used to identify the manufacturing lot of the sensor (step 58). For example, a small RFID chip may be located at the base of the sensor and may be read by the transmitter when the sensor and transmitter are mated (step 60). In another implementation, the RFID may be read by a receiver, or a smartphone or other device (step 62). Alternatively, the RFID device may be located on the applicator, and the transmitter may read the identification information (and therefore the calibration information) again when the applicator is used to attach the sensor to a patient.

[0143] In yet another implementation, near field communication (NFC) may be used on the packaging or any other component of the system to communicate identifying information (step 66).

[0144] Other types of communication schemes may be used to communicate information from the sensor to the transmitter. For example, a mechanical sensor on the transmitter may enable communication of the transmitter's code information, e.g., bumps, vertical pins, mechanical system sensing direction, to the base or other mechanical elements readable by the transmitter (step 72). A magnetic sensor may be used for the same purpose (step 78), and in the same way an optical reader on the transmitter may be used to read, for example, bar codes or QR codes as well as other identifying marks or colors (step 74). Resistive sensors, or other sensors that detect the state of connection, may also be used (step 76). For example, sensors of different codes may have corresponding different lengths. The sensor may have multiple contact pads to which the sensor is aligned. Sensors of different codes may have different resistivities, the measurement of which may determine the code.

[0145] FIG. 8 shows a diagram 80 illustrating a method of code communication with user data input (step 28). Perhaps most simply, a code may be provided to the user at the time of purchase, which is simply manually entered (step 84) into the receiver, smartphone, or other device having a UI that allows for data entry. For example, the user may enter text, numbers, colors, etc. The sensor may also be shipped with a card, e.g., a SIM card, which may be inserted into the receiver (step 90) to allow calibration information to be communicated, but without the need for the user to enter a manual code. The transmitter may be equipped with a switch system (step 88), and the user may adjust the position of the switch on the transmitter according to the instructions on the sensor received. For example, the transmitter switch may be a four-position switch or a DIP switch, and by appropriate adjustment, the user may provide the transmitter with a code associated with the sensor. The receiver or smartphone may also be capable of scanning a label associated with the sensor via an integrated camera or barcode reader, allowing information to be communicated in that manner. The scan may include a bar label, a QR code, etc.

[0146] One variation is described below in Figures 9-11. This implementation improves or enables factory calibration with human data from the field. In more detail, factory calibration parameters (e.g., sensitivity and baseline over time) are often best identified using human data. Bench data correlates with human data, but the correlation is still not perfect and often there is an offset in the correlation. Having access to the human data generated by each lot of sensors produced during manufacturing is generally the best data set for generating factory calibration information. Because factory calibration parameters can vary between lots, characterizing each lot can be advantageous when improvements are made.

[0147] 9-11 show how data collected on a person using a portion of many sensors can be used to generate or adjust factory calibration numbers for the remainder of the lot. There are several permutations of this method.

[0148] In a location arrangement, calibrated sensors are sent to market for use by patients. Once such sensors have been calibrated in the connected system, for example via blood glucose calibration techniques or other calibration techniques, including calibration techniques that use only the CGM signal itself, the calibration information may be sent back to the manufacturer via the cloud or other internet-based network. The information may be used to generate factory calibration settings for the remainder of the manufacturing lot of sensors that were not sent to market, which may then be shipped.

[0149] In another implementation, there is an initial factory calibration setting that is shipped with the product. Again, cloud or network information may be monitored and a determination may be made as to how closely the actual parameters match the initial factory calibration setting. Adjustments may then be made to the factory calibration settings of sensors that have not been shipped based on this determined proximity. In this and the previous implementation, sensor product releases may be staggered so that subsequent shipments improve accuracy. This implementation may also be implemented even when all of the sensors have been shipped, as adjustments may be made through the network or through the cloud.

[0150] 9, a factory 148 is shown having a manufacturing lot or batch 150 of sensors, which are generally made with the same (or very similar) manufacturing process. The lot or batch 150 may be divided into a first portion 154 and a second portion 156. The first portion 154 may remain at the factory 148 temporarily, while the second portion 156 may be sent to a group of users 158.

[0151] 10, data from second portion 156 may be used at factory 148 to inform the factory calibration of first portion 154 and convert it to a calibrated first portion 154'. If first portion 154 has already been shipped, within a group of users 158, the first portion may be calibrated before or after insertion and is designated first portion 154''. Calibration of the first portion following shipment may be performed as described above with access to network or cloud resources for factory calibration information, particularly if it has been updated with data from sensors in the field.

[0152] 11 is a flow chart 160 illustrating the above method. First, a manufacturing lot or batch of sensors is manufactured under known and reproducible conditions in a factory (step 162). The lot or batch is split into at least two portions (step 164). For convenience, two portions are described here, but the manufacturing lot may be split into any number of portions for staggered release.

[0153] In this embodiment, the second portion is sent to the user (step 166). The second portion is then calibrated (step 168), which may be performed in a known manner, for example, using a priori information, bench calibration values, user data, fingerstick calibration, etc. Calibration may also be performed using techniques disclosed herein.

[0154] The calibration information from the second portion may then be sent to the factory (step 170). A calibration for the first portion of the sensor may then be generated or adjusted based on the data from the second portion (step 172). That is, if a factory calibration has been generated for the first portion, it may be adjusted as necessary. If a factory calibration has not been generated, the received data from the second portion may be used to inform the calibration of the first portion, such as an average of the determined sensitivity from the field. The adjustment or generation may be performed at the factory (step 174), or it may be performed following shipment, before or after insertion into a patient (step 176), with the transmitter, receiver, or other monitoring device, e.g., a smartphone, in network communication with a server or other network resource operated by the factory 148.

[0155] Once a sensor is inserted into a user and the initial calibration is complete, any subsequent calibrations are referred to as "ongoing" or "ongoing" calibrations. Schematically, this is shown in diagram 102 of FIG. 12. A user, patient, or host 112 has an indwelling sensor 114 that is connected to a transmitter 115. In many cases, the transmitter is used multiple times with different sensors. In other cases, the transmitter may be disposable.

[0156] The transmitter 115 enables communication of the signal measured by the sensor 114 to a device such as a receiver, or dedicated device 104, or smartphone 108. The receiver 104 is shown with a display 106, and the smartphone 108 is shown with a display 110. The displays 106 or 110 may be used to display a user clinical value of an analyte concentration, e.g., glucose concentration, relying on the relationship described above in which the measured current or counts are related to the clinical value of the analyte concentration by a linear relationship with a slope that represents the sensitivity.

[0157] Systems and methods according to the present principles describe the development or determination of this linear relationship based largely or exclusively on characteristics of the sensor signal itself, and in some implementations do not rely on external data, as previous systems did. In addition, such a "self-aware" system using "self" or "automatic" calibration can be used not only to more accurately measure subsequent analyte concentrations, but also to retroactively correct the results of previous measurements. In this way, when such a system is displayed on a display, such as display 106 or display 110, it more accurately conveys the measured data. In other words, retroactive processing may be used to correct or modify the previous calibration and even to update the data measured therefrom. In this way, when the display shows historical data as well as current data, at least the historical data is updated, i.e., the display changes to reflect calibration parameters that are known or more reliably known than the previous calibration.

[0158] This method is illustrated by a flow chart 116 in FIG. 13, where the first step is the receipt or determination of a seed value (step 118), which may be received or determined, for example, using the initial calibration procedure described above. The seed value may then be used to determine the calibration (step 120). For example, if the received calibration parameter is a particular value of slope or sensitivity m, it may be used to correlate the measured counts to clinical values ​​of analyte concentrations and may be used to immediately begin notifying the user of those measured analyte concentrations, e.g., glucose measurements. That is, an analyte may be measured using the sensor (step 122), and the measurements may be displayed to the user (step 124), based at least in part on the seed value received in step 118.

[0159] In some cases, a change in calibration occurs (step 126), which may be detected in a variety of ways, including those described below. The calibration, specifically the calibration parameters including sensitivity and baseline, may then be adjusted (step 128). After updating the calibration parameters, the display may be updated (step 130).

[0160] As mentioned above, updating the display may refer not only to adjusting the currently measured value of the analyte concentration, but also to recalculating the historical value and changing its display based on the adjusted calibration. For example, the sensitivity may be "seeded" with an initial value, but following receipt of data, it may be determined that it is actually 10% lower than the initial seed value. In this case, it is not only the ongoing displayed analyte value that is adjusted, but in one implementation, the historical value may also be adjusted to reflect the updated sensitivity. This example illustrates a situation in which the seed value is updated with a measurement value. In some cases, an already determined value (determined by the seed or measurement value) may be updated with a later determined value. This situation may arise, for example, when the sensor calibration parameters "drift." For example, if it is determined that the calibration of the sensor has drifted, changes may be made to the calibration parameters so that the receiver, smartphone, or other monitoring device continues to display an accurate value of the analyte concentration. In one implementation, if it can be determined when a drift has occurred, certain historical values, i.e., values ​​measured following the drift, may be updated in the display, while other historical values, e.g., values ​​measured before the drift, do not need to be updated.

[0161] In one implementation, if the determined seed value and the initial seed value are close, e.g., within 10%, the initial seed value (or other calibration parameter) may simply be adjusted accordingly, however, if the values ​​are further apart the user may be prompted for intervention, e.g., with an optional finger stick.

[0162] 14 and 15 are graphs showing analyte concentration over time before (14) and after (15) a change in sensitivity. Specifically, FIG. 14 shows a graph 132 in which a plot 138 of analyte concentration over time is shown. Axis 134 represents the value of analyte concentration and axis 136 represents time. Following a change in sensitivity, the graph becomes graph 140 with transformed historical analyte values ​​146. The change in sensitivity may be considered an updated sensitivity or an updated seed value, depending on the implementation. Several other methods of adjusting the calibration using sensor signal characteristics may be used, including, but not limited to, the mean sensor signal, the standard deviation of the sensor signal, or the CV (coefficient of variation) or interquartile range, or other higher order or rank statistics.

[0163] In more detail, and in contrast to previous efforts, preferred embodiments describe systems and methods for periodically or substantially continuously post-processing (e.g., updating) a substantially real-time graphical display of glucose data (e.g., a trend graph representing glucose concentration over the previous number of minutes or hours) with processed data that has been processed in response to calibration updates, e.g., as a result of sensor drift, system error, etc.

[0164] 16, specifically block 137, a sensor data receiving module, also referred to as a sensor data module, or processor module, receives sensor data (e.g., a data stream) including one or more time separated sensor data points. In some embodiments, the data stream is stored in the sensor for further processing, and in some alternative embodiments, the sensor periodically transmits the data stream to a receiver or other monitoring device, such as a smartphone, which may be in wired or wireless communication with the sensor. In some embodiments, raw and / or filtered data is stored in the sensor and / or transmitted and stored in the receiver.

[0165] In block 139, the processor module is configured to process the sensor data in various ways. The processor module may also be used in combination with a calibration module 143 to determine if a change in calibration has occurred, as described above and in more detail below. More particularly, in block 143, the calibration module uses the data in the data stream to detect changes in calibration, and more specifically, changes in sensitivity.

[0166] In block 141, the output module provides output to a user via a user interface (not shown). The output represents an estimated glucose value, which is determined by converting the sensor data into a meaningful clinical glucose value. The user output may be in the form of, for example, a numerical estimated glucose value, a directional trend in glucose concentration, and / or a graphical display of the estimated glucose data over a period of time. Other indications of the estimated glucose value are also contemplated, for example audio and tactile. In some embodiments, the output module displays both "real-time" glucose values ​​(e.g., numerical values ​​representing the most recently measured glucose values) as well as graphical displays of processed and / or post-processed sensor data.

[0167] In one embodiment, the estimated glucose values ​​are represented by numerical values. In another exemplary embodiment, the user interface graphically represents the estimated glucose data trends over a period of time (e.g., 1, 3, and 9 hours, respectively). In alternative embodiments, other time periods may be represented. In alternative embodiments, pictures, animations, charts, graphs, value ranges, and numerical data may be selectively displayed.

[0168] The processor module may be further configured to perform post-processing steps, such as to periodically or substantially continuously post-process (e.g., update) a displayed graphical representation of data corresponding to a time period according to received data, e.g., more recently received data. For example, glucose trend information (e.g., relating to the previous 1, 3, or 9 hour trend graph) may be updated to better represent actual glucose values ​​taking into account newly determined calibration values. In some embodiments, the post-processing module post-processes segments of data (e.g., 1, 3, or 9 hour trend graph data) every few seconds, minutes, hours, days, or anywhere in between, and / or when requested by a user (e.g., in response to a button activation such as a request to display a 3 hour trend graph).

[0169] Generally, post-processing involves processing performed by a processor module (e.g., in a handheld receiver unit) on "recent" sensor data (e.g., data including time points within the last few minutes or hours) after its initial display of the sensor data and prior to what is commonly referred to in the art as "retrospective analysis" (e.g., analysis accomplished retrospectively at one time, as opposed to intermittently, periodically, or continuously, on an entire data set, such as for display of the sensor data for a physician's analysis). Post-processing can include programming performed to recalibrate the sensor data (e.g., to better match a reference value), fill data gaps (e.g., data eliminated by noise or other issues), smooth (filter) the sensor data, compensate for time lags in the sensor data, etc. Preferably, the post-processed data is displayed on the personal handheld unit (e.g., on 1, 3, and 9 hour trend graphs on the receiver or smartphone, etc.) in "real time" (e.g., including recent data within the last few minutes or hours) and may be updated (post-processed) automatically (e.g., periodically, intermittently, or continuously) or selectively (e.g., in response to request) when new or additional information is available (e.g., new reference data, new sensor data, etc.). In some alternative embodiments, post-processing may be triggered as a function of the duration of the calibration episode change, e.g., data associated with a calibration event change that extends beyond about 30 minutes may be processed and / or displayed separately from data for the first 30 minutes of the calibration episode change.

[0170] In one exemplary embodiment, the processor module filters the data stream to recalculate the data for a prior period of time and periodically or substantially continuously displays a graphical representation of the recalculated data for that period of time (e.g., a trend graph). In another exemplary embodiment, the processor module adjusts the data for a time lag from the data for the prior period of time (e.g., removes the time lag induced by real-time filtering) and displays a graphical representation of the time lag adjusted data for that period of time (e.g., a trend graph). In another exemplary embodiment, the processor module algorithmically smooths one or more sensor data points over a moving window (e.g., including time points before and after the one or more sensor data points) relative to the data for the prior period of time and displays a graphical representation of the updated, averaged, or averaged data for that period of time (e.g., a trend graph).

[0171] In some embodiments, the processor module is configured to filter the sensor data in response to determining the start of the calibration event change and display a graphical representation of the filtered sensor data. In some embodiments, the processor module is configured to display a graphical representation of unfiltered sensor data (e.g., raw data) in response to determining the end of the calibration event change. In some embodiments, the processor module is configured to display a graphical representation of unfiltered sensor data except when a calibration event change is determined. It has been found that the adaptive filtering described herein, including selective filtering during the calibration event change, increases the accuracy of the displayed data, reduces the display of noisy data, and / or reduces data gaps and / or cuts off sooner compared to conventional sensors.

[0172] Calibration Routine As mentioned above, it is desirable to provide a more convenient calibration routine for users, particularly Type II users or users using systems for weight loss optimization and / or sports and fitness optimization.

[0173] One way to reduce the need for user-based calibration is to use more enhanced factory calibration, and certain details regarding methods related to factory calibration can be found in USSN 13 / 827,119, filed March 14, 2013, published as US 2014-0278189-A1, and USSN 62 / 053,733, filed September 22, 2014, both of which are owned by the applicant hereof and are incorporated by reference in their entireties herein.

[0174] Other techniques may also be used to facilitate the calibration requirements. For example, referring to flow chart 145 of FIG. 17, if the previous sensor session showed generally reliable results (step 147), the same calibration parameters may simply be used for the new sensor session from the previous session (step 149). Specifically, the calibration parameters from the old sensor session may be transmitted to the new sensor session in a number of ways, for example, by using glucose signal transmitter technology if the calibration parameters are stored on the sensor electronics, or by conveying calibration state variables to the new session if the calibration parameters are carried on a monitoring device, e.g., a smartphone. This technique may be particularly useful if the sensors are somewhat related, e.g., from the same batch of the same type, the same package, etc.

[0175] This technique is not necessarily limited to only using a single previous session, e.g., going back to only one session. For example, recurring patterns, e.g., historical patterns, may be learned by analysis of several previous sensor sessions. For example, the system may learn that a user has a habit of eating pizza on Fridays and has done so over the last seven sessions, and thus the algorithm may learn not to treat such events as outliers. Patient habits may also be learned, e.g., that the patient likes to have many small meals versus only a few large meals. Failure modes may also be learned, e.g., if the patient is prone to a particular failure mode due to a particular way of attaching and / or using their device. As discussed below in connection with FIG. 18, certain glucose trace characteristics that constitute a recurring event may be advantageously learned from previous sensor sessions, and often multiple previous sensor sessions are necessary to distinguish common events from outliers. In addition, if other event data, e.g., meal or exercise data, is available, correlations between such recurring events and the input meal or exercise data may be learned.

[0176] Referring now to FIG. 18, a flow chart 178 for another method of calibration is shown. In particular, certain glucose trace characteristics are known to exhibit repeatable events that, when they reoccur, reoccur at known and repeatable glucose values. As an example, steady state values, certain trend values ​​including certain slopes, etc. tend to be repeatable for a given patient. Such repeatable events tend to give rise to characteristic glucose trace properties and / or patterns that are repeatable and detectable. For example, it is a feature of many biological systems that analyte values ​​are highly repeatable when they are not changing rapidly, i.e., when they are at a steady state. In other words, when an analyte value is at a steady state for a first time and then at a steady state for a second time period, the analyte value, e.g., concentration, will generally be at or near the same value during each of the steady state periods. This concept can be used to calibrate an analyte sensor.

[0177] For example, when a user is at steady state with respect to an analyte, the value of the analyte may be measured and stored, and when the user is again at steady state, their analyte value is likely to be the same as the previously measured value, and the sensor reading the analyte concentration value may be calibrated accordingly.

[0178] Different analytes may achieve different steady states in various ways. For example, uric acid concentrations change very little throughout a typical day. If a user has not exercised or eaten for several hours, their glucose values ​​may be at a steady state. If a user has not exercised for several hours, their lactate values ​​may be at a steady state. In general, many analyte values, including glucose, achieve a steady state when a biological system has not changed significantly over a period of time. As described, the steady state values ​​are reproducible, especially for prediabetic or non-diabetic patients, and for humans who are primarily using the system for weight loss or sports optimization. Thus, whenever a steady state is detected in the analyte value, the sensor measuring the analyte may be calibrated.

[0179] In some cases, the system may prompt the user to fast or exercise to achieve a steady state, which may then be measured and subsequently used for such calibration. Additionally, the system may detect a steady state but prompt the user for verification, for example, by asking the user, "Are you fasting?"

[0180] In some cases, the steady state value may be determined based on the demographics of the user and therefore does not require any measurement. For example, for a non-diabetic patient, a typical glucose value may be about 80-100 mg / dL. Often, if a person is very non-diabetic, their value may be around 80, while if they are progressing to pre-diabetes, their value may approach 100. Thus, providing the system with only certain information about the user may allow a degree of calibration to be performed, particularly for certain applications, including where the user does not need to have the accuracy determine the exact value, but rather where accuracy only to certain ranges is sufficient, e.g., hyperglycemic, hypoglycemic, or euglycemic.

[0181] Apart from steady state, other repeatable events that may be used include slope, typical or similar post-meal response, e.g., post-breakfast response if the user has the same type of breakfast every day. Other repeatable events include, for example, range of low to high measurements each day, i.e., daily high and low ranges, certain types of excursions, certain types of temporal patterns, decay rates, and slopes, rates of change, etc.

[0182] As a specific example, if a user has a characteristic breakfast, which results in a characteristic post-prandial glucose trend, if a change in the trend occurs, it can be inferred, at least to some extent, that a change in the sensitivity of the sensor has occurred and the sensor needs to be recalibrated.

[0183] Additional data may also be used to aid in calibration. For example, if the user is diabetic and somehow measures their blood glucose several times a day, such values ​​may be used as calibration values. This is especially true if their blood glucose values ​​vary significantly. In this case, if the system detects a local steady state, it may prompt the user to measure and enter a blood glucose value in order to correlate the value with the steady state.

[0184] Historical data may in some cases be used to determine steady state calibration values, for example, previous data from a blood record, data from a previous session, estimates from measurements such as A1C that track glucose averages over time, etc.

[0185] In some cases, it is not necessary to focus on a particular value. A determination that the user is within a certain range of values ​​may be sufficient for a Type II user. The range may be determined and used to provide the user with information about whether a goal is being achieved or about the program they are undergoing.

[0186] The use of steady state estimation can be used not only for initial calibration, but also for update calibration: whenever the system is considered to be in a subsequent steady state, the values ​​measured during the subsequent steady state can be assumed to be the values ​​originally determined for the user. Other calculations can also be used, including the use of weighted averages, slow moving averages (see below), etc.

[0187] Referring to FIG. 18, the method of flowchart 178 allows for calibration of glucose concentration or other analyte values ​​using information generated directly from the device, ie, the analyte concentration signal itself.

[0188] Thus, with reference to flowchart 178, the first step in a calibration routine according to these principles is to detect that the system is in a steady state (step 180), for example a first known steady state.

[0189] Generally, steady state may be detected without any action by the user (step 182). In some cases, however, steady state may be prompted to allow steady state to be achieved (step 184), for example, by waiting a predetermined period following a meal or exercise routine (step 186), or by waiting a predetermined period after an event. The system or routine may request or prompt user information regarding, for example, fasting, for example prompting the user to enter the duration since the user's last meal or related parameters. In the case of glucose, the routine may be configured to require a low rate of change, or a rate of change below a predetermined value, for example, less than 0.25 mg / dL per minute. In some cases, if calibration has not yet occurred, the rate of change may be based on counts, or microamps, or other "raw" signal values. The rate of change may be determined by a calculation of the derivative.

[0190] Once a steady state is detected, calibration parameters such as sensitivity may be determined (step 188) based on the measured counts (or current) at the steady state and the known steady state glucose value, which may be known or estimated. Another steady state may then be detected (step 192) and, if drift has occurred, the second steady state is associated with a different count number. The different count number may be used to determine the extent of drift, and the system may be recalibrated using the new (second) count number measured at the new (second) steady state along with the already known steady state glucose value (step 194).

[0191] In some cases, if the change is substantial, e.g., exceeds a predetermined threshold, the user may be prompted for additional data, e.g., fingerstick, exercise or diet data, etc. (step 196). The difference may also be used to determine the cause of the drift (step 198).

[0192] A specific example will now be described. A continuous glucose monitor may be used for a patient who does not have diabetes. In patients who do not have diabetes, their glucose values ​​are typically in the range of 70-90 mg / dL. Their glucose levels only deviate from this range after meals or during vigorous exercise, and in some cases such deviations may be detected by changes in the sensor signal or via auxiliary measurements using a heart rate monitor or accelerometer. During these glucose excursions, the current measured by the sensor changes quickly, which can be easily detected by the monitoring device. The rate of change may be calculated, for example, using an FIR filter over the glucose values ​​of the last 20 minutes, but may range from a simple rate of change defined as the difference between two glucose values ​​at the start and end of the period divided by the duration of the period. During rapid rates of change, systems and methods according to the present principles may avoid such rapidly changing data for calibration if they use steady state occurrences as repeatable events for calibration, but rather may wait until values ​​stabilize before performing a calibration event (as noted, certain calibration methods may take advantage of such. For example, as noted above, in some cases repeatable events usable for calibration may include transient noise events or patterns. That is, certain aspects of high change regions or peaks may be useful for non-steady state repeatable event calibration, and often transient events that include rapid or slow rates of change in analyte concentration values ​​may form signal characteristics whose patterns are estimated and used as repeatable events). In one example, the absolute rate of change threshold may be set at 0.25 or 0.5 mg / dL / min over the last 25 minutes of glucose data. As noted above, uncalibrated units may also be used. Thus, calibration may be inhibited if the absolute rate of change threshold is exceeded. In other implementations, a different calibration value may be used, which may also be configured to depend on the direction of the rate of change, or the calibration value used may be a function of the rate of change itself.

[0193] With reference to FIG. 19 and as described with respect to step 194 of FIG. 18, the steady state may be used to update the calibration values ​​and to determine the initial values. Furthermore, the system may be used even when the calibration parameters change, for example, when the sensor enzyme layer changes over time with use or other drift occurs. For example, with reference to graph 202 of FIG. 19, the value of the analyte may be related to a calibrated or uncalibrated value at time t0 (axis 206), and this steady state value may be presumed to be reproducible whenever the user is at a steady state in that analyte value. This knowledge may result in an initial calibration line 208, the slope of which is the sensor sensitivity if the axis 204 has units of counts. At a subsequent time t1, which is also presumed to be at a steady state, even if the number of counts measured is different, knowledge of the same steady state value allows a subsequent calibration line 210 to be drawn and the system may thus be recalibrated. The degree of recalibration required may be very useful in determining the cause and treatment of the drift.

[0194] Systems and methods according to the present principles may be most effective when the baseline or background signal is sufficiently stable and predictable (or eliminated through advanced membrane or sensor technology). When the sensor is activated, a current is generated. If the baseline is small enough or estimated with sufficient accuracy, the remaining current may come from the analyte of interest. The algorithm may measure the current over a set period of time, and if the current is stable, for example, within a predefined limit, the algorithm may infer that the glucose (or other analyte) is not changing and is within a narrow range. The algorithm may then use, for example, a glucose value of about 80 mg / dL (or whatever is determined to be a typical value for the user) and correlate it with the current generated during that time of having stable glucose to automatically calibrate the device. In one implementation, the glucose value was set to 100 mg / dL (it is iterated that this may vary depending on factors such as rate of change, duration of wear, time of day, characteristics of the user, e.g., state of progression to diabetes, etc.). The system and method may use a regression model to calculate the slope and baseline at two points. The first point is the current generated during stable glucose (and approximated glucose level for a non-diabetic user), and the second point is zero glucose (using an estimate for the background signal). The slope of the line may be determined, for example, using the regression slope (estimated for counts / BG) and a weighted average of the previous slope estimates. Following calibration, the glucose data may be presented to the user. The baseline in this implementation was estimated to be zero, although a different non-zero baseline value may be used. Calibration may also be updated periodically, for example, every few minutes or hours depending on the implementation.

[0195] The above technique may be used in combination with factory calibration information generated during manufacture of the device, or it may also be used with externally generated glucose information, which may further account for changing sensitivity over time by incorporating predetermined drift curves or other drift compensation techniques, as described in more detail in U.S. Patent Application No. USSN 13 / 446,848, filed April 13, 2012, and published as US 2012 / 0265035-A1, which is owned by the present applicant and is incorporated by reference in its entirety herein.

[0196] Systems and methods according to the present principles may further be used to calibrate another, e.g., adjacent, sensor, e.g., under the same membrane, using the calibration information for one sensor. Such a calibration may be performed when drift parameters, if caused by the membrane, may be presumed to be the same for both sensors. For example, if both sensors, e.g., a glucose sensor and a lactate sensor, are under the same membrane layer, and one or more calibration parameters have been determined ex vivo, the calibration parameters may be presumed to have a similar relationship in vivo, and thus measurements of one may be used to determine the other. For example, if the lactate sensor has a known offset of calibration (or other relationship, or scaling, or correlation coefficient) from the glucose sensor as measured ex vivo and then in vivo, the determination of the calibration for that glucose sensor may be used to calibrate the lactate sensor. For example, if the calibration of the glucose sensor is deemed to drift 50%, then the calibration of the lactate sensor may be presumed to have drifted 50%. As a result, updating one or more calibration parameters of one sensor may result in updating one or more calibration parameters of the other sensor.

[0197] Further details of such embodiments can be found in U.S. Patent Application Nos. USSN 12 / 770,618, filed April 29, 2010, published as US-2011 / 0004085-A1, and USSN 12 / 829,264, filed July 1, 2010, published as US-2011 / 0024307-A1, and [545PR], all of which are commonly owned and incorporated by reference in their entireties herein.

[0198] Additionally, systems and methods according to the present principles may start using factory calibration information and then incorporate automatic calibration techniques over time to obtain more accurate glucose information. If the signal does not follow or is outside of the predetermined parameters, the system may request a calibration value using known techniques, e.g., SMBG or fingerstick calibration. The system and method may then incorporate this glucose information into the original parameters and adjust the set point, e.g., from 100 mg / dL to a more appropriate and accurate value. That is, while the above techniques intended for use in certain applications may generally be configured to avoid the need for fingerstick calibrations, if available, systems and methods according to the present principles may advantageously apply them for calibration purposes or otherwise.

[0199] Systems and methods according to the present principles may be configured to determine a confidence level or range, which may change as the resolution or accuracy of the data changes. More specifically, the display may generate a value and trend graph, or it may show a range or other UI element. The range may change over time, contracting or expanding as confidence in the accuracy changes. For example, during initial warm-up, a factory calibration value may be utilized. However, the accuracy may not be accurate enough to have additional information. During this time, the display may show a range rather than a value.

[0200] A further feature of systems and methods according to the present principles is that they may request information when the user is setting up within the system and adjust which technology to use depending on that information. For example, the system may prompt the user to input whether the user has type I diabetes, type II diabetes, or is non-diabetic, and may select different technologies depending on the answer. Systems and methods according to the present principles may further ask if the user is interested in weight loss optimization, sports and fitness optimization, or other similar optimization routines, and may adjust the algorithms accordingly. The device may also be used in a "blind" mode, for example, for an extended period of time, such as 14 days, and only receive blood glucose values. These blood glucose values ​​may be used to learn what the patient's resting blood glucose is, which may better guide the automatic calibration's steady state blood glucose value estimation. After an extended period of time, the user may then use the device in automatic calibration mode.

[0201] Besides using the characteristics of the steady state value of the analyte to gather additional information, a "slow moving average", i.e., an average value measured over, for example, 1-3 days, may also be used, since such a slow moving average is also generally constant, especially over the use of a sensor session. Its variation may therefore be used both qualitatively and quantitatively to detect and quantify drift. For example, the value of the slow moving average of glucose concentration is shown by graph 220 in FIG. 20 and by the more schematic graph 212 in FIG. 21, where the sensor count is shown on axis 214 against time on axis 216. As can be seen, the slow moving average G1 measured at time t1 may be reduced to a slow moving average G2 at time t2. The slow moving average may be used to quantify drift, since the advanced sensor is highly selective for glucose, and therefore the only thing that contributes to the change in the slow moving average is the sensitivity.

[0202] The details of the use of the slow moving average are described in detail below, but note here that it need not have a continuous period. For example, the slow moving average may be measured by sampling a common period over many days. For example, a slow moving average of a nighttime period may be measured, which may then only consider, for example, the period from 11 pm to 7 am. The slow moving average then constitutes an average of the data measured over several days, but only during this period. Other exemplary periods during which such separate or non-continuous slow moving averages may be measured may include, for example, post-meal periods, etc. Such events may be time-based, where a user has very consistent timing of such events, or may be event-based, where, for example, events are recorded or tracked by one or more sensors. For example, movement events may be recorded by an accelerometer, and meal events may be recorded by detecting glucose spikes.

[0203] More specifically, instead of using daily averages, averages may be measured that are specific to a qualitative or quantitative type of period that is specific and important to the user, for example, a period may be used that is specific to the considered time frame over which the average is measured. For example, the time frame may be considered as "4 hours after a meal". Period data may be used that is measured over many days, but only considers that particular period, and therefore measures only the variation from the average of the glucose response during that particular set of periods. In other words, the average may be determined by putting together and averaging all of the glucose values ​​from individual periods over several days. In this way, if drift is detected, it is defined in terms of the average value obtained by measuring the average over such similar periods. For example, a patient's daily average may be 100, but their nighttime average may be 85, and their "regular" walking daytime average may be 120. Measurement of drift may be made in terms of this defined "local" average. Exemplary periods may include, for example, after dinner, 9 am to noon, during sleep, etc.

[0204] A slow moving average of the calibrated or even uncalibrated values ​​may be measured, but the slow moving average may itself require additional data to perform the initial calibration. For this reason, implementations may include determining the initial glucose value using other methods and obtaining, for example, a day's worth of data, from which an initial slow moving average may be determined, which may then be compared to subsequently measured slow moving averages to determine, for example, sensitivity drift corrections, etc. Such systems and methods may be particularly beneficial because the slow moving average may be checked very frequently, for example, every 5 minutes, compared to previous systems, as opposed to SMBG calibration, which may only be performed as frequently as the user is willing to measure readings.

[0205] Flowchart 222 of FIG. 22 illustrates one implementation of the use of a slow moving average. In a first step, an analyte value is measured using a sensor (step 224). A first slow moving average may then be determined (step 226). Analyte values ​​may then continue to be measured (step 228), which may form a reference for a displayed value of analyte concentration (step 232), the displayed value being based on the initial (or subsequent) value of sensitivity. A second and subsequent slow moving average may then be determined (step 230), the period of the slow moving average being generally more than half a day, e.g., 1-3 days. The slow moving average may often be measured as frequently as required, e.g., every 5 minutes, every hour, etc. In some implementations, the time constant of the filter may be changed, for example, when the user has an actual peak value, so that the effect (user's high analyte concentration) does not represent an actual drift or change in the sensitivity of the filter. The use of a slow moving average may also be replaced by other forms of filtering, such as order statistics filtering or time domain filtering.

[0206] If the first and second values ​​(or second or subsequent values, or indeed any set of values) of the slow moving average change, in one implementation, drift may be presumed to have occurred, and the calibration may be adjusted based on the drift, e.g., the difference between the slow moving averages (step 234). Implementations described below consider other potential causes of variation in the slow moving average. The display may be updated (step 236) based on the recalibrated sensor (recalibrated based at least in part on the measured drift). In some cases, the seed value (or other value used as a basis for the display) may also be modified to reflect (and compensate for) the drift (step 238).

[0207] Referring to flow chart 240 of FIG. 23A, the recalibration (or other recalibration) based on data collected from the change in the slow moving average may also be used in post-processing to update the display of the historical values, defined herein as "already displayed" values, based on the recalibration. In a first step, a value of analyte concentration may be measured by the sensor (step 242). The measured value may be displayed as a clinical value based on the calibration, e.g., based on a previously determined sensitivity value (step 244). The sensitivity may change based on (including solely based on) the sensor signal information (step 246), e.g., based on a change in the slow moving average or the steady state value. The display may then be updated based on the change (step 248), specifically, the historical already displayed values ​​may be redisplayed based on the recalibration so that the historical values ​​are more accurately represented. Other forms of using signal characteristics (or features) to detect sensitivity changes (or drift) include the CV (coefficient of variation), standard deviation, or interquartile range of the signal and its relationship to the corresponding glucose parameter.

[0208] Once a change is detected, it may be analyzed or "discriminated" to determine the cause and / or magnitude of the change. It is common to find a change in sensitivity due to drift, but it may also have other causes, including pump problems, e.g., pump blockage, or other problems, e.g., membrane breakage, etc. It is also desirable to distinguish these changes from those due to actual changes in glucose values. As at least a first step in this latter determination, the measured change in glucose value may be compared to a threshold for such a change that is physiologically feasible. If the change is not physiologically feasible, the change may be deemed to be due, at least in part, to drift or a system anomaly.

[0209] Another way to distinguish signal drift behavior is by comparing the signal drift curve with known signal drift curves, specifically, the envelopes of such curves. FIG. 4 shows one such curve, but for a given type of sensor, there are envelopes of such curves, i.e., there are patterns of how the sensitivity changes. If the way the sensitivity is changing follows one of these curves, it can be inferred that the change is due to sensitivity drift rather than actual glucose value or system anomaly. Further details regarding sensitivity profile curves are described, for example, in U.S. Patent Application No. 13 / 796,185, entitled “Systems And Methods For Processing Analyte Sensor Data”, filed September 19, 2013, owned by the applicant of the present application and incorporated herein by reference in its entirety.

[0210] If the sensitivity changes by shifting to another sensitivity of a known sensitivity curve, the known calibration parameters for that curve may be used in subsequent data analysis and display. If the sensitivity changes outside the limits of the known sensitivity curve, the change may be inferred to be due to a system problem or anomaly, e.g., an error or artifact, as described above. However, in a particular implementation, a particular sensor may have a sensitivity curve that is known to be within a band. A known failure mode may shift the sensitivity to another curve within the band or to another band, i.e., the sensitivity may shift to a known other band of the curve with a known failure mode.

[0211] In this regard, it should be noted that, in general, a given type of sensor is functional to achieve the required goal up to a certain tolerance. For example, 80% of the sensors of a lot may function as required (see FIG. 23B). The remaining 20%, on the other hand, may not follow the predicted sensitivity curve (see FIG. 23C). A majority of these remaining sensors, e.g., 75%, may follow a known failure mode, which results in them following a known alternative sensitivity curve, family of curves, or band of curves. By identifying which of these sensors follow an alternative sensitivity curve and adjusting the sensor calibration accordingly, the "failure" of the sensor can be largely ameliorated. For example, if the failure mode is such that all 75% have signal values ​​that tend to decrease in the same way, after the failure mode is determined, the "failure" can be ameliorated by adjusting the sensor reading "up". Such an aspect may be particularly important as sensor sessions become longer and longer, e.g., from 7-day sessions to 14-day sessions. In the failure mode shown in Figure 23C, the decrease in sensitivity begins at about day 8. The ability to detect and ameliorate such failure modes is particularly important because even as the onset of session sensor failure modes becomes increasingly better characterized, the end of session sensor failure modes, especially with longer sessions, remains difficult to quantify.

[0212] In some implementations, consideration of glucose signal variability in combination with the slow moving average may be used to distinguish glucose signal fluctuations from fluctuations in sensor sensitivity. For example, if the slow moving average decreases but the variability remains the same or remains within a specified range or band, the cause of the decrease in the slow moving average is likely to be a sensitivity change. Alternatively, if the slow moving average decreases but the variability also decreases, the cause of the decrease is likely to be a real and actual change in glucose concentration.

[0213] Changes, fluctuations, errors, artifacts, and other signal behaviors that are not physiologically feasible may be the source of various remedial actions by the system, some of which have been mentioned above. For example, a recalibration may be performed and the results may be propagated backwards towards historical data. The user may be prompted to provide a fingerstick calibration. Part of the remedial action algorithm may be to determine whether to correct via recalibration or to prompt for a fingerstick or other calibration point. For example, if a signal is received that is outside of physiologically possible limits, the user may be prompted for an additional calibration point, e.g., a fingerstick. Alternatively, the user may be prompted to provide other types of additional information, e.g., to input data corresponding to recent exercise or meals or other recent changes in user behavior. As a specific example, if the slow moving average of the user's glucose concentration was 100 mg / dL for the first three days of a session but suddenly became 200 on the fourth day, systems and methods according to the present principles may prompt for a fingerstick calibration. If the slow moving average on day 4 was 105 mg / dL, the system may increase / decrease or adjust the sensitivity accordingly, for example, lowering the value again to 100 mg / dL.

[0214] If fingerstick calibration is performed, the use of fingerstick calibration data may vary depending on the use of the device. For example, if the device is used adjunctively, i.e., non-therapeutic (which is the case for many type II users), if the fingerstick shows drift, it may still be possible to use the sensor if the drift is not substantial. In many cases, the techniques described herein may be used to ameliorate the effects of drift and still allow accurate readings to be displayed. If the device is used non-adjunctively, e.g., for type I patients who use insulin, if the fingerstick shows drift, improvements or adjustments, e.g., recalibration, may be made more aggressively, and if this cannot be done or accurate sensor readings cannot be returned after recalibration, instructions to discontinue use of the sensor may be displayed to the user.

[0215] Pattern analysis may be performed to determine if the change or variation is of a known type, e.g., characteristic of a known sensitivity change. Pattern analysis may determine if the change or variation meets criteria for a certain behavior, e.g., exceeds a known certain threshold. As described above, the behavior of a day may be used in determining the slow moving average. If, after analysis, the system determines that the data for that day is not reliable, another day's data may be used. The data may be displayed in ranges or bands, or with confidence intervals or other indicators, rather than as highly accurate numbers. If a slow moving average is used, the time constant of the slow moving filter may be adjusted to include or exclude shorter term variations.

[0216] These aspects are summarized in flow chart 250 of FIG.

[0217] Referring first to FIG. 24, an analyte concentration value may be measured by a sensor (step 252). Such values ​​are generally measured as current, e.g., as amperes (picoamps), and equivalently as counts. A slow moving average may be defined by measuring the counts over an extended period of time, e.g., several hours, half a day, 24 hours, or 2-3 days. Because the sensor varies from unit to unit, the slow moving average generally only becomes meaningful once an initial period, e.g., 24 hours, has passed. Thus, the next step is to determine a first slow moving average over a first period of time (step 254).

[0218] In one implementation, a value for initial use of the first apparent sensitivity may be assumed via a seed value as in the method described above. The first apparent sensitivity may then be determined based on the first slow moving average (step 256). Specifically, if a first slow moving average is assumed, the first apparent sensitivity may be based on a relationship between the assumed first slow moving average and the measured first slow moving average.

[0219] A second slow moving average may then be determined over a second period of time (step 258), and again, optionally, a second apparent sensitivity may be based thereon (step 260).

[0220] If the slow moving average or apparent sensitivity is deemed to change between the first and second time periods, remedial action may be required. Thus, the next step is to determine whether the change is consistent with a predetermined criterion (step 262). The predetermined criterion may include many factors, such as known sensitivity changes over time, the envelope of the sensitivity profile curve, possible physiological changes, changes associated with errors such as pump anomalies, etc. (step 270). For example, such a step may require a determination as to whether the change is consistent with a criterion associated with sensitivity drift, anomalies, or an actual change in the average glucose concentration value (step 264). If consistent with drift, e.g., if the sensitivity is determined to have drifted by comparison to a known sensitivity profile curve, a correction may be made automatically. In either case, the sensitivity may be adjusted based at least in part on the difference between the two slow moving averages (step 268), since that difference provides a quantitative indication of the degree of change or drift the sensor has undergone.

[0221] If the change is not consistent with drift, it may be determined whether the change is consistent with other causes for which predefined criteria exist. If the change is not consistent with known behavior, e.g., does not match any predefined criteria, the user may be prompted to enter information to explain the change, e.g., diet or exercise information, fingerstick calibration values, data from other external sources, etc. (step 266). In some cases, the data entered by the user may be used in a recalibration routine, e.g., along with a slow moving average or quantitative differences in sensitivity, to determine a new or updated sensitivity.

[0222] 25 shows another flow chart 286 of an exemplary method using a slow moving average or steady state value. In a first step, after an initial calibration, analyte concentration values ​​continue to be monitored at the sensor (step 288). The initial calibration may be based on a number of factors, including a population average, data from a previous session, bench data, in vitro data, or other a priori data (step 290).

[0223] Based on the measurements and the initial value, a clinical value of the analyte concentration is calculated and / or displayed. An updated calibration may then be calculated based on only the measurements, e.g., only the signal from the sensor (step 294). The adjustment may be based on a change in a slow moving average, a change in a steady state value, or other criteria.

[0224] Following the updated calibration, the clinical values ​​may be calculated and / or recalculated based on the updated calibration (step 298) and the display may then be updated (step 300), including updating already displayed (historical) values ​​to updated values ​​based on the updated calibration.

[0225] Yet another implementation of the present principles is illustrated by the flow chart 302 of Figure 26. The first step is to receive on the monitoring device a seed value for a calibration parameter, e.g., sensitivity (step 304). The seed value may be based on a number of factors, including user self-characterization of the condition, population averages, data from previous sessions, bench data, in vitro data, or other a priori data (step 306).

[0226] The monitoring device may then continue to receive sensor data and detect when the analyte concentration value is at steady state (step 308). For example, this may be done when a set of received signals over a predetermined period of time is within a predetermined range or band of values. A correlation of the measured signal value, for example in current or counts, with a known or estimated steady state value may then be performed (step 310).

[0227] Following the correlation step, the monitoring device continues to receive signals from the sensors (step 312). A clinical value of the analyte concentration is calculated and displayed based on the received signals, the known or estimated steady state value (even if the host is no longer in steady state), and the seed value (step 314).

[0228] Behavior may be detected outside of predefined parameters, as described above in connection with FIG. 24, and the user may be prompted to input external data, e.g., fingerstick calibration values ​​(step 316). Recalibration may be computed and / or recalculated based on the received signal, the external data, and optionally a seed value, followed by subsequent display (step 318). In some implementations, the known steady state values ​​and / or seed values ​​may be reset and redisplayed based on the calculations performed and the recalculated historical values ​​(step 320).

[0229] 27-33 show detailed methods for determining calibration parameters, such as sensitivity and baseline, using a probabilistic approach. Certain aspects of the probabilistic approach are described in U.S. Patent Application No. 13 / 827,119, entitled "Advanced Calibration For Analyte Sensors," filed March 14, 2013, owned by the present applicant, and incorporated herein by reference in its entirety. In this application, which is incorporated by reference, including what is referred to herein as a "signal-based calibration algorithm," a priori calibration distribution information is converted into a posteriori calibration distribution information, which is modified with real-time inputs and from which calibrated data points are determined. In this manner, calibration errors can be avoided, for example, when regression results in deviating sensitivity and / or baseline values ​​due to improper estimations on reference data. In addition to the methods described in the applications incorporated by reference above, other methods of determining calibration parameters, such as sensitivity and baseline, can also be used. These methods include techniques that allow for factory calibration that can be used for the duration of a sensor session.

[0230] In Figures 27-33, distributions are again used for calibration parameters such as sensitivity and baseline, which are optimized based on data that is then known, e.g., the sensor count distribution acquired over the first 24 hours of sensor session use. Referring initially to flow chart 322 of Figure 27, the first step is to receive an initial value of analyte concentration from the sensor, as well as an initial value or distribution of sensitivity and optionally baseline (step 324). In some cases, the effect of the baseline may be reduced to essentially zero to simplify the calculation. The initial value of sensitivity may be from the sources mentioned above, e.g., entered by a user, drawn from a population average, transferred from a previous session, or may be other sources of seed values ​​(step 326). The initial value may also be used as part of the criteria for a slow moving average filter or in the calculation. If the initial value of sensitivity or baseline is a distribution of such values, it may be developed, at least initially, from demographic considerations.

[0231] The analyte signal is then monitored from the sensor (step 328). A number of clinical values ​​are then calculated and displayed (step 330) based on the monitored signal and the initial value or distribution of sensitivity, or the initial value of average glucose. For simplicity, the baseline is presumed to be zero or negligible. The initial value or distribution of sensitivity is assumed such that a displayed analyte concentration value can be provided to the user even if it is less accurate during this initial period than thereafter once additional data has been collected.

[0232] A value distribution of the monitored signal may then be determined (step 332). A representative value, e.g., mean, median, mean, etc., of the value distribution of the monitored signal may be calculated (step 334). As an initial sensitivity, the representative value may be divided by an initial value of the analyte concentration based on an initial assumed sensitivity (step 338).

[0233] An initial value or value distribution of the sensitivity and / or multiple concentration values ​​may then be optimized to match the value distribution of the monitored signal (step 336). More specifically, the sensor count is the product of the sensitivity and the analyte concentration, and thus the concentration is equal to the sensor count divided by the sensitivity. A median sensor count may be determined, for example, after a day of data has been acquired, and a search may be performed that optimizes or provides the best fit to the distribution of sensor counts given the distributions or samples from the sensitivity parameter space and the baseline parameter space. For example, if a user's long-term glucose values ​​over the course of a day were in the range of 100-200, certain limits may be inferred as to what the sensitivity and baseline could be. Thus, the sensitivity and long-term glucose values ​​within the distribution may be selected such that their product best optimizes the measured representative value of the sensor count. Additionally, the sensitivity and long-term glucose values ​​may be selected such that their values ​​are "most likely" (step 340), meaning that their values ​​are closest to the center of their corresponding distributions. The slow moving average may be monitored (step 342) and changes therein detected and analyzed as described above (step 344).

[0234] In other words, there is data regarding an assumed initial average glucose value, or sensitivity, and a distribution of sensor counts following day 1. From the distribution, a median sensor count can be obtained.

[0235] Sensor count SC=f SC (SC), which has a normal distribution.

[0236] The sensitivity equation has the following form:

[0237] y=mx+b

[0238] If b=0 is assumed and the equation is further specified to be the mean value,

[0239] Median sensor count = m * mean glucose value

[0240] Also, both the sensor count and sensitivity are considered as normally distributed.

[0241] f SC (SC)=m*GV

[0242] Furthermore, m is known to be a slow function of travel time due to drift, and therefore

[0243] f SC (SC)=m(t)*GV

[0244] It is therefore apparent that the sensor count and glucose value are connected by a multiplicative "constant" that is in fact a slow transit time function.

[0245] The distribution of sensor counts may also reveal the potential distribution of sensitivities, i.e., the potential initial sensitivities m, and specifically the following aspects:

[0246] Mean GV = (median SC / m 中央値 )

[0247] therefore,

[0248] m 中央値 = Median SC / Average GV

[0249] For example, if the median SC is 131,000 and the mean GV is 131, then m is 1000 counts / (mg / dL). The distribution of m may be checked to determine if this value is reasonable or unlikely. A similar determination may also be made for "b" if it is not negligible.

[0250] Representative values ​​of the monitored signal (sensor count) may be converted into an estimate of long term glucose.

[0251] Long Term Glucose = (Long Term Sensor Count) / (Sampled Sensitivity)-Sampled Baseline (mg / dL)

[0252] Graph 346 of FIG. 28 shows an exemplary distribution of sample sensitivity, graph 348 from FIG. 29 shows an exemplary distribution of sample baselines, and graph 350 of FIG. 30 shows an exemplary distribution of sampled long-term glucose values. As an example, if the representative sensor count is 113,536, given the three distribution constraints described above, the optimal slope is 890 counts / (mg / dL), the optimal estimate of long-term glucose is 153.5685, and the optimal baseline is -26 mg / dL. These points are shown on the same set of graphs 346, 348, and 350, and are reproduced on FIGS. 31, 32, and 33 as points 354, 358, and 362, respectively.

[0253] In variations, the distribution may be made more "granular" such that different distributions may be provided for different demographic populations or groups. Other variations will also be understood.

[0254] As mentioned above, a slow moving average filter may be used as part of drift quantification since advanced sensors are highly selective to glucose and therefore the main factor contributing to the change in the slow moving average is sensitivity. To initially seed the slow moving average filter, an initial seed value of the average glucose concentration may be multiplied by the average sensitivity to obtain an initial number of counts. Thereafter, S t = alpha filter t α*S t-1 +(1-α)S TX t Sensors t

[0255] After a period of time, e.g., one day, enough data may be received so that it can be corrected to an actual average value, which can be used to determine drift. The above steps may then be repeated, determining the subsequent best combination of slope, baseline, and glucose in the above manner that results in the best raw counts to match the estimate of raw counts determined from the day 1 data, for example.

[0256] The distribution may in some implementations be modified according to measured data as more is acquired. In this way, a better calibration may be obtained. At the start, only the assumed distribution was used. Afterwards, actual measured data is available and can be used. The filter may be reseeded with the median or other representative sensor counts, and the distribution will generally converge to the actual measured data. Fingerstick data is available and can be used for even faster convergence.

[0257] Seeding or reseeding may be performed in many ways and adjusted to allow for faster convergence of the drift curve based on the filter seed parameters. For example, the initial seed value for the filter may be based on a predicted signal level estimated from a predicted average glucose level and sensor sensitivity. The initial seed value for glucose may also be based on subject demographics such as the user's age and the duration of their diabetes. The initial seed value may also be based on data such as fasting glucose levels, glycated hemoglobin (A1C) tests, current diabetes treatment (e.g., oral medication, basal insulin use, or intensive insulin therapy), or medical record information or laboratory tests such as downloaded self-monitored blood glucose levels.

[0258] In another implementation, an initial seed value may be used to first start a filter running in the forward direction. After a representative set of sensor readings has been collected, e.g., 24 hours after the sensor readings, a second filter may then run in the reverse direction. If a representative set of sensor readings is available, the forward or reverse filter may be seeded with a typical signal value, such as the median sensor reading, or the filter may be seeded with a typical signal value adjusted for predicted drift, such as starting the forward filter with 0.9*median and the reverse filter with 1.1*median. These techniques have the advantage that when more than one filter, such as a forward filter and a reverse filter, are used, their seed values ​​may then be further optimized to minimize the difference between the two filter outputs. For example, one exemplary method is to minimize the mean square error between the two filter outputs.

[0259] In the system and method according to the present principles described above, redefining the seed value daily helps minimize the mean absolute error in the signal domain. In one implementation, using an approximate estimate of the signal trajectory from day 1 each day, for example, a smooth trend of the noisy filter output can create a trajectory for the drift on day 2. Drift rate estimates can be compared from two or more different methods, and the difference or error between the two can be fed into an algorithm, such as a signal-based calibration algorithm as disclosed in the patent application (Serial No. 13 / 827,119) incorporated by reference above, that determines the distribution of sensitivity, particularly with respect to the certainty interval. Also in this manner, signal features can be extracted, including features corresponding to noise, level, drift, model, power, energy, etc. In this manner, it can be determined whether the drift correction is on track. For example, if the drift slope differs significantly from what the factory calibration model suggests, for example, by more than a certain threshold percentage, the user can be prompted for feedback or to provide a fingerstick calibration.

[0260] In some implementations, a smoothing filter may be used to compensate for signal drift in real time. In one case, a double exponential smoothing filter is used. Such a filter may estimate a multiplicative decay trend without seasonality, but may estimate additive or multiplicative seasonality to improve performance. The double exponential filter acts to recover the underlying changes, i.e., drift, of the sensor signal as a function of time. There are three main underlying equations that govern the double exponential smoothing filter:

[0261]

number

[0262] A table of the parameters in the equations can be seen in Table I below.

[0263] [Table 1]

[0264] In the above equations and in this setting, α and β can be considered small. α is small because it is desirable not to have the filter affected by glucose excursions. β is small because the underlying trends being recovered are slow moving in nature. The following parameters were used to generate one set of exemplary results (Table II (estimating a 5 minute sampling rate)):

[0265] [Table 2]

[0266] In the above equation, the slope may be the initial sensitivity calculated by the algorithm, or a value determined from another method or from the method for determining the sensitivity value described above. The average glucose in the above equation may be an average of historical glucose values ​​from previous sessions, or based on, for example, A1C values ​​reported by the user. In one implementation, data was generated using the initial sensitivity estimated by the algorithm using a 2-hour calibration and self-reported A1C numbers from a cohort of test subjects.

[0267] The drift correction curve was estimated using the following equation:

[0268]

number

[0269] The sensor signal was then drift corrected using the following equation:

[0270]

number

[0271] Glucose values ​​were then calculated using the following equation:

[0272]

number

[0273] In the above equation, the slope is that estimated by the algorithm at the initial calibration, and the baseline is the slope multiplied by 1 mg / dL.

[0274] To illustrate the effectiveness of the double exponential filter, Figures 34 and 35 show example glucose traces. Figure 34 shows a CGM trace 364 along with reference and calibration values. Figure 35 shows the raw sensor signal 366 along with the output 367 from the double exponential filter. In this case, the sensor was calibrated once using two activation values ​​entered by the user.

[0275] The estimated drift curve for the above sensor is seen by curve 368 in Figure 36. As can be seen, the drift curve is easily seen in both the upward and downward directions, and knowledge of the drift detected and quantified by the double exponential filter allows for correction of the drift. One advantage of this implementation is that there is no need to rely on any known curve shape to correct for the drift.

[0276] 37-39 are further charts illustrating drift correction according to the principles described above.

[0277] Besides the use of double exponential filters, other filters may also be used. For example, a Kalman filter may be used, which includes process noise, also known as model noise, and measurement noise estimates. Gaussian filters may also be used, as well as traditional Butterworth low-pass filters, moving median filters, moving average filters, etc., as long as the filters help to reveal the underlying trends. A filter bank or series of filters may be used that combines multiple filters to obtain an average or combined trend of the underlying signal. When multiple filters are used, different types of filters may be used within a single bank, and filter settings may vary between filters. The use of multiple filters may improve signal correction at the end of subsequent periods, for example, at the end of a second day's worth of data, at the end of a third day's worth of data, etc. Without being bound by theory, it is presumed that in the use of these filters, the day-to-day variability of the average glucose measurements is not significant compared to the change in slope or underlying signal change.

[0278] In other variations, the signal may be pre-processed prior to drift estimation filtering to remove data gaps and outliers. Additionally, the calibration may be automatically updated in a manner that reduces the occurrence of unexpected spikes in CGM readings. Such methods include making changes when the signal is stable, e.g., changing the calibration settings in the middle of the night, or slowly blending the calibration changes into the current settings over a period of time, e.g., an hour or more.

[0279] Other useful techniques that can be used with filtering include various decomposition techniques. For example, empirical model decomposition can be used to split the signal into a series of intrinsic mode functions over time. Other time and frequency based decompositions can also be used, including Fourier and wavelet transforms.

[0280] In another variation, other signal-based parameters can be determined and used in the calibration. For example, referring to FIG. 40, it can be seen that the coefficient of variation (CV) of the sensor signal has a strong correlation with the glucose variation of the signal, e.g., the glucose standard deviation. Thus, in determining the calibration parameters, this correlation can be used to select calibration parameters that satisfy the error model of signal CV-glucose standard deviation.

[0281] More specifically, a priori information as described may be used in a factory calibration, including information obtained prior to a particular calibration. For example, such information includes information from a previous calibration, such as information obtained prior to sensor insertion. Calibration information includes information useful for calibrating a continuous glucose sensor, such as, but not limited to, sensitivity (m), change in sensitivity (dm / dt), and other signal and time derivative aspects, as described above. Importantly, a priori information also includes distribution information, such as ranges, distribution functions, and distribution parameters, including mean and standard deviation.

[0282] Specifically, with regard to the standard deviation of the distribution of glucose values, it may be advantageously used, for example, in determining the limits of likely glucose values, as well as possible combinations of sensitivity and baseline. The standard deviation may also be used in determining the level of confidence from the a priori calibration distribution information, such as when it is a feedback of a posteriori calibration distribution information from a previous calibration (where the level of tightness or looseness of the distribution is quantified by the standard deviation). In the same way, the level of confidence may be determined, for example, from the a posteriori calibration distribution information, based on the level of tightness or looseness of the distribution, which may again be quantified by the standard deviation.

[0283] As a specific example, Figure 41 shows the glucose signal over a 10 day period. From this, the signal standard deviation and the signal mean may be calculated. The signal coefficient of variation may then be determined by:

[0284] Signal CV = signal standard deviation / signal mean

[0285] In the case of FIG. 41, the signal CV was calculated to be 0.4230.

[0286] If a relationship has been determined between the signal CV and the glucose standard deviation (see, e.g., the lines in FIG. 42), then the calculated signal CV determined above may be used to determine a predicted glucose standard deviation, which may then be used in the above calculations as well as other calculations. For example, an error model may be established using this correlation, and the error model used in the calibration in U.S. patent application Ser. No. 13 / 827,119, incorporated by reference above.

[0287] FIG. 43 shows an exemplary distribution of the difference between the measured and predicted standard deviations.

[0288] Other relationships may also be used. For example, referring to FIG. 44, a relationship between average glucose value and glucose standard deviation is seen. Specifically, it can be seen that users with higher standard deviations tend to have higher average glucose. This relationship may be used in determining, setting, inferring, or otherwise selecting the calibration value. As another example, and referring to FIG. 45, another useful indicator is the type of patient. Specifically, FIG. 45 shows the clear difference in glucose standard deviation between non-diabetic patients and type I and type II diabetic patients. Accordingly, based on the type of patient, the model selected for the patient population may be modified based on the type of population, for example, the standard deviation may be strengthened with respect to the CV error model, the mean value may be shifted, etc.

[0289] As mentioned above, the system may adjust the data for time lag from the data over a previous period of time (e.g., to remove time lag induced by real-time filtering) and display a graphical representation of the lag-adjusted data over that period (e.g., a trend graph). The system may also compensate for time lag. For example, FIG. 46 shows data points separated by a time lag, where Δ represents the individual rate of change between two adjacent points. By use of the following or similar equations, such time lag may be compensated for:

[0290] Glucose(t) = [drift corrected sensor(t) + 5*ROC(t)] / mb(mg / dL)

[0291] For example, if the current point is in light noise, all filtered sensor counts may be used. EXAMPLES

[0292] An exemplary calibration routine is now described, the steps of which can be seen in flow chart 400 of FIG. 47. In a first step, the sensitivity profile versus time is characterized using a bench test that measures the response of the sensor in glucose solution over a period of one or more days (step 402). As this is typically a destructive test, it may be performed on a representative set of sensors from a production lot or production line. This test may be repeated periodically or when the process changes, e.g., when there are changes in raw materials or equipment. The sensitivity of each sensor may then be measured in a non-destructive bench test (step 404). The results of steps 402 and 404 are used to estimate the in vivo initial and final sensitivity of each sensor (step 406).

[0293] Note that the destructive testing measures long term drift for a group of sensors, determining, for example, that a sensor drifts 10% over the first two days. The non-destructive bench testing provides a starting point for each sensor in logarithms, for example, that a given sensor might have a starting sensitivity of 20. Combining the two tests, for example, it might be determined that a given sensor starts at 20 and is expected to drift 10%, for example, to 22.

[0294] This step therefore maps or converts bench values ​​to in vivo values ​​using a function trained or optimized on well-characterized in vivo data, such as clinical trials. As another example, a subject's sensor may start at 24 and drift to 26 in vivo.

[0295] The electrical characteristics of the transmitter (such as gain and offset) are calibrated during manufacture (step 408) and these calibration coefficients are stored on the transmitter (step 412) so that the raw sensor signal can be corrected for part-to-part differences in the electronics before running the algorithm.

[0296] The sensor is then packaged with a single-use transmitter (step 414). The sensor's identifier, e.g., an identification number, is read by optical bar code and its estimated in vivo sensitivity is retrieved from a manufacturing database and written to the transmitter, e.g., using wireless (NFC or Bluetooth®) communication (step 416).

[0297] When the transmitter first detects that it is connected to a functioning sensor, a session timer is started and the algorithm begins (step 418). The algorithm begins converting the sensor signal to glucose using the CGM model (step 420), with the model parameters set to the previous information about the sensor sensitivity recorded in step 416.

[0298] When a representative set of signal data, e.g., 24 hours, is available, seed parameters for the forward and reverse filters may be set using the signal median and the estimated amount of drift (step 422). These seed values ​​are then further optimized to minimize the mean squared error between the two signal filters (step 424).

[0299] The signal-based calibration algorithm uses the average values ​​of the forward and backward filtered signals and the raw sensor signal, and then adjusts the model parameters, e.g., sensitivity and baseline, to meet some criteria (step 426). Time-based input data is then used to update the algorithm. An exemplary signal-based calibration algorithm is that disclosed in the above-incorporated patent application (Serial No. 13 / 827,119), specifically in

[0188] , Example 4 showing a Bayesian learning approach for drift estimation and correction.

[0300] In the implementation of Figure 47, the model is adjusted to meet the criterion that the mean glucose values ​​match the predicted diabetes mean, and may be further adjusted to meet another criterion that the CGM glucose variability matches the mean glucose levels. To calculate these metrics, the algorithm has an estimated relationship between mean glucose and glucose variability. For example, a non-diabetic patient may have a mean glucose level of 85 mg / dL and a standard deviation of 15 mg / dL. A diabetic patient may have a mean glucose of 170 mg / dL and a standard deviation of 65 mg / dL.

[0301] The optimized model parameters are then used by the CGM model to convert subsequent sensor readings into glucose values ​​that are then displayed (step 428).

[0302] Steps 424-428 are repeated when a new set of signal data is available.

[0303] Similar methods may be used to detect unacceptable amounts of sensor change (typically loss of sensitivity from day 7 onwards) and to stop displaying potentially inaccurate readings.

[0304] In one preferred embodiment, the analyte sensor is an implantable glucose sensor, as described with reference to U.S. Patent Publication No. 6,001,067 and U.S. Patent Publication No. US-2005 / 0027463-A1. In another preferred embodiment, the analyte sensor is a transcutaneous glucose sensor, as described with reference to U.S. Patent Publication No. US-2006 / 0020187-A1. In yet other embodiments, the sensor is configured to be implanted within or outside a host's blood vessel, as described in U.S. Patent Publication No. US-2007 / 0027385-A1, co-pending U.S. patent application Ser. No. 11 / 543,396, filed Oct. 4, 2006, co-pending U.S. patent application Ser. No. 11 / 691,426, filed Mar. 26, 2007, and co-pending U.S. patent application Ser. No. 11 / 675,063, filed Feb. 14, 2007. In one alternative embodiment, the continuous glucose sensor includes a transcutaneous sensor, such as described in U.S. Patent No. 6,565,509 to Say et al. In another alternative embodiment, the continuous glucose sensor includes a subcutaneous sensor, such as described in U.S. Patent No. 6,579,690 to Bonnecaze et al. and U.S. Patent No. 6,484,046 to Say et al. In another alternative embodiment, the continuous glucose sensor includes a pluggable subcutaneous sensor, such as described in U.S. Patent No. 6,512,939 to Colvin et al. In another alternative embodiment, the continuous glucose sensor includes an intravascular sensor, such as described in U.S. Patent No. 6,477,395 to Schulman et al. In another alternative embodiment, the continuous glucose sensor includes an intravascular sensor, such as described in U.S. Patent No. 6,424,847 to Mastrototaro et al.

[0305] The connections between elements shown in the drawings represent exemplary communication paths. Additional communication paths, either direct or through intermediaries, may be included to further facilitate the exchange of information between the elements. The communication paths may be bidirectional communication paths that enable the elements to exchange information.

[0306] The various operations of the methods described above may be performed by any suitable means capable of performing the operations, such as various hardware and / or software component(s), circuits, and / or module(s). Generally, any operation illustrated in the figures may be performed by a corresponding functional means capable of performing the operation.

[0307] The various example logic blocks, modules, and circuits described in connection with this disclosure (such as the blocks of FIGS. 2 and 4) may be implemented or performed using a general purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array signal (FPGA) or other programmable logic device (PLD), discrete gate or transistor logic, discrete hardware components, or any combination thereof, designed to perform the functions described herein. A general purpose processor may be a microprocessor, but in the alternative, the processor may be a commercially available processor, controller, microcontroller, or state machine. A processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration.

[0308] In one or more aspects, the functions described may be implemented in hardware, software, firmware, or any combination thereof. If implemented in software, the functions may be stored or transmitted as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media, including any medium that facilitates transfer of a computer program from one place to another. Storage media may be any available medium that can be accessed by a computer. By way of example, and not limitation, such computer-readable media may include various types of RAM, ROM, CD-ROM, or other optical disk storage, magnetic disk storage, or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer. Also, any connection is properly termed a computer-readable medium. For example, if the software is transmitted from a website, server, or other remote source using coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave, the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included within the definition of the medium. Disk and disc, as used herein, include compact discs (CDs), laser discs, optical discs, digital versatile discs (DVDs), floppy disks, and Blu-ray® discs, where disks typically reproduce data magnetically while discs reproduce data optically with a laser. Thus, in some aspects, computer readable media may include non-transitory computer readable media (e.g., tangible media). Additionally, in some aspects, computer readable media may include transitory computer readable media (e.g., signals). Combinations of the above should also be included within the scope of computer readable media.

[0309] The methods disclosed herein include one or more steps or actions for achieving the described method. The method steps and / or actions may be interchanged with one another without departing from the scope of the claims. In other words, unless a specific order of steps or actions is specified, the order and use of specific steps and / or actions may be modified without departing from the scope of the claims.

[0310] Certain aspects may include a computer program product for performing the operations presented herein. For example, such a computer program product may include a computer-readable medium having instructions stored (and / or encoded) thereon, the instructions being executable by one or more processors to perform the operations described herein. For certain aspects, the computer program product may include packaging materials.

[0311] Software or instructions may also be transmitted over a transmission medium. For example, if the software is transmitted from a website, server, or other remote source using coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave, the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included within the definition of transmission media.

[0312] Further, it should be understood that modules and / or other suitable means for implementing the methods and techniques described herein may be downloaded and / or otherwise obtained by a user terminal and / or base station, where applicable. For example, such devices may be coupled to a server to facilitate the transfer of means for implementing the methods described herein. Alternatively, the various methods described herein may be provided via a storage means (e.g., RAM, ROM, a physical storage medium such as a compact disk (CD) or floppy disk, etc.), such that the user terminal and / or base station may obtain the various methods upon coupling or providing the storage means to the device. Furthermore, any other suitable technique for providing the methods and techniques described herein to a device may be utilized.

[0313] It is to be understood that the claims are not limited to the precise configuration and components illustrated above. Various modifications, changes and variations can be made in the arrangement, operation and details of the methods and apparatus described above without departing from the scope of the claims.

[0314] Unless otherwise defined, all terms (including technical and scientific terms) shall have their ordinary and customary meanings as indicated to those skilled in the art, and shall not be limited to any special or customized meanings unless expressly defined as such herein. It should be noted that the use of a particular term when describing a particular feature or embodiment of the present disclosure should not be interpreted as implying that the term is redefined herein to include any particular characteristic of the feature or embodiment of the present disclosure to which it pertains. Particularly in the appended claims, the terms and phrases used in this application, and variations thereof, should be interpreted as open-ended as opposed to limiting, unless expressly stated otherwise. As examples of the above, the term "including" should be interpreted to mean "including without limitation," "including but not limited to," etc.; the term "comprising," as used herein, is synonymous with "including," "containing," or "featuring," is inclusive or open-ended, and does not exclude additional, unrecited elements or method steps; the term "having" should be interpreted as "having at least," the term "including" should be interpreted as "including but not limited to," the term "example" is used to provide illustrative examples of the items under discussion rather than an exhaustive or exclusive list of them, and should not be construed as being "known," "conventional," "ordinary," or "conventional." Adjectives such as, "standard," and terms of similar import should not be construed to limit the subject matter described to that available in a given period or at a given time, but rather should be construed to encompass known, conventional, or standard technology that may be available or known at any time now or in the future, and the use of terms such as "preferably," "preferred," "desired," or "desirable" and terms of similar import should not be understood to imply that a particular feature is critical, essential, or even important to the structure or function of the invention, but rather should be intended merely to highlight alternative or additional features that may or may not be utilized in a particular embodiment of the invention.Similarly, groups of items joined by the conjunction "and" should not be construed as requiring the presence of every single one of those items in the group, but rather, unless otherwise stated, should be construed as "and / or." Similarly, groups of items joined by the conjunction "or" should not be construed as requiring mutual exclusivity between the groups, but rather, unless otherwise stated, should be construed as "and / or."

[0315] When a range of values ​​is provided, it is understood that the upper and lower limits, and each intervening value between the upper and lower limits of that range, are included within an embodiment.

[0316] With respect to virtually any plural and / or singular term herein, one skilled in the art can convert from plural to singular and / or from singular to plural as appropriate to the context and / or application. Various singular / plural permutations may be expressly set forth herein for clarity. The indefinite articles "a" or "an" do not exclude plurals. A single processor or other unit may accomplish the functions of several items recited in the claims. The mere fact that certain criteria are recited in mutually distinct independent claims does not indicate that a combination of these criteria cannot be used to advantage. Reference signs in the claims should not be construed as limiting the scope.

[0317] It will be further understood by those skilled in the art that, when a specific number is intended in an introduced claim recitation, such intent shall be clearly stated in the claim, and that in the absence of such a statement, no such intent exists. For example, to aid in understanding, the following appended claims may include the use of the introductory phrases "at least one" and "one or more" to introduce the claim recitation. However, the use of such phrases should not be construed as implying that the introduction of a claim recitation with the indefinite article "a" or "an" limits a particular claim that includes such an introduced claim recitation to an embodiment that includes only one such recitation, even when the same claim includes the introductory phrases "one or more" or "at least one" and an indefinite article such as "a" or "an" (e.g., "a" and / or "an" should typically be construed to mean "at least one" or "one or more"), which also applies to the use of definite articles used to introduce claim recitations. In addition, even if a specific number of recitations of an introduced claim is explicitly recited, a person skilled in the art will recognize that such recitation should typically be interpreted to mean at least the recited number (e.g., a mere recitation of "two recitations" without other modifiers typically means at least two recitations, or more than two recitations). Furthermore, when a conventional expression similar to "at least one of A, B, and C, etc." is used, such a structure is generally intended to include any combination of the recited items, including, for example, a single member, in the sense that a person skilled in the art would understand the conventional expression (e.g., "a system having at least one of A, B, and C" includes, but is not limited to, a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, and C, etc.).When a conventional expression similar to "at least one of A, B, or C, etc." is used, such a structure is generally intended in the sense that one of ordinary skill in the art would understand the conventional expression (e.g., "a system having at least one of A, B, or C" includes, but is not limited to, a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, and C, etc.). It will be further understood by those of ordinary skill in the art that virtually any disjunction and / or phrase indicating two or more alternative terms, whether in the description, claims, or drawings, should be understood to contemplate the possibility of including one of the terms, either of the terms, or both terms. For example, the phrase "A or B" is understood to include the possibilities of "A" or "B" or "A and B."

[0318] It is to be understood that all numbers expressing quantities of ingredients, reaction conditions, and the like used herein are in all instances modified by the term "about." Accordingly, unless indicated to the contrary, the numerical parameters set forth herein are approximations that may vary depending upon the desired properties sought to be obtained. At the very least, and not as an attempt to limit the application of the doctrine of equivalents to the scope of any claims in any application claiming priority to this application, each numerical parameter should be construed in light of the number of significant digits and ordinary rounding approaches.

[0319] All references cited herein are incorporated herein by reference in their entirety. To the extent that publications and patents or patent applications incorporated by reference conflict with the disclosure contained herein, the present specification is intended to supersede and / or take precedence over any such conflicting material.

[0320] Headings are included herein for reference and to aid in locating the various sections. These headings are not intended to limit the scope of the concepts described therein. Such concepts may have applicability throughout the entire specification.

[0321] Moreover, while the foregoing has been described in some detail by way of illustration and example for purposes of clarity and understanding, it will be apparent to those skilled in the art that certain changes and modifications may be practiced. Thus, the description and examples should not be construed as limiting the scope of the invention to the specific embodiments and examples described herein, but rather as embracing all modifications and alternatives that fall within the true scope and spirit of the invention. [Explanation of symbols]

[0322] 2. Drug delivery pumps 4 Glucose Meter 8. Continuous Analyte Sensor Systems 10 Continuous Analyte Sensors 12 Sensor Electronics 14, 16, 18, 20 Display Devices 100 Systems 406 Network 490 Cloud-Based Analyte Processors

Claims

1. 1. A method of calibrating an analyte concentration sensor in a biological system using only a signal from the analyte concentration sensor, wherein at steady state, an analyte concentration value in the biological system is known, the method comprising: a. detecting, on a monitoring device, when an analyte concentration value measured by an analyte concentration sensor placed within the biological system constitutes a first repeatable event; b. correlating, on the monitoring device or on a device or server operatively coupled to the monitoring device, the measured analyte concentration value when the biological system is in the detected first repeatable event with the known analyte concentration value.

2. The method of claim 1 , wherein the correlating comprises determining a functional relationship between the sensor readings and the known analyte concentration values.

3. The method of claim 2 , wherein the functional relationship includes a multiplicative constant.

4. The method of any one of claims 1 to 3, wherein the detection comprises waiting a predetermined time following input of an event to the monitoring device.

5. The method of claim 4 , wherein the event is diet or exercise.

6. a. detecting an occurrence of a second recurring event subsequent to said correlation, said second recurring event being different from said first recurring event; 6. The method of claim 1, further comprising: recalibrating the analyte concentration sensor by correlating a sensor reading when the biological system is in the second detected repeatable event with the known analyte concentration value.

7. The method of claim 6 , wherein the sensor readings have a first raw value at an initial calibration and a second raw value at a recalibration, the first and second raw values ​​being different.

8. 7. The method of claim 6, wherein the recalibration is configured to occur when sensor readings are substantially stable or within a predetermined range of readings for a threshold period of time, thereby reducing the occurrence of unexpected spikes in readings.

9. a. subsequent to said correlation, displaying a graph or table showing currently measured and historical values ​​of said analyte concentration calibrated at least in part based on said correlation; 8. The method of claim 7, further comprising: subsequent to said recalibration, updating said display of said graph or table showing currently measured and historical values ​​of said analyte concentration according to said recalibration.

10. The method of claim 9 , wherein the updating changes the representation of the historical value of the analyte concentration.

11. a. determining a difference between the first raw value and the second raw value; 8. The method of claim 7, further comprising: comparing the difference-based quantity to a predetermined standard; and determining whether the sensor calibration has drifted based on the comparison.

12. The method of claim 11 , further comprising determining a quantitative amount that the sensor calibration has drifted.

13. The method of claim 12 , further comprising adjusting the sensor calibration based on the determined quantitative amount.

14. The method of claim 11 , wherein the quantity is a gradient between the first raw value and the second raw value.

15. The method of claim 14 , wherein if the gradient exceeds a predetermined threshold, future calibrations are inhibited until the gradient no longer exceeds a predetermined threshold.

16. The method of claim 15 further comprising prompting a user to input a measurement.

17. The method according to any one of claims 1 to 11, wherein the sensor is a glucose sensor.

18. a. receiving a signal from the sensor following said correlation; 20. The method of claim 1, further comprising: displaying a value corresponding to the received signal, the displayed value being based on the received signal and the known analyte concentration value.

19. The method of any one of claims 1 to 18, further comprising determining the known analyte concentration value by prompting a user to input a measurement value.

20. The method of any one of claims 1 to 19, further comprising determining the known analyte concentration value by accessing a population mean.

21. 21. The method of any one of claims 1 to 20, wherein the recurring event is selected from the group consisting of steady state, post-prandial rise, daily high and low glucose range, decay rate, or rate of change.

22. 1. A method for compensating for drift in an analyte concentration sensor in a biological system using only a signal from the analyte concentration sensor, comprising: a. measuring an analyte level using an indwelling analyte concentration sensor; b. determining a first slow moving average of the analyte measurement over a first time period and referencing a calibration of the sensor based at least in part on the first slow moving average; c. determining a second slow moving average of the measurement of the analyte over a second period of time subsequent to the first determination; d. adjusting the calibration of the sensor based at least in part on a difference between the first slow moving average and the second slow moving average.

23. 23. The method of claim 22, wherein the first period of time is greater than about 12 hours in duration.

24. 24. The method of claim 23, wherein the first period of time is greater than about 24 hours in duration.

25. The method of any one of claims 22 to 24, wherein the duration of the first period of time is the same as the duration of the second period of time.

26. a. subsequent to said referencing said calibration of said sensor based at least in part on said first slow moving average, displaying a graph or table showing at least historical values ​​of analyte concentration calibrated at least in part on said first slow moving average; The method of any one of claims 22-25, further comprising: subsequent to said adjusting, updating said display of said graph or table showing at least historical values ​​of said analyte concentration according to said adjusted calibration.

27. 27. The method of claim 26, wherein the updating changes the indication of the historical value of the analyte concentration.

28. 27. The method of claim 26, wherein the displayed graph or table further indicates a currently measured value of the analyte concentration.

29. The method of any one of claims 22 to 27, wherein the referencing a calibration of the sensor based at least in part on the first slow moving average further comprises basing the calibration on a seed value.

30. 30. The method of claim 29, wherein the seed value is a value received from a population average or from a previous session.

31. 31. The method of claim 30, further comprising, subsequent to said adjusting, varying said seed value based at least in part on said adjusting.

32. 32. The method of claim 31, further comprising varying the seed value based on a difference between the first slow moving average and the second slow moving average.

33. 33. The method of any one of claims 22 to 32, wherein the adjustment of the calibration is configured to occur when sensor readings are substantially stable or within a predetermined range of readings for a threshold period of time, thereby reducing the occurrence of unexpected spikes in readings.

34. 1. A method for compensating for drift of an analyte concentration sensor in a biological system using only a signal from the analyte concentration sensor, comprising: a. measuring an analyte level using an indwelling analyte concentration sensor; b. determining a first slow moving average of the analyte measurements over a first set of time periods, the first set including event-based time periods, and referencing a calibration of the sensor based at least in part on the first slow moving average; c. determining a second slow moving average of the measurement of the analyte over a second set of time periods subsequent to the first determination, the second set including event-based time periods; d. adjusting the calibration of the sensor based at least in part on a difference between the first slow moving average and the second slow moving average.

35. 35. The method of claim 34, wherein the first and second event-based periods are selected from the group consisting of a post-meal period, a sleep period, and a post-breakfast period.

36. 35. The method of claim 34, wherein the adjustment of the calibration is configured to occur when sensor readings are substantially stable or within a predetermined range of readings for a threshold period of time, thereby reducing the occurrence of unexpected spikes in readings.

37. 1. A method for compensating for drift of an analyte concentration sensor in a biological system using only a signal from the analyte concentration sensor, comprising: a. measuring an analyte level using an indwelling analyte concentration sensor; b. determining a first slow moving average of the analyte measurements over a first time period and referencing a first apparent sensitivity of the sensor based at least in part on the first slow moving average; c. subsequent to the first determination, determining a second slow moving average of the measurements of the analyte over a second time period and referencing a second apparent sensitivity of the sensor based at least in part on the second slow moving average; d. determining whether a change in apparent sensitivity of the sensor between the first apparent sensitivity and the second apparent sensitivity meets a predetermined criterion; e. if the change in apparent sensitivity meets a predetermined criterion, adjusting the actual sensitivity of the sensor to an adjusted value based on the difference between the first apparent sensitivity and the second apparent sensitivity; f. if the change in apparent sensitivity does not match a predetermined criterion, prompting a user to enter data so that a reason for the change in apparent sensitivity may be determined.

38. 38. The method of claim 37, wherein the determining comprises determining whether the change in apparent sensitivity is due to sensitivity drift or a change in the slow moving average.

39. 40. The method of claim 38, further comprising, if the change in apparent sensitivity is due to a change in the slow moving average, prompting the user to enter data relating to the change.

40. 40. The method of claim 37, wherein the predetermined criteria comprises known behavior regarding sensitivity changes over time for the sensor.

41. 41. The method of claim 40, wherein the known behavior regarding sensitivity change constitutes an envelope of acceptable sensitivity change over time.

42. 38. The method of claim 37, wherein the adjusted value is based at least in part on the second slow moving average.

43. 41. The method of claim 40, wherein the predetermined criteria further comprises known values ​​for physiologically likely analyte changes.

44. 38. The method of claim 37, wherein the prompting of the user includes prompting the user to input a calibration value.

45. 40. The method of claim 37, wherein the prompting of the user includes prompting the user to input diet or exercise information.

46. 46. ​​The method of claim 44 or 45, wherein upon receiving the calibration value or the meal or exercise information from the user, the method further comprises determining whether the change in apparent sensitivity is due to sensitivity drift or a change in the slow moving average.

47. 47. The method of claim 46, further comprising adjusting the actual sensitivity of the sensor based on the received calibration value or diet or exercise information.

48. 48. The method of any one of claims 37 to 47, wherein the adjustment of the actual sensitivity of the sensor is configured to occur when sensor readings are substantially stable or within a predetermined range of readings for a threshold period of time, thereby reducing the occurrence of unexpected spikes in readings.

49. 1. A method for checking the calibration of an analyte concentration sensor system in a biological system using only a signal from the analyte concentration sensor, comprising: a. measuring analyte values ​​over time using an indwelling analyte concentration sensor after an initial calibration; b. calculating a clinical value of an analyte concentration based on the measurement and the initial calibration; c. adjusting an initial calibration to an updated calibration, the updated calibration being based on the measurements of the analyte over time or only a subset thereof; and d. calculating a clinical value of said analyte concentration based on the measurements and said updated calibration.

50. 50. The method of claim 49, wherein the initial calibration is based on a population average or on data entered by a user.

51. 51. The method of claim 49 or 50, wherein the adjustment is based on a slow moving average of the measurement of the analyte over time.

52. 52. The method of any one of claims 49 to 51, wherein the adjustment is based on a steady state value of the analyte.

53. 53. The method of any one of claims 49 to 52, wherein the initial calibration is based on data determined prior to a session associated with the indwelling analyte concentration sensor.

54. 54. The method of claim 53, wherein the data is determined a priori, on the bench, or in vitro.

55. 55. A method according to any one of claims 49 to 54, wherein the adjustment is configured to occur when the sensor readings are substantially stable or within a predetermined range of readings for a threshold period of time, thereby reducing the occurrence of unexpected spikes in readings.

56. 1. A method of calibrating an analyte concentration sensor in a biological system using a signal from the analyte concentration sensor, wherein at steady state, an analyte concentration value in the biological system is known, the method comprising: a. receiving, on a monitoring device, seed values ​​for calibration parameters; b. detecting, on said monitoring device, when an analyte concentration value measured by an analyte concentration sensor placed within the biological system is at steady state; c. correlating, on the monitoring device or on a device or server operatively coupled to the monitoring device, the measured analyte concentration value when the biological system is at the detected steady state with the known analyte concentration values; d. receiving a signal from said sensor following said correlation; e. calculating and displaying a value corresponding to the received signal, the calculated value being based on the received signal, the known analyte concentration value, and the seed value.

57. 57. The method of claim 56, wherein the received seed value is received from a source that includes factory calibration information.

58. a. detecting behavior in the received signal outside of predefined parameters; 58. The method of claim 56 or 57, further comprising: b. prompting a user to input external calibration information.

59. 60. The method of claim 58, wherein the displayed value is further based on the external calibration information.

60. 59. The method of claim 58, wherein the external calibration information is received from an SMBG or a fingerstick calibration.

61. The method of any one of claims 58 to 60, further comprising resetting the known calibration value to a new known calibration value, the resetting being based at least in part on the external calibration information.

62. The method of any one of claims 58 to 60, further comprising resetting the seed value to a new seed value, the resetting being based at least in part on the external calibration information.

63. 62. The method of any one of claims 56 to 61, further comprising modifying the display based on the determined accuracy of the value.

64. 64. The method of claim 63, wherein the modification of the display includes displaying ranges rather than values, or vice versa.

65. A method according to any one of claims 56 to 64, wherein the received seed values ​​for calibration parameters are user entered characteristics of a medical condition.

66. 66. The method of claim 65, wherein the user-entered characteristics of a medical condition include an indication of Type I diabetes, Type II diabetes, non-diabetes, or pre-diabetes.

67. A method according to any one of claims 56 to 66, wherein the received seed values ​​for calibration parameters are values ​​based on one or more user entered blood glucose values.

68. and displaying the values ​​corresponding to the received signals includes displaying a graph or table showing currently measured and historical values ​​of the analyte concentration. a. Detecting that a change in calibration has occurred; b. adjusting one or more calibration parameters of the analyte concentration sensor according to the change in calibration; c) following said adjustment, updating said display of said graph or table showing currently measured and historical values ​​of said analyte concentration according to said adjusted calibration parameters.

69. The detection that a change in calibration has occurred comprises: a. Detecting changes in the slow moving average, or 69. The method of claim 68, comprising: b) detecting a change in said steady state value.

70. 70. A method according to any one of claims 56 to 69, wherein the adjustment is configured to occur when sensor readings are substantially stable or within a predetermined range of readings for a threshold period of time, thereby reducing the occurrence of unexpected spikes in readings.

71. 1. A method of calibrating an analyte concentration sensor, following insertion of the sensor into a patient, using only parameters based on or derivable from the sensor signal; a. receiving at least an initial value of an analyte concentration and an initial value or a distribution of initial values ​​of a sensor sensitivity; b. monitoring a signal from the analyte concentration sensor for a duration following insertion of the sensor; c. calculating a plurality of analyte concentration values ​​based on the initial values ​​or value distributions of the monitored sensor signal and the sensor sensitivity over the duration; d. determining a distribution of values ​​of the monitored signal over said time duration; e. optimizing the initial value or the value distribution of the sensor sensitivity and the plurality of analyte concentration values ​​to match the value distribution of the monitored signal; f) determining updated sensitivities based on said optimization.

72. 72. The method of claim 71, wherein the receiving comprises receiving an initial value distribution of sensor sensitivity, and the calculation of a plurality of analyte concentration values ​​is based on the monitored sensor signal and representative values ​​from the initial value distribution of the sensor sensitivity.

73. 73. The method of claim 72, wherein the representative value is selected from an average or a midpoint or a median.

74. Said determination of updated sensitivity may comprise: a. dividing the representative value by the initial value of the analyte concentration; 74. The method of claim 73, further comprising: b. updating said sensitivity value to be equal to the result of said division.

75. 75. The method of any one of claims 71 to 74, wherein the initial value of an analyte population is a population mean.

76. The method of any one of claims 71 to 75, wherein the initial values ​​of the analyte populations are input by a user.

77. The method of any one of claims 71 to 76, wherein the initial values ​​of the analyte populations are transferred from a previous session.

78. 78. The method of any one of claims 71 to 77, wherein the optimization comprises optimizing the product of the initial value or value distribution of the sensor sensitivity and the plurality of analyte concentration values.

79. 80. The method of claim 78, wherein the optimizing of the product comprises optimizing the product to match the value distribution of the monitored signal while adjusting parameters of the value distribution of the sensor sensitivity and a plurality of analyte concentration values ​​to most closely match corresponding population means.

80. 80. The method of any one of claims 63-79, wherein the receiving further comprises receiving an initial value distribution of a baseline, and the optimizing further comprises optimizing the value distribution of the baseline, along with the value distribution of the sensor sensitivity and the plurality of analyte concentration values, to match the value distribution of the monitored signal.

81. 81. The method of claim 80, wherein the initial value distribution of the baseline follows a normal distribution.

82. 82. The method of any one of claims 63 to 81, wherein at least the initial value of the analyte concentration is used as part of a seed input to a slow moving average filter.

83. The method according to any one of claims 71 to 82, wherein the initial value distribution of sensor sensitivity is defined by a normal distribution.

84. A method according to any one of claims 71 to 83, wherein the determined value distribution of the monitored signal follows a log-normal distribution.

85. The method of any one of claims 63 to 84, wherein the duration is one day.

86. 86. The method of any one of claims 63 to 85, further comprising continuing to determine an updated sensitivity based on a previous updated sensitivity and the received analyte concentration value.

87. 83. The method of claim 82, further comprising detecting a slow moving average of the monitored analyte concentration values.

88. 88. The method of claim 87, further comprising prompting a user to enter data if the absolute value of the change in the slow moving average is greater than a predetermined threshold over a predetermined unit of time.

89. 90. The method of claim 87, further comprising determining whether the change is due to a systematic error or a change in the actual sensitivity of the sensor if the absolute value of the change in the slow moving average is greater than a predetermined threshold over a predetermined unit of time.

90. 90. The method of claim 89, wherein the determining whether the change is due to a systematic error or a change in the actual sensitivity of the sensor comprises determining whether the subsequent behavior of the sensitivity is consistent with a known sensitivity profile.

91. 91. The method of claim 90, wherein the known sensitivity profile is an envelope of a sensitivity curve.

92. 92. The method of claim 91, further comprising updating the sensitivity based at least in part on the value of the change in the slow moving average if the absolute value of the change in the slow moving average is determined to be due to a change in actual sensitivity of the sensor.

93. 90. The method of claim 89, wherein the determining whether the change is due to a systematic error or due to a change in the actual sensitivity of the sensor comprises determining whether subsequent behavior of the analyte concentration values ​​is consistent with a known envelope of physiological feasibility.

94. 92. The method of claim 91, further comprising prompting the user to input data if the absolute value of the change in the slow moving average is determined to be due to a systematic error.

95. 1. A method for calibrating an analyte concentration sensor in a biological system using a signal from the analyte concentration sensor, comprising: a. receiving or determining seed values ​​for calibration parameters associated with an analyte concentration sensor; b. determining, at least in part, a calibration of the analyte concentration sensor using the seed value; c. measuring an analyte concentration value using the analyte concentration sensor; and d. displaying measurements calibrated at least in part using said seed value.

96. 96. The method of claim 95, wherein the receiving or determining is performed on a monitoring device in operative signal communication with the analyte concentration sensor.

97. 97. The method of claim 96, wherein the display is performed on the monitoring device or on a mobile device in signal communication with the monitoring device.

98. and displaying the measurements includes displaying a graph or table showing at least historical values ​​of the analyte concentration. a. Detecting that a change in calibration has occurred; b. adjusting one or more calibration parameters of the analyte concentration sensor in accordance with the detected change in calibration; c) following said adjusting, updating said display of said graph or table showing at least historical values ​​of said analyte concentration according to said adjusted calibration parameters.

99. 99. The method of claim 98, wherein the adjustment of the calibration is configured to occur when sensor readings are substantially stable or within a predetermined range of readings for a threshold period of time, thereby reducing the occurrence of unexpected spikes in readings.

100. 99. The method of claim 98, wherein the updating changes the indication of the historical value of the analyte concentration.

101. The method of any one of claims 95 to 100, wherein the seed value is based at least in part on a code.

102. 102. The method of claim 101, wherein the code is entered into a monitoring device by a user.

103. 102. The method of claim 101, wherein a monitoring device is configured to receive the code without substantial involvement of the user.

104. A method according to any one of claims 95 to 100, wherein the seed value is based at least in part on an impedance measurement.

105. The method of any one of claims 95 to 100, wherein the seed value is based at least in part on information associated with a manufacturing lot of the sensor.

106. The method of any one of claims 95 to 100, wherein the seed value is based at least in part on a population average.

107. The method of any one of claims 95 to 100, wherein the seed value is based at least in part on a previous analyte value of the user.

108. 1. A method for calibrating and compensating for drift of an in-dwelling analyte concentration sensor in a biological system using only a signal from the analyte concentration sensor, wherein at steady state, an analyte concentration value in the biological system is known, the method comprising: a. detecting, on a monitoring device, when an analyte concentration value measured by an analyte concentration sensor placed in a biological system is at steady state; b. correlating, on the monitoring device or on a device or server operatively coupled to the monitoring device, the measured analyte concentration value when the biological system is at the detected steady state with the known analyte concentration values; c. determining a first slow moving average of the measurement of the analyte over a first time period and referencing a calibration of the sensor based at least in part on the first slow moving average and the known analyte concentration value; d. determining a second slow moving average of the measurement of the analyte over a second period of time subsequent to the first determination; e. adjusting the calibration of the sensor based at least in part on a difference between the first slow moving average and the second slow moving average.

109. 109. The method of claim 108, wherein the adjustment of the calibration is configured to occur when sensor readings are substantially stable or within a predetermined range of readings for a threshold period of time, thereby reducing the occurrence of unexpected spikes in readings.

110. 1. A method of calibrating a first portion of a number of sensors, a second portion of which is in use, comprising: a. receiving calibration data from some of the second portions of sensors; b. updating one or more calibration parameters of the first portion based on the received data.

111. 111. The method of claim 110, wherein the updating is performed before the first portion is incorporated into a user.

112. 111. The method of claim 110, wherein the update is performed after the first portion is incorporated by a user.

113. 113. The method of claim 112, wherein the updating is accomplished by transmitting, over a network, new or updated calibration parameters to a monitoring device or sensor electronics module associated with the sensor.

114. A method according to any one of claims 110 to 113, wherein the second part of the sensor is configured to be calibrated using an a priori calibration.

115. A method according to any one of claims 110 to 114, wherein the second part of the sensor is configured to be calibrated using user data.

116. The method of any one of claims 110 to 115, wherein the second portion of the sensor is configured to be calibrated using an ex vivo bench calibration.

117. A method according to any one of claims 110 to 116, wherein the second part of the sensor is configured to be calibrated using a blood measurement.

118. 1. A method for compensating for drift of an analyte concentration sensor in a biological system using only a signal from the analyte concentration sensor, comprising: a. measuring the value of an analyte as a function of time using an indwelling analyte concentration sensor; b. filtering the measurements using a double exponential smoothing filter; c) subsequent to said filtering, displaying said filtered measurements versus time.

119. The double exponential smoothing filter is [0010] is governed by, where 【number】 and wherein 【number】 The method of claim 118, wherein:

120. The glucose signal as a function of time is [0025] wherein: [0030] where: [0045] The method of claim 119,

121. 1. A method for calibrating an analyte concentration sensor in a biological system using only a signal from the analyte concentration sensor, comprising: a. receiving data corresponding to analyte concentration values ​​over a period of time; b. determining a coefficient of variation of the received data; c. determining a standard deviation of the analyte concentration values ​​based at least in part on the determined coefficient of variation; d. using the determined standard deviation in calculating a calibration parameter, the calibration parameter being associated with the analyte concentration value.

122. 122. The method of claim 121, wherein the determining the coefficient of variation of the received data comprises dividing a standard deviation of the received data by a mean of the received data.

123. 1. A method for compensating for drift of an analyte concentration sensor in a biological system using only a signal from the analyte concentration sensor, comprising: a. measuring an analyte level using an indwelling analyte concentration sensor; b. determining a first slow moving average of the measurement of the analyte over a first time period and referencing a calibration of the sensor based at least in part on the first slow moving average; c. determining a second slow moving average of the measurement of the analyte over a second period of time subsequent to the first determination; d. adjusting the calibration of the sensor based at least in part on a seed value and a difference between the first slow moving average and the second slow moving average.

124. 123. The method of claim 122, wherein the seed value is determined using a signal median, a drift value, or both.

125. A method according to any one of claims 122 to 124, wherein both forward and backward filters are used, further comprising optimizing seed values ​​to minimize the mean squared error between the two signal filters.

126. 126. The method of claim 125, further comprising adjusting a sensitivity and baseline for the sensor according to a signal-based calibration algorithm, the signal-based calibration algorithm using an average of signals from the forward and reverse filters along with a raw sensor signal.

127. 127. The method of claim 126, further comprising adjusting the sensitivity and the baseline based on one or more criteria.

128. 128. The method of claim 127, wherein the criteria include that the average glucose value should be consistent with a predicted diabetes average value.

129. 128. The method of claim 127, wherein the criteria include that CGM glucose variability should be consistent with average glucose levels.

Images