Methods and apparatus for information gathering, error detection and analyte concentration determination during continuous analyte sensing

Abstract

To provide methods for information gathering and / or error correction for CGM glucose concentration determination.SOLUTION: A continuous glucose monitoring (CGM) device may include a wearable portion having a sensor configured to produce glucose signals from interstitial fluid, a processor, a memory and a transmitter circuitry. The memory may include a computer program code stored therein that, when executed by the processor, causes the CGM device to (a) measure and store a plurality of glucose signals using the sensor and memory; (b) for a presently-measured glucose signal, employ the plurality of presently-measured glucose signals stored in the memory and the pre-determined gain function to compute a compensated glucose value; and (c) communicate the compensated glucose value to a user of the CGM device.SELECTED DRAWING: Figure 9

Description

Technical Field

[0001] Related Applications This application claims the benefit and priority of U.S. Provisional Patent Application No. 62 / 898,507, filed on September 10, 2019, entitled "METHODS AND APPARATUS FOR INFORMATION GATHERING, ERROR DETECTION AND ANALYTE CONCENTRATION DETERMINATION DURING CONTINUOUS ANALYTE SENSING", which is hereby incorporated by reference in its entirety for all purposes.

[0002] The present disclosure generally relates to determining analyte concentration in an analyte-containing fluid using continuous analyte sensing.

Background Art

[0003] Continuous analyte sensing in in vivo and / or in vitro samples, such as continuous glucose monitoring (CGM), has become a routine detection operation, particularly in diabetes care. By providing real-time glucose concentration, treatment / clinical actions can be applied timely and blood glucose status can be better controlled.

[0004] During CGM operation, the biosensor is typically inserted subcutaneously and operates continuously in an environment surrounded by tissue and interstitial fluid (ISF). The biosensor can operate continuously at a constant potential with respect to a reference electrode such as an Ag / AgCl electrode or a composite reference counter electrode. The biosensor may also operate with two working electrodes, where one is dedicated to measuring the analyte of interest, such as glucose, by a glucose-specific enzyme such as glucose oxidase. The other electrode is dedicated to measuring the background signal generated from interfering species such as uric acid, acetaminophen, or the like. In this dual-electrode operation scheme, the interference signal can be continuously subtracted from the main signal of the analyte of interest by either a simple subtraction method or another algorithmic method.

[0005] Optical sensors can also be used for continuous glucose monitoring using fluorescence, absorbance, reflectance, and / or the like. For example, optical oxygen sensors that rely on fluorescence or fluorescence quenching are utilized to indirectly measure glucose by measuring the oxygen concentration in ISF that has an inverse relationship with the glucose concentration. (See, e.g., Stein et al., "Microscale Enzymatic Optical Biosensors using Mass-Transport Limiting Nanofilms. 1. Fabrication and Characterization using Glucose as a Model Analyte," Anal Chem, Author Manuscript, 2008, www.ncbi.nlm.nih.gov / pmc / articles / PMC2518633 / )

[0006] An analytical method may be used to determine the accuracy of the analytical method using a reference concentration to measure the concentration of an analyte in a sample. For a biosensor deployed subcutaneously and exposed to interstitial fluid, the defined glucose signal responds to ISF glucose. However, it is difficult to directly determine the ISF glucose concentration because ISF samples are not readily available for reference ISF glucose measurements. Additionally, the associated therapeutic effects based on glycemic status depend on capillary glucose delivered to cells through the capillary system.

[0007] As is known, ISF glucose lags capillary glucose by about 5 to 15 minutes depending on whether the living system is fasting or in a glucose transition phase. Therefore, ISF glucose can only function as an indicator of capillary glucose considering the time delay, and it is difficult to provide an accurate prediction of capillary glucose using a CGM biosensor. Furthermore, due to system calibration (in-situ calibration or factory calibration), tissue effects on the biosensor membrane, and signal noise due to changes in sensitivity over time to other known factors, the ISF glucose timing profile relative to capillary glucose becomes more difficult to define.

[0008] One conventional method to reduce the glucose delay of ISF and thus increase the accuracy of CGM is by filtering. Another method is so-called delay correction, where the estimated glucose value is compared with the measured glucose value and used to correct the delay. However, due to the errors associated with the measurement of ISF glucose due to the above factors, the filtering method or the delay correction method may prove to be not very significant.

[0009] Improved CGM methods and devices are desired. SUMMARY OF THE INVENTION

[0010] In some embodiments, a method of fabricating a continuous glucose monitoring (CGM) device includes: (1) creating a gain function based on a plurality of sensor progression parameters of a glucose signal measured by a CGM sensor, where each sensor progression parameter is based on the glucose signal at a point of interest and the glucose signal measured prior to the glucose signal at the point of interest; (2) providing a CGM device including a sensor, a memory, and a processor; (3) storing the gain function in the memory of the CGM device; and (4) storing computer program code in the memory of the CGM device, the computer program code causing the CGM device, when executed by the processor, to: (a) measure and store a plurality of glucose signals using the sensor and memory of the CGM device; (b) calculate a plurality of sensor progression parameters for a currently measured glucose signal based on the currently measured glucose signal and the plurality of previously measured glucose signals stored in the memory; (c) calculate a corrected glucose value using the plurality of sensor progression parameters and the stored gain function; and (d) communicate the corrected glucose value to a user of the CGM device.

[0011] In some embodiments, a continuous glucose monitoring (CGM) device includes a wearable portion having a sensor configured to generate a glucose signal from interstitial fluid, a processor, a memory coupled to the processor, and a transmitter circuit coupled to the processor. The memory includes a gain function based on a plurality of sensor progression parameters of the glucose signal, and each sensor progression parameter is based on the glucose signal at the point of interest and the glucose signals measured prior to the glucose signal at the point of interest. The memory includes computer program code stored in the memory, and when the computer program code is executed by the processor, causes the CGM device to (a) measure and store a plurality of glucose signals using the sensor and memory of the wearable portion; (b) calculate a plurality of sensor progression parameters for the currently measured glucose signal based on the currently measured glucose signal and the plurality of previously measured glucose signals stored in the memory; (c) calculate a corrected glucose value using the plurality of sensor progression parameters and the stored gain function; and (d) communicate the corrected glucose value to a user of the CGM device.

[0012] In some embodiments, a continuous glucose monitoring (CGM) device includes a wearable portion having a sensor configured to generate a glucose signal from interstitial fluid, a current sensing circuit coupled to the sensor and configured to measure the glucose signal generated by the sensor, and a transmitter circuit configured to transmit the measured glucose signal. The CGM device also includes a portable user device having a memory, a processor, and a receiver circuit configured to receive the glucose signal from the wearable portion. The memory includes a gain function based on a plurality of sensor progression parameters of the glucose signal, each sensor progression parameter being based on the glucose signal at a point of interest and the glucose signal measured prior to the glucose signal at the point of interest. The memory includes computer program code stored in the memory, which, when executed by the processor, causes the CGM device to: (a) acquire and store a plurality of glucose signals using the sensor of the wearable portion and the memory of the portable user device; (b) calculate a plurality of sensor progression parameters for the currently measured glucose signal based on the currently measured glucose signal and the plurality of previously measured glucose signals stored in the memory; (c) calculate a corrected glucose value using the plurality of sensor progression parameters and the stored gain function; and (d) communicate the corrected glucose value to the user of the CGM device.

[0013] In some embodiments, a method for correcting errors during continuous glucose monitoring (CGM) measurements includes: (a) providing a CGM device including a sensor, a memory, and a processor, the CGM device having a gain function stored in the memory, the gain function being based on a plurality of sensor progression parameters of a glucose signal, each sensor progression parameter being based on the glucose signal at a point of interest and the glucose signal measured prior to the glucose signal at the point of interest; (b) using the sensor and the memory to measure and store a plurality of glucose signals; (c) calculating, for a currently measured glucose signal, a plurality of sensor progression parameters based on the currently measured glucose signal and the plurality of previously measured glucose signals stored in the memory; (d) calculating a corrected glucose value using the plurality of sensor progression parameters and the stored gain function; and (e) communicating the corrected glucose value to a user of the CGM device.

[0014] In some embodiments, a method for determining an analyte concentration during continuous monitoring measurements includes: (a) subcutaneously inserting a biosensor into a subject, the biosensor including a counter electrode, a reference electrode, and a working electrode having a chemical composition configured to oxidize the analyte at a point of interest; (b) applying a constant voltage to the working electrode having the chemical composition, thereby generating a continuous current from the working electrode; (c) sensing and storing in a memory a working electrode current signal from the working electrode; (d) aggregating the working electrode current signal at the point of interest and a portion of the working electrode current signal stored in the memory that was measured prior to the working electrode current signal at the point of interest; (e) generating a gain function value from a predetermined gain function using the working electrode current signal at the point of interest and the portion of the working electrode current signal aggregated from the memory; (f) modifying a system gain using the gain function value generated from the predetermined gain function; and (g) determining the analyte concentration of the working electrode current signal at the point of interest based on the modified system gain and the working electrode current signal at the point of interest.

[0015] In some embodiments, a continuous analyte monitoring (CAM) device is provided that includes a wearable portion having a biosensor configured to be subcutaneously inserted into a subject, the biosensor including a counter electrode, a reference electrode, and a working electrode having a chemical composition configured to oxidize an analyte of interest and generate an analyte signal from interstitial fluid; a processor; a memory coupled to the processor; and a transmitter circuit coupled to the processor. The memory includes an analyte signal of interest and a predetermined gain function based on analyte signals measured prior to the analyte signal of interest. The memory includes computer program code stored in the memory, which when executed by the processor causes the CAM device to: (a) apply a constant voltage to the working electrode having the chemical composition, thereby generating a continuous current from the working electrode; (b) sense a working electrode current signal from the working electrode and store it in the memory; (c) aggregate the working electrode current signal of interest and a portion of the working electrode current signal stored in the memory measured prior to the working electrode current signal of interest; (d) generate a gain function value from the gain function using the working electrode current signal of interest and the portion of the aggregated working electrode current signal from the memory; (e) correct the system gain using the gain function value generated from the predetermined gain function; and (f) determine an analyte concentration of the working electrode current signal of interest based on the corrected system gain and the working electrode current signal of interest.

[0016] In one embodiment, a method of fabricating a continuous analyte monitoring device includes: (a) operably coupling an analyte sensor to a host for use during a continuous analyte monitoring process; (b) continuously recording an analyte signal during the continuous analyte monitoring process; (c) recording a reference analyte concentration during the continuous analyte monitoring process; (d) establishing a data pairing between the analyte signal and the reference analyte concentration; (e) calculating a relative analyte error referenced to the reference analyte concentration; (f) calculating sensor progression parameters by aggregating sensor progression information and referencing analyte data points of interest to previously measured analyte data points; (g) setting at least one of a relative analyte error referenced to the reference analyte concentration and a relative gain error referenced to a reference gain as a target for statistical analysis, and performing statistical analysis by setting sensor progression parameters as input variables to obtain a gain function; and (h) recording a gain function including selected sensor progression parameters and their weighting factors as factory calibration components for storage in the continuous analyte monitoring device.

[0017] Other features, aspects, and advantages of embodiments according to the present disclosure will become more fully apparent from the following detailed description, the appended claims, and the accompanying drawings, which illustrate numerous exemplary embodiments and implementations. Various embodiments according to the present disclosure may also take other different applications, and some details thereof may be modified in various respects without departing from the spirit and scope of the appended claims. Accordingly, the drawings and description are to be regarded as illustrative in nature and not as restrictive. The drawings are not necessarily drawn to scale. BRIEF DESCRIPTION OF THE DRAWINGS

[0018]

Figure 1A

Figure 1B

Figure 1C

Figure 1D

Figure 1E

Figure 1F

Figure 2

Figure 3A

Figure 3B

Figure 4A

Figure 4B

Figure 4C

Figure 4D

Figure 4E

Figure 4F

Figure 5A

Figure 5B

Figure 6A

Figure 6B

Figure 7A

Figure 7B

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

[0019] Overview Methods and devices for continuous glucose monitoring (CGM) have been developed to more closely monitor a person's glucose level and detect shifts in glucose levels. A CGM system generates a "continuous" glucose signal during operation, such as a continuous electrochemical and / or optical signal, but the measurement of the generated glucose signal is typically not truly continuous and is performed every few minutes. A CGM system having implantable and non-implantable portions can be worn for several days before removal and replacement. A CGM system can include a sensor portion that is inserted to be located under the skin and a non-implantable processing portion that is adhered to an outer surface of the skin, such as the abdomen or the back of the upper arm. Unlike a blood glucose monitoring (BGM) system that measures the glucose concentration in blood, a CGM system measures the glucose concentration in interstitial fluid or a sample of non-direct capillary blood.

[0020] A CGM system can provide frequent measurements of a person's glucose level without the need for each such measurement to involve collection of a blood sample, such as by a fingerstick. A CGM system may still occasionally employ the use of a fingerstick and a BGM system, such as the Contour NEXT One® by Ascensia Diabetes Care AG of Basel, Switzerland, for calibration of the CGM system.

[0021] As described above, during CGM, the biosensor may operate continuously at a constant potential with respect to the reference electrode or the composite reference counter electrode. Since potential pulses for each data point can destabilize the glucose signal that results in a potential degradation of the signal quality, methods equivalent to gated amperometry used in the BGM field are not employed during CGM. Thus, there is a lack of meaningful information that can be readily utilized to assist glucose measurement through the algorithm method during CGM.

[0022] Within the BGM test field, the gated amperometry method described in U.S. Patent Publication No. 2013 / 0256156, entitled "Gated Amperometry Methods," applies a bias voltage to the test strip, and a group of signal / data points are measured in response to the bias applied for the final analyte determination. In the segmented signal processing method of optical and electrochemical sensors described in U.S. Patent Publication No. 2013 / 0071869, entitled "Analysis Compensation Including Segmented Signals," data points within a single process are used together to provide information for endpoint analyte determination. These patent publications describe individual tests in a transient process where one analyte determination is independent of all other analyte determinations. Thus, a group of signal / data points from a single sensor test is used only for the purpose of providing a single glucose measurement / analyte determination for the test strip / cartridge. Subsequent glucose / analyzer measurements each rely on a new group of data points (and a new test strip / cartridge).

[0023] In contrast, during a continuous glucose monitoring process, according to the embodiments described herein, each measured signal / data point is its own endpoint for analyte determination in the data continuum, but is also related to its adjacent data points in short-term and / or long-term relationships. According to the embodiments provided herein, due to the continuous nature of CGM, previous signal / data points may contain information related to the signal / data points measured thereafter. That is, each data point may be related to its adjacent (e.g., previous) data points, or even data points obtained at much earlier times. The relationship of the current (point of interest) data point to many of the previously measured data points in the continuum has been found to include sensor error and / or status information (referred to herein as "sensor progress information"). In some embodiments, the previous data points can be a source of information suggesting a sensor error source or sensor state. The parameters related to the current data point or point of interest data point and the previously measured data points are referred to herein as sensor progress parameters (SPP).

[0024] As described herein, sensor progression parameters in CGM or other continuous analyte monitoring methods can be determined by referring to the current analyte signal in the form of a ratio, difference, relative difference, and / or the like to previously measured analyte signals in the data continuum. According to one or more embodiments described herein, methods for information aggregation and / or error correction for CGM glucose concentration determination are provided. In some embodiments, the method may include generating a gain function value using previously measured glucose signals from the CGM sensor along with the CGM glucose signal at the current point of interest. The gain function value is generated from the gain function (using the SPP) and can be used to adjust the error in the gain used to determine the glucose concentration from the glucose signal at the point of interest, as well as the error due to ISF glucose delay. In some embodiments, if each and every glucose reading in the CGM data continuum is accurately determined point-by-point by correcting and / or reducing the error from signal deviation and ISF delay, there will be little or no ISF delay relative to the reference glucose profile.

[0025] For example, the gain function value may be determined from a gain function that uses a sensor progression parameter (SPP), which is calculated from the glucose signal at the current point of interest and the previously measured glucose signal (e.g., using ratios, differences, etc., as described below). That is, the gain function may be a function of the SPP (i.e., gain function = f(SPP)). For example, the operating electrode current signal and / or the background current signal from a biosensor (e.g., a CGM sensor) may be periodically sampled and stored in memory. For subsequent glucose signals at points of interest (e.g., the operating electrode current signal), the stored current signals can be used together with the glucose signal at the point of interest to calculate the gain function value from the gain function. The gain function value can then be used during the calculation of the glucose concentration of the glucose signal at the point of interest to reduce the error in the calculated glucose concentration of the glucose signal at the point of interest (e.g., by adjusting the system gain using the gain function value).

[0026] As further described below, the gain function can be determined using statistical techniques such as, for example, multivariate regression. In some embodiments, the gain function is pre-determined by the CGM device manufacturer and stored in the memory of the CGM device for use during glucose monitoring with the CGM device.

[0027] In one or more embodiments, the method can include generating a series of sensor progression parameters by taking the ratio of the glucose signal from the current data point to the glucose signal from a previous data point. These ratios, and / or combinations of these ratios and / or other related terms, can be employed within a predetermined gain function that enables error correction for sources of error in the CGM glucose signal, such as gain changes over time and ISF delays. This can increase the accuracy of the CGM and / or assist in treatment measures taken in response to CGM glucose measurements. Biosensor systems according to these and other embodiments are provided.

[0028] Although mainly described with respect to glucose concentration measurements during continuous glucose monitoring, it will be understood that the embodiments described herein may be used with other continuous analyte monitoring systems (e.g., cholesterol, lactate, uric acid, alcohol, or other analyte monitoring systems).

[0029] As an example, one or more gain functions can be developed that include a selected group of sensor progression parameters, such as ratios based on glucose signals taken at different times, combinations of such ratios, and other cross terms in a linear combination. A cross term can include, for example, terms of an SPP (e.g., ratio, difference, etc.) in relation to other types of parameters such as initial glucose (G RAW ), normalized gain, background / interference signal, motion parameter, temperature value, different ratios, and / or the like. Non-linear combinations may also be used. A sample gain function can take the following form: (1) Gain Function = c1*R_t1 + c2*R_t2 + c3*R_t3 + c4*R_t4 +... c n *R_tn where c1, c2, c3, c4... c nwhere the weighting factor and R_t1, R_t2, R_t3, R_t4...R_tn are ratios of glucose data points captured at different times (e.g., the "point of interest" glucose data point divided by the glucose data point measured and / or sensed before the glucose data point of interest), combinations of ratios of glucose data points, or other cross terms. Thus, the gain function represents the relative error in the gain and / or the relative error in glucose, and is derived from information aggregated in the form of sensor progression parameters from the current (point of interest) glucose signal and previously measured glucose signals (in some embodiments, for example, glucose signals measured up to 12 hours or more before the current glucose signal, minutes before). More specifically, in some embodiments, the gain function can be derived by multivariate regression or other statistical analysis techniques from the ratio, difference, and / or relative difference of the current glucose signal to past glucose signals, and their cross terms in the form of sensor progression parameters with each other and / or other parameters. In some embodiments, the gain function may be based on, and / or include, dozens or hundreds (or more) of sensor progression parameters such as ratios, differences, relative differences, and / or cross terms. Exemplary gain functions, and methods for determining such gain functions, are described below.

[0030] According to embodiments provided herein, the error in the raw or uncorrected glucose signal Signal Raw can be corrected and / or otherwise adjusted by using sensor progression parameters in the gain function that reference the current data point to past data points. For example, in some embodiments, the corrected glucose signal G Comp can be calculated as follows: (2) G Comp = Signal Raw * Gain * (1 / (1 + Gain Function)) where the gain is, for example, the calibration glucose value (G BGM ) from a blood glucose meter, divided by the CGM sensor current (Signal CGMDividing by (Gain = G BGM / Signal CGM ) represents the (system) gain determined from in - situ calibration, etc.

[0031] In some embodiments, one or more gain functions are determined and stored in the memory of a CGM device, such as one or more other parts of a wearable or CGM device, and can be used to calculate a corrected glucose value based on glucose signal ratios (and / or other relationships) measured by an interstitial CGM sensor and sensor progression parameters such as an uncorrected glucose signal.

[0032] Conventional expressions of BGM accuracy are percentages within ±x% accuracy limits such as ±20%, ±15%, or ±10%, which represent the percentage difference (100% * 〔G BGM -G Ref 〕 / G Ref ) of the BGM glucose value relative to a reference glucose value, and determine the percentage of data points within a certain accuracy limit in a sample population. The smaller the accuracy limit, the higher the accuracy.

[0033] For CGM glucose determination, the measurement accuracy can be defined by the following mean absolute relative difference (MARD): (3) MARD = Σ[Abs([G CGM -G REF / G REF )] / n) Wherein, G CGM is the CGM - measured glucose value, and G REFis, for example, a reference glucose value measured by BGM, and n is the number of data points. The expression of MARD combines the mean and standard deviation of the sample population with respect to the reference glucose value to create a composite MARD value, where the smaller the MARD value, the better the accuracy. The BGM rule of accuracy is not used to evaluate the error within a specific accuracy limit, but depending on the mean and standard deviation of the error in the data population, the inherent accuracy expressed as a percentage within the ±x% accuracy limit can be estimated to be approximately 2.5 times the MARD value. Therefore, a MARD value of 10% can have an approximate accuracy of the data within ±25%, or an approximate accuracy of 25%. Conversely, a BGM system with an accuracy of ±10% is predicted to have a MARD value of 4%. The embodiments described herein may enable a reduction in the MARD value of the CGM device (e.g., in some embodiments, about 7 to 10% or less).

[0034] According to the embodiments provided herein, the sensor progression parameter that refers to previously measured data points from the current (point of interest) data point can be expressed by the ratio of the signal from the current data point to the signal of the previously measured data point. This may form a network of information embedded in the sensor progression parameter, which is sent to the calculation of the CGM glucose value of the current data point to improve accuracy. The ratio formed from the current or present data point and the previous data point may be referred to herein as the "current - past ratio" for convenience. The current - past ratio can be calculated for the operating electrode current Iw, the background current Ib, and Iw - Ib, or optical signals such as fluorescence, absorbance, and / or reflectance signals, and / or the like.

[0035] Figure 1A shows a reference of a current data point to past data points in the collection of glucose signals captured over a 3 - hour period, and the use of the collected glucose signals to calculate sensor progression parameters such as ratios, according to the embodiments described herein. Referring to Figure 1A, past glucose signal data points measured 3 hours, 2 hours, 1 hour, 9 minutes, 6 minutes, and 3 minutes before the currently measured glucose signal data point are shown (indicated by the circles above R_3hr, R_2hr, R_1hr, R_9min, R_6min, and R_3min, respectively). Using these past data points, a ratio to the current data point can be calculated (referred to as R_3hr, R_2hr, R_1hr, R_9min, R_6min, and R_3min in Figure 1A). Other numbers of ratios may be calculated and / or other time increments may be used. Ratios to future data points can be calculated similarly as CGM sensing progresses during CGM operation. As shown as R_4hr, R_8hr, and R_12hr in Figure 1B, ratios can be used that employ even "older" data points, such as data points acquired 4, 8, or even 12 hours before. Longer or shorter ranges of past data points may be used.

[0036] Examples of exemplary current - past ratios are shown below, where Iw t represents the current data point of the operating electrode current at time t (the time of the point of interest), and Iw t-xmin represents the past data point of the operating electrode current at time t - x min, measured x minutes before the current data point. For example, based on the current operating electrode current and the operating electrode current 3 minutes before, the current - past ratio R_3min of the operating electrode current is (4) R_3min = Iw t / Iw t-3min is. In this particular case, data points are acquired regularly at 3-minute intervals. The ratio for a longer period may be based on a time that is a multiple of 3 minutes. For example, the present-past ratio for the working electrode current is 6 minutes, 9 minutes, 1 hour, 3 hours, and 12 hours earlier than the working electrode current at the current point of interest, and is shown in the following equations (5) to (9) respectively. (5)R_6min=Iw t / Iw t-6min (6)R_9min=Iw t / Iw t-9min (7)R_1hr=Iw t / Iw t-1hr (8)R_3hr=Iw t / Iw t-3hr (9)R_12hr=Iw t / Iw t-12hr Other measurement intervals may be used. For example, if the data acquisition rate is based on measurements taken every 5 minutes, 10 minutes, or 15 minutes, the present-past ratio can be a multiple of 5 minutes, 10 minutes, or 15 minutes. Similar ratios can be determined for, for example, the background current Ib, the current difference between the working electrode and the background current, or the like, as shown by the following equations (10) to (17). (10)R_6min=Ib t / Ib t-6min (11)R_9min=Ib t / Ib t-9min (12)R_1h=Ib t / Ib t-1hr (13)R_3h=Ib t / Ib t-3hr (14)R_6min=(Iw t -Ib t ) / (Iw t-6min -Ib t-6min ) (15)R_9min=(Iw t -Ib t ) / (Iw t-9min -Ib t-9min) (16)R_1h=(Iw t -Ib t ) / (Iw t-1hr -Ib t-1hr ) (17)R_3h=(Iw t -Ib t ) / (Iw t-3hr -Ib t-3hr )

[0037] Cross terms that include ratios and other parameter combinations, and / or combinations of multiple ratios, can also be determined as further described below. Thus, for each measured data point, there is a set of parameters associated with the measured data point that can be obtained using previous data points. As described above, for sensor progression parameters, the glucose signal at the current or "point of interest" can be referenced to past glucose signals measured 6 hours, 8 hours, 10 hours, or even 12 hours or more before the glucose signal at the point of interest is measured. In some embodiments, the sensor progression parameters may be calculated with respect to a ratio, difference, or other relationship between the current (point of interest) glucose signal and a previously measured glucose signal, where the signal may be an electrochemical current or an optical signal such as fluorescence, absorbance, or reflectance.

[0038] In some cases, after a CGM sensor is inserted into a patient, a warm-up period (e.g., a 3-hour warm-up period or a shorter or longer warm-up period) can be employed. In such cases, there can be a period of several hours (e.g., 3 hours or a long warm-up time) during which only the data points collected during the warm-up period can be obtained. After the warm-up period, as more data points are collected, ratios or other sensor progression parameters can be calculated based on increasingly older data points (e.g., 4 hours, 5 hours, 6 hours, etc.). In some embodiments, the sensor progression parameters can be calculated using the current glucose signal and past glucose signals measured up to 12 hours prior. Other cut-off points (e.g., longer or shorter than 12 hours) may be used.

[0039] Figures 1C, 1D, and 1E show Tables 100a, 100b, and 110c of exemplary ratios of the operating electrode current Iw, the background electrode current Ib, and the Iw - Ib current difference of a CGM sensor, respectively, using previous data points measured up to 1 hour before the last data point (acquired at time = 1 hour) according to embodiments provided herein. The labels of the ratios in Table 100a (e.g., w_3min, w_6min, etc.) indicate that the ratio is from the operating electrode (e.g., enzyme electrode) current Iw. The labels of the ratios in Table 100b indicate that the ratio is from the background current Ib, and the labels of the ratios in Table 100c indicate that the ratio is from the Iw - Ib current value. For each row, the ratio is calculated by dividing the leftmost data point by the previously measured data point. For example, the latest Iw, Ib, or Iw - Ib signal (acquired at time = 1 hour) is divided by the previously measured current signal (at 3 - minute intervals, going back to time = 0). As seen in Tables 100a, 100b, and 100c, a number of ratios can be developed for each measured glucose signal based on the previously measured glucose signal.

[0040] Different sensor progression parameters contain different information for the current signal based on the specific previously measured signal used. For example, Figure 1F shows a graph 100d of exemplary ratios of the operating electrode current signal taken over 50 hours, calculated by dividing each operating electrode current signal by the operating electrode current signals measured 3 minutes, 30 minutes, and 2 hours before (w_3min, w_30min, and w_2hr, respectively). In this plot, the different ratios at any given time t have different magnitudes, resulting in different temporal profiles of the sensor progression parameters. As shown in Figure 1F, at each time point, there are different ratios representing different information from previous data points that can be used to enhance the accuracy of glucose measurement (as described below).

[0041] As described above, there are two major error sources in the error during CGM measurement: signal error and ISF glucose delay. The signal error, which is the first error source, can be caused by sensitivity changes over time or even sensitivity changes during the calibration period. This can be seen in FIG. 2, which shows a normalized gain versus time graph 200 of an example of a series of in-situ CGM calibrations for two CGM sensors (sensor 1 and sensor 2) according to an embodiment provided herein. Specifically, each plateau or horizontal region, such as plateau 202 in FIG. 2, represents the normalized gain calculated by dividing the BGM glucose value determined using a blood glucose meter by the glucose signal (e.g., Iw or Iw - Ib) of the CGM sensor (sensor 1 or sensor 2), and then dividing this gain by the initial gain (referred to as Gain_1). For example, according to the following equation (18), (18)Gain = G Ref-cal / Signal cal where, in the formula, G Ref-cal is the reference glucose value from the blood glucose meter, and Signal cal is the raw glucose signal measured from the CGM sensor (e.g., the operating electrode current, the operating electrode current minus the background electrode current, or the like). Gain_1 is the initial gain of the CGM sensor, and (19)Gain_1 = G Ref-cal_1 / Signal cal_1 where, as a result, the normalized gain, Gain / Gain_1, is (20)Gain / Gain_1=(G Ref-cal / Signal cal ) / (G Ref-cal_1 / Signal cal_1 ).

[0042] Gain (also referred to as system gain) is defined similarly to electronic gain and has the physical dimension of [concentration / signal]. Thus, if the BGM concentration is in [mg / dL] and the sensor current signal is in [nanoAmps or nA], the unit of gain is [mg / dL] / [nA]. -1 is.

[0043] Each step gain of the gain curve 200 in FIG. 2 represents the gain used to convert the CGM sensor signal to glucose concentration, which can be done by the following equation. (21)G Raw =Gain*Signal where G Raw represents the initial (uncorrected) glucose value, Gain is the gain determined by calibration (G Ref-cal / Signal cal ), and Signal is the glucose signal from the CGM sensor (e.g., Iw or Iw - Ib).

[0044] By providing a series of in - situ calibrations, the set of gains forms the gain curve in FIG. 2, which reflects the sensor sensitivity changes and provides cross - sectional calibration to the CGM sensor during the process of CGM sensor deployment (e.g., typically about 1 - 2 weeks or 7 - 14 days). However, further changes in sensitivity between in - situ calibrations can be a source of error in the long - term monitoring process. The gain curve in FIG. 2 is specific to the sensor used and depends on in - situ calibrations (e.g., taken periodically during the CGM process) using a reference glucose value such as the BGM glucose value. That is, the gain curve in FIG. 2 is calculated and / or adjusted based on the data points measured during the CGM process.

[0045] Another source of error is the apparent ISF (glucose) delay, as shown in the schematic of FIG. 3A. In particular, FIG. 3A shows a graph of glucose versus time as measured using a BGM (reference glucose profile, curve 302), and a CGM (CGM glucose profile, curve 304). When considering the glucose profile from the reference (BGM) glucose measurement and the glucose profile from the CGM sensor, the two glucose profiles are separated or shifted, such that there is a time delay where the ISF (CGM) glucose profile (curve 304) is delayed by a time delay Δt with respect to the reference (BGM) glucose profile (curve 302). The time delay Δt varies depending on whether the glucose was measured during fasting or during a glucose change phase. As described above, conventional methods for reducing this time delay are by filtering, i.e., delay correction. However, although these methods may function to some extent, the time delay may still exist because the nature of the time delay changes.

[0046] As shown in FIG. 3A and according to the embodiments described herein, when each individual error of the glucose concentration ΔG on the CGM glucose profile is reduced / removed, there is no obvious shift of the CGM glucose profile (curve 304) from the reference (BGM) glucose profile (curve 302). By error correction for each point, improvement in the accuracy of CGM glucose measurement becomes possible, as further described below.

[0047] Relationship G Raw =Gain*Signal shows that the relative change in glucose ΔG / G is equal to the relative change in the sensor conversion gain, ΔGain / Gain, which holds the signal constant. That is, (22)ΔG / G = ΔGain / Gain = (Gain act -Gain cal ) / Gain cal =Gain act / Gain cal -1, and where Gain act is the actual gain that fully accounts for the error of the CGM system, while Gain cal is the gain from in-situ calibration (e.g., fingertip stick readings from BGM). Gain act is the Gain from a pair of data points (e.g., glucose signal and reference glucose value) within one calibration period of the study act =G BGM / Signal act and can be determined by. In contrast, Gain cal can be determined after each in-situ calibration (Gain cal is obtained as Gain cal =BGM cal / Signal cal ). Gain act , and Gain cal may be the same in the absence of error, or they may be different if there is some error. At the same time, as follows, the relative change in glucose ΔG / G is also equal to the relative change in the signal, holding the gain constant, (23) ΔSignal / Signal = (Signal act - Signal ideal ) / Signal ideal = Signal act / Signal ideal - 1, and where Signal act is the real-world signal that includes a part of the error leading to the error of the actual glucose G act , while Signall ideal is the calibration gain Gain calis the ideal (error-free) signal for performing glucose measurements without error. In each of the above, the glucose, gain, and relative change in the signal refer to the ideal value, calibrated value, or true value (through the denominator term). As can be seen from Equation (21), the relative change in gain is opposite to the relative change in the signal that holds the glucose constant. This can also be seen from Equations (22) and (23), where the complete gain Gain act (considering all system errors) is present in the numerator, while the ideal signal Signal ideal is present in the denominator of Equation (23). This means that the relative signal error of ΔSignal / Signal is equal in magnitude but opposite in direction to the relative gain change of ΔGain / Gain. Any change in the signal ΔSignal / Signal is assumed to be due to a change in the sensor gain, ΔGain / Gain, but in the opposite direction. Therefore, taking into account the signal error, Gain cal can be adjusted to Gain cal / (1 + ΔGain / Gain). Thus, the final glucose value G final is (24) G final = Signal * Gain / (1 + ΔGain / Gain), where the modification factor 1 / (1 + ΔGain / Gain) represents the relative change in gain that defines the instantaneous calibration state, but is in the opposite direction to the relative signal change. G final can also be referred to herein as the corrected glucose value G comp , and the signal can be referred to as the raw or uncorrected glucose signal Signal Raw . Then, Equation (24) can be written as G Comp = Signal Raw * Gain * (1 / (1 + Gain Function)), which is Equation (2) above. For a transfer function having a non-linear relationship or a stepwise calculation of the analyte concentration, the correction relationship can be expressed as (25) G comp = G raw / (1 + Gain Function), where in the formula, G raw is the initial glucose of glucose from these other conversion functions.

[0048] G Final = Signal Raw * Gain / (1 + ΔGain / Gain), considering the above relationship, the goal is to find sensor progression parameters such as ratios or other parameters that satisfy and / or define ΔGain / Gain of the gain function. From the previous consideration of the current - past ratio as information aggregated from previous data points, in some embodiments, the gain function can be derived from these ratios and their cross - terms. There may be no explicit or obvious correlation between some single ratios and the gain function ΔGain / Gain, or the relative glucose change ΔG / G. However, combining a set of multiple ratio terms and their selective cross - terms can provide the necessary correlation between the relative gain change and the gain function ΔGain / Gain. For example, in some embodiments, multivariate regression can be employed using ΔGain / Gain or ΔG / G as the regression target and a number of current - past ratio terms and cross - terms as input parameters that provide information aggregated from previous data points. In some embodiments, up to 2000 or more combined terms of current - past ratio terms and / or cross - terms can be employed as input parameters. Fewer or more ratio terms and / or cross - terms may be employed, and other relationships between the current (point of interest) data point and previous data points (e.g., differences or other relationships) can also be employed. Figure 3B shows the correlation between the relative glucose error ΔG / G and the gain function defined by a group of ratios and their cross - terms. The gain function is a function of the sensor progression parameter (SPP), i.e., the gain function = f(SPP). R 2 = 64% indicates a strong correlation. R 2The larger the value, the stronger the correlation, and more precisely, the gain function approaches the relative gain change (and the correction result becomes better based on the gain function).

[0049] As an example, using a number of CGM sensors, CGM data for dozens of users was collected over 167 hours. The entire process of the 167-hour CGM data was divided into three segments: (1) 3 - 21 hours, (2) 12 - 45 hours, and (3) 40 - 167 hours. These segments are respectively identified by reference numerals 204a, 204b, and 204c on the gain curve of FIG. 2. Fewer or more segments may be used. Referring to FIG. 2, the largest change in gain occurs within the second segment 204b. For the first segment 204a, the ratio of previous data points taken up to 3 hours before the signal of the point of interest (due to the 3-hour warm-up time) may be adopted. For example, in some embodiments, data collected during the warm-up period may be used to calculate the current-past ratio starting from 3 hours. For the second segment 204b, the ratio of previous data points taken up to 12 hours before the signal of the point of interest may be adopted, focusing on the relatively large change in gain. For the third segment 204c, the previous data points taken up to 12 hours before the signal of the point of interest may be adopted. As described above, older previous data points may be used.

[0050] Multivariate regression can be performed using any suitable data analysis and / or statistical software package to obtain the gain function for each of the segments 204a, 204b, and 204c. For example, Minitab software available from Minitab, LLC of State College, Pennsylvania, or another similar software package may be adopted.

[0051] Using multivariate regression for individual data points and their ratio parameters, the following exemplary gain function examples can be determined. Other ratios, data point relationships, clusters, and / or gain functions may be adopted.

[0052] Figures 4A, 4B, and 4C show exemplary gain functions (referred to as gain function 1, gain function 2, and gain function 3 in FIGS. 4A - C) for segments 204a, 204b, and 204c of FIG. 2 according to the embodiments described herein. FIGS. 4D, 4E, and 4F are lists of sensor progression parameters (e.g., ratios) and cross-stream definitions for gain function 1, gain function 2, and gain function 3, respectively, according to the embodiments described herein. This information is also described in the following appendix section. Other and / or additional numbers of gain functions, sensor progression parameters, cross-streams, coefficient values, and / or constants may be employed. These gain functions and terms of the gain functions are merely representative. Other types and / or numbers of gain functions may be used.

[0053] During operation, the gain function may be stored in the memory of the CGM device and is used to generate a gain function value that is used to calculate a corrected glucose value based on the currently measured glucose signal (e.g., the working electrode current or optical signal) from the CGM sensor and a past glucose signal taken more than 12 hours before the currently measured glucose signal. Using such a gain function, the error in the CGM glucose value caused by gain changes and ISF delays may be significantly reduced. For example, some uncorrected glucose values from the CGM sensor were observed to have a MARD value of 18% - 25%, while the corrected glucose values determined using the gain function were observed to have a MARD value of 7% - 10% according to the embodiments described herein.

[0054] Figures 5A and 5B show exemplary consensus error grid plots 500a and 500b for the raw glucose values and the corrected glucose values of a CGM sensor, respectively, according to embodiments provided herein. The clinical significance of regions A, B, C, D, and E is described below in Table 1 based on Joan L. Parkes et al., “A New Consensus Error Grid to Evaluate the Clinical Significance of Inaccuracies in the Measurement of Blood Glucose,” Diabetes Care, Volume 23(8), pp. 1143 - 1148 (2000).

Table 1

[0055] As shown in FIGS. 5A and 5B, the combined data for regions A and B for glucose values with error correction is greater than 99% (FIG. 5B) compared to less than 98% for uncorrected glucose values (FIG. 5A). Further, the glucose values in region A increase significantly. This performance improvement can also be seen in the effective reduction of the ISF delay for the corrected CGM glucose values. For example, FIGS. 6A and 6B show the time for BGM glucose values, corrected CGM glucose values (G Comp ), and uncorrected CGM glucose values (G Raw ) against a first CGM sensor (sensor 1 in FIG. 6A) and a second CGM sensor (sensor 2 in FIG. 6B) according to embodiments provided herein. By using a gain function, the corrected glucose values (G Comp ) for both CGM sensors appear to essentially contain no ISF delay compared to the raw glucose values (G Raw ).

[0056] FIG. 7A shows a high-level block diagram of an exemplary CGM device 700 according to embodiments provided herein. Although not shown in FIG. 7A, various electronic components and / or circuits are of course configured to couple to a power source such as a battery, among others. The CGM device 700 includes a bias circuit 702 configured to couple to a CGM sensor 704. The bias circuit 702 can be configured to apply a bias voltage, such as a continuous DC bias, to an analyte-containing fluid through the CGM sensor 704. In this exemplary embodiment, the analyte-containing fluid may be human interstitial fluid, and the bias voltage can be applied to one or more electrodes 705 (e.g., working electrode, background electrode, etc.) of the CGM sensor 704.

[0057] In some embodiments, the CGM sensor 704 may include two electrodes, and the bias voltage can be applied across the pair of electrodes. In such a case, a current can be measured through the CGM sensor 704. In other embodiments, the CGM sensor 704 can include three electrodes such as a working electrode, a counter electrode, and a reference electrode. In such a case, the bias voltage may be applied between the working electrode and the reference electrode, and the current can be measured, for example, through the working electrode. The CGM sensor 704 includes a chemical substance that reacts with the glucose-containing solution in a redox reaction and affects the concentration of charge carriers and the time-dependent impedance of the CGM sensor 704. Exemplary chemical substances include glucose oxidase, glucose dehydrogenase, or the like. In some embodiments, a mediator such as ferricyanide or ferrocene may be used.

[0058] The bias voltage generated and / or applied by the bias circuit 702 can be in the range of about 0.1 to 1 volt with respect to the reference electrode, for example. Other bias voltages may be used.

[0059] The current through the CGM sensor 704 in the analyte-containing fluid in response to the bias voltage is the current measurement (I from the CGM sensor 704 meas) can be transmitted to the circuit 706 (also referred to as a current sensing circuit). The current measurement circuit 706 can be configured to sense and / or record a current measurement signal having a magnitude indicative of the magnitude of the current transmitted from the CGM sensor 704 (e.g., using a suitable current-voltage converter (CVC), etc.). In some embodiments, the current measurement circuit 706 may include a resistor having a known nominal value and a known nominal accuracy (e.g., in some embodiments, 0.1% - 5%, or even less than 0.1%), through which the current transmitted from the CGM sensor 704 passes. The voltage generated across the resistor of the current measurement circuit 706 represents the magnitude of the current and the current measurement signal (or raw glucose signal Signal Raw ) may also be referred to as.

[0060] In some embodiments, the sample circuit 708 may be coupled to the current measurement circuit 706 and may be configured to sample the current measurement signal, and may create digitized time-domain sample data representative of the current measurement signal (e.g., a digitized glucose signal). For example, the sample circuit 708 may be any suitable A / D converter circuit configured to receive the current measurement signal, which is an analog signal, and convert it to a digital signal having a desired number of bits as an output. The number of bits output by the sample circuit 708 may be 16 bits in some embodiments, although more or fewer bits may be used in other embodiments. In some embodiments, the sample circuit 708 may sample the current measurement signal at a sampling rate within the range of about 10 samples per second to 1000 samples per second. Faster or slower sampling rates may be used. For example, a sampling rate such as about 10 kHz - 100 kHz may be used to downsample and further reduce the signal-to-noise ratio. Any suitable sampling circuit may be employed.

[0061] Referring further to FIG. 7A, the processor 710 may be coupled to the sample circuit 708 and may be further coupled to the memory 712. In some embodiments, the processor 710 and the sample circuit 708 are configured to communicate directly with each other via a wired path (e.g., via a serial connection or a parallel connection). In other embodiments, the coupling between the processor 710 and the sample circuit 708 may be through the memory 712. In this arrangement, the sample circuit 708 writes digital data to the memory 712 and the processor 710 reads digital data from the memory 712.

[0062] The memory 712 may store therein one or more gain functions 714 for use in determining a glucose value corrected based on a raw glucose signal (from the current measurement circuit 706 and / or the sample circuit 708). For example, in some embodiments, as described above, three or more gain functions may be stored in the memory 712 for use with data collected in different segments (periods) of CGM. The memory 712 may also store a plurality of instructions therein. In various embodiments, the processor 710 may be a computational resource such as, but not limited to, a microprocessor, a microcontroller, an embedded microcontroller, a digital signal processor (DSP), a field programmable gate array (FPGA) configured to operate as a microcontroller, or the like.

[0063] In some embodiments, when the plurality of instructions stored in memory 712 are executed by processor 710, processor 710 is caused to (a) measure a glucose signal (e.g., a current signal) from interstitial fluid by CGM device 700 (via bias circuit 702, CGM sensor 704, current measurement circuit 706, and / or sample circuit 708), (b) store the glucose signal in memory 712, (c) calculate sensor progression parameters such as a ratio (and / or other relationship) of the glucose signal at a point of interest to a previously measured glucose signal, (d) calculate a corrected glucose value (e.g., concentration) using the calculated sensor progression parameters and a stored gain function, and (e) communicate the corrected glucose value to the user. The instructions may include instructions to cause the above actions.

[0064] Memory 712 may be any suitable type of memory, such as, but not limited to, one or more of volatile memory and / or non-volatile memory. Volatile memory may include, but is not limited to, static random access memory (SRAM) or dynamic random access memory (DRAM). Non-volatile memory may include, but is not limited to, electrically programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), flash memory (e.g., of either a NOR configuration or a NAND configuration, and / or of either a stacked or planar arrangement, and / or of a type of EEPROM in any arrangement of single level cell (SLC), multi-level cell (MLC), or a combination of SLC / MLC), resistive memory, filamentary memory, metal oxide memory, phase change memory (e.g., chalcogenide memory), or magnetic memory. Memory 112 may be packaged, for example, as a single chip or as multiple chips. In some embodiments, memory 112 may be embedded, along with one or more other circuits, in an integrated circuit such as, for example, an application specific integrated circuit (ASIC).

[0065] As described above, the memory 712 can have a plurality of instructions stored therein that, when executed by the processor 710, cause the processor 710 to perform various operations specified by one or more of the stored instructions. The memory 712 can further have a portion reserved for one or more "scratch pad" storage areas that can be used for read or write processing by the processor 710 in response to the execution of one or more of the plurality of instructions.

[0066] In the embodiment of FIG. 7A, the memory 712 including the bias circuit 702, the CGM sensor 704, the current measurement circuit 706, the sample circuit 708, the processor 710, and the gain function 714 can be disposed within the wearable sensor portion 716 of the CGM device 700. In some embodiments, the wearable sensor portion 716 can include a display 717 (e.g., without using an external device) for displaying information such as glucose concentration information. The display 717 can be any suitable type of human-sensing display, such as, but not limited to, a liquid crystal display (LCD), a light emitting diode (LED) display, or an organic light emitting diode (OLED) display.

[0067] Referring further to FIG. 7A, the CGM device 700 may further include a portable user device portion 718. The processor 720 and the display 722 may be disposed within the portable user device portion 718. The display 722 may be coupled to the processor 720. The processor 720 may control the text or image shown by the display 722. The wearable sensor portion 716 and the portable user device portion 718 may be communicatively coupled. In some embodiments, the communication coupling between the wearable sensor portion 716 and the portable user device portion 718 may be by wireless communication via a transmitter circuit and / or a receiver circuit, such as the transmit / receive circuit TxRx724a of the wearable sensor portion 716 and the transmit / receive circuit TxRx724b of the portable user device portion 718. Such wireless communication may be by any suitable means, including but not limited to standard-based communication protocols such as the Bluetooth® communication protocol. In various embodiments, the wireless communication between the wearable sensor portion 716 and the portable user device portion 718 may alternatively be by near field communication (NFC), radio frequency (RF) communication, infrared (IR) communication, or optical communication. In some embodiments, the wearable sensor portion 716 and the portable user device portion 718 may be connected by one or more wires.

[0068] The display 722 may be any suitable type of human-sensing display, such as, but not limited to, a liquid crystal display (LCD), a light emitting diode (LED) display, or an organic light emitting diode (OLED) display.

[0069] Referring now to FIG. 7B, there is shown an exemplary CGM device 750 that is similar to the embodiment shown in FIG. 7A but has a different partitioning of components. In the CGM device 750, the wearable sensor portion 716 includes a bias circuit 702 coupled to the CGM sensor 704 and a current measurement circuit 706 coupled to the CGM sensor 704. The portable user device portion 718 of the CGM device 750 includes a sample circuit 708 coupled to the processor 720 and a display 722 coupled to the processor 720. The processor 720 is further coupled to a memory 712 that has a gain function 714 stored therein. In some embodiments, the processor 720 within the CGM device 750 may also perform the aforementioned functions implemented, for example, by the processor 710 of the CGM device 700 of FIG. 7A. The wearable sensor portion 716 of the CGM device 750 may be smaller, lighter, and thus less invasive than the CGM device 700 of FIG. 7A as it does not include a sample circuit 708, a processor 710, a memory 712, etc. Other component configurations may be employed. For example, as a variation of the CGM device 750 of FIG. 7B, the sample circuit 708 may remain on the wearable sensor portion 716 (such that the portable user device 718 receives the digitized glucose signal from the wearable sensor portion 716).

[0070] Figure 8 is a schematic side view of an exemplary glucose sensor 704 according to an embodiment provided herein. In some embodiments, the glucose sensor 704 can include a working electrode 802, a reference electrode 804, a counter electrode 806, and a background electrode 808. The working electrode 802 can include a conductive layer coated with a chemical that reacts with the glucose-containing solution in a redox reaction (which affects the concentration of charge carriers and the time-dependent impedance of the CGM sensor 704). In some embodiments, the working electrode 802 can be formed from platinum or platinum with a roughened surface. Other working electrode materials can be used. Exemplary chemical catalysts (e.g., enzymes) for the working electrode 802 include glucose oxidase, glucose dehydrogenase, or the like. The enzyme component can be immobilized on the electrode surface by a crosslinking agent such as, for example, glutaraldehyde. An outer membrane layer can be added over the enzyme layer to protect the overall internal components including the electrode and the enzyme layer. In some embodiments, a mediator such as ferricyanide or ferrocene can be used. Other chemical catalysts and / or mediators can be used.

[0071] In some embodiments, the reference electrode 804 can be formed from Ag / AgCl. The counter electrode 806 and / or the background electrode 808 can be formed of a suitable conductor such as platinum, gold, palladium, or the like. Other materials can be used for the reference electrode, the counter electrode, and / or the background electrode. In some embodiments, the background electrode 808 can be the same as the working electrode 802, but does not include a chemical catalyst and a mediator. The counter electrode 806 can be separated from the other electrodes by a separation layer 810 (e.g., polyimide or another suitable material).

[0072] FIG. 9 is a flowchart of an exemplary method 900 of fabricating a CGM device according to embodiments provided herein. Referring to FIG. 9, at block 902, a gain function is created based on a plurality of sensor progression parameters, such as a plurality of ratios of glucose signals (measured by a CGM sensor). The glucose signal can be an electrochemical current, an optical signal, or the like. Each sensor progression parameter (SPP) is based on the glucose signal at a point of interest and the glucose signal measured prior to the glucose signal at the point of interest. For example, the SPP ratio includes the ratio of the glucose signal at the point of interest to the glucose signal measured prior to the glucose signal at the point of interest. In some embodiments, the glucose signal measured up to 12 hours prior to the glucose signal at the point of interest can be used. Shorter or longer periods can be used. In at least some embodiments, multivariate regression or similar statistical techniques are employed using hundreds or even thousands of sensor progression parameters, such as ratios, differences, and / or other cross-stream, and / or other cross-streams, to calculate the relevant SPPs (e.g., ratios, differences, etc.), cross-stream, and coefficients employed in the gain function. A plurality of gain functions (e.g., 2, 3, 4, 5, etc.) can be determined for use during different periods of CGM monitoring. For example, one or more gain functions can be stored in the memory 712 of the wearable sensor portion 716 (FIG. 7A) or the portable user device 718 (FIG. 7B).

[0073] At block 904, the gain function is stored in the memory of the CGM device (e.g., in the form of parameter names and their coefficients). For example, one or more gain functions can be stored in the memory 712 of the wearable sensor portion 716 (FIG. 7A) or the portable user device 718 (FIG. 7B).

[0074] In block 906, computer program code is stored in the memory of the CGM device, and when the computer program code is executed by a processor, the CGM device is caused to: (a) use the sensors of the CGM device to measure a plurality of glucose signals; (b) store the glucose signals in the memory of the CGM device; (c) for the currently measured glucose signal, calculate a plurality of sensor progression parameters such as ratios, differences, etc. based on the currently measured glucose signal and the plurality of previously measured glucose signals stored in the memory; (d) use the plurality of sensor progression parameters and the stored gain function to calculate a corrected glucose value (e.g., concentration); and (e) communicate the corrected glucose value to the user of the CGM device. For example, the computer program code can be stored in the memory 712 of the wearable sensor portion 716 (FIG. 7A) or the portable user device 718 (FIG. 7B). The CGM sensor 704 can be used to measure glucose signals that can be stored in the memory 712. These stored glucose signals can be used to calculate a plurality of ratios to the currently measured glucose signal. Next, the calculated ratios can be used together with the stored gain function to calculate a corrected glucose value (e.g., using processor 710, processor 720, and / or the above equation (2)). Once calculated, the corrected glucose value can be communicated to the user (e.g., via the display 717 of the wearable sensor 716 and / or the display 722 of the portable user device 718).

[0075] FIG. 10 is a flowchart of an exemplary method 1000 for determining glucose concentration during continuous glucose monitoring (CGM) measurements according to embodiments provided herein. Referring to FIG. 10, at block 1002, a CGM device is provided that includes a sensor, a memory, and a processor. The CGM device includes one or more gain functions stored in the memory. Each gain function is based on a plurality of sensor progression parameters of the glucose signal, such as a ratio, each including a ratio of the glucose signal at a point of interest to the glucose signal measured prior to the glucose signal at the point of interest. In some embodiments, the glucose signal measured up to 12 hours prior to the glucose signal at the point of interest may be used. Shorter or longer periods can be used. For example, one or more gain functions may be stored in the memory 712 of the wearable sensor portion 716 (FIG. 7A) or the portable user device 718 (FIG. 7B). The glucose signal can be an electrochemical current, an optical signal, or the like. In any of the embodiments described herein, additional calibration information, such as in-situ and / or factory calibration data, can be stored in the CGM or other analyte monitoring device (e.g., in the memory of the device such as memory 712) for use during glucose and / or other analyte determination.

[0076] In block 1004, multiple glucose signals are measured and stored in the memory of the CGM device. For example, the wearable sensor portion 716 can be applied to the user of the CGM device 700 using a suitable inserter. During the insertion process, the CGM sensor 704 is inserted through the user's skin and contacts the interstitial fluid. The bias circuit 702 may apply a bias voltage (e.g., a continuous DC bias) to the CGM sensor 704, and then the current measurement circuit 706 can sense the current signal generated by the applied bias voltage (e.g., the operating electrode current, and in some embodiments, the background electrode current). The sample circuit 708 can digitize the sensed current signal, and the processor 710 (or processor 720 in the embodiment of FIG. 7B) can store the current signal in the memory 712.

[0077] In block 1006, multiple sensor progression parameters, such as multiple ratios, are calculated based on the currently measured glucose signal and the multiple previously measured glucose signals stored in the CGM memory. For example, each sensor progression parameter employed by the gain function stored in the memory of the CGM device can be calculated for the currently measured glucose signal (e.g., using the processor 710 or 720 and the memory 712 of FIG. 7A or FIG. 7B).

[0078] In block 1008, a corrected glucose value is calculated based on the currently measured glucose signal using the calculated sensor progression parameters and the stored gain function. As discussed, different gain functions can be used for different CGM usage periods (e.g., 3 - 21 hours, 12 - 45 hours, 40 - 167 hours, or the like), so the gain function used can depend on the period during which the currently measured glucose signal is measured. For example, the processor 710 (or processor 720) calculates the ratios and other cross terms used in the stored gain function, and the formula G Comp =Signal RawIt can be used to calculate a corrected glucose value using *Gain*(1 / (1 + Gain Function)). Alternatively, this can be considered as adjusting the system *Gain* based on the gain function and then using the adjusted gain function to calculate the corrected glucose value (e.g., concentration).

[0079] In block 1010, the corrected glucose value is communicated to the user of the CGM device. For example, display 117 (FIG. 7A) or display 722 (FIG. 7B) may display the glucose value. Alternatively, the glucose value may be used as part of a trend line, graph, or image. In some embodiments, the corrected glucose value may not be displayed until much later in the future and / or until the user requests to display the glucose value.

[0080] In some embodiments, other analytes can be measured using continuous monitoring according to the embodiments provided herein. For example, the concentration of cholesterol, lactate, uric acid, alcohol, or the like can be detected using an analyte or other biosensor, the analyte signal of interest, the previously measured analyte signal, and one or more appropriate gain functions.

[0081] Figure 11 shows an exemplary method 1100 for determining analyte concentration during continuous monitoring measurements using a biosensor subcutaneously inserted into a subject, according to an embodiment provided herein. Referring to Figure 11, the method 1100 for determining analyte concentration during continuous monitoring measurements includes subcutaneously inserting a biosensor into the subject (block 1102). In some embodiments, the biosensor can include a counter electrode, a reference electrode, and a working electrode having a chemical composition configured to oxidize the analyte of interest. For example, the CGM sensor 704 or another analyte sensor can be inserted by the user. A bias voltage, such as a constant voltage, can be applied to the working electrode to generate a continuous current from the working electrode (block 1104). At block 1106, the working electrode current signal from the working electrode can be sensed and stored in a memory (e.g., the memory 712 of the CGM device 700 or 750, or the memory of another continuous analyte monitoring device). For example, the working electrode current signal can be sensed (e.g., sampled) and stored periodically.

[0082] At block 1108, the working electrode current signal at the point of interest and a portion of the working electrode current signal stored in the memory that was measured prior to the working electrode current signal at the point of interest can be aggregated (e.g., read from the memory). A gain function value can then be generated from a predetermined gain function using the working electrode current signal at the point of interest and the portion of the working electrode current signal aggregated from the memory (block 1110). For example, a ratio and / or difference based on the working electrode current signal at the point of interest and the portion of the working electrode current signal aggregated from the memory can be used within the predetermined gain function.

[0083] The system gain can be modified using the gain function value generated from the predetermined gain function (block 1112). For glucose, the system gain can be based on in-situ calibration using the BGM glucose value. The system gain for other analytes can similarly be determined based on a reference analyte concentration and the working electrode current signal (e.g., Gain = A Calb / (Iw) or A Calb / (Iw - Ib), where A Calb is the reference analyte value). In block 1114, based on the corrected system gain and the operating electrode current signal at the point of interest, the corrected analyte concentration (A Comp ) of the operating electrode current signal at the point of interest can be determined as follows. (25) A Comp = Signal Raw *Gain*(1 / (1 + Gain Function)), where Gain*(1 / (1 + Gain Function)) represents the corrected system gain.

[0084] Exemplary analytes include glucose, cholesterol, lactate, uric acid, alcohol, or the like. In some embodiments, the background current signal is stored in the operating electrode signal in memory and can be used to remove signals from interfering substances such as vitamin C and acetaminophen.

[0085] In yet another embodiment, a continuous analyte monitoring (CAM) device may include a wearable portion having a biosensor configured to be subcutaneously inserted into a subject. The biosensor may include a counter electrode, a reference electrode, and a working electrode having a chemical composition configured to oxidize an analyte of interest and generate an analyte (e.g., glucose) signal from interstitial fluid. The wearable portion may also have a processor, a memory coupled to the processor, and a transmitter circuit (e.g., as shown in FIG. 7A for CGM device 700) coupled to the processor. The memory may include an analyte signal of interest and a predetermined gain function based on analyte signals measured prior to the analyte signal of interest. Further, the memory may include computer program code stored in the memory, and the computer program code, when executed by the processor, causes the CAM device to: (a) apply a constant voltage to a working electrode having a chemical composition, thereby generating a continuous current from the working electrode; (b) sense a working electrode current signal from the working electrode and store it in the memory; (c) aggregate the working electrode current signal of interest and a portion of the working electrode current signal stored in the memory measured prior to the working electrode current signal of interest; (d) generate a gain function value from the gain function using the working electrode current signal of interest and the portion of the working electrode current signal aggregated from the memory; (e) correct the system gain using the gain function value generated from the predetermined gain function; and (f) determine an analyte concentration of the working electrode current signal of interest based on the corrected system gain and the working electrode current signal of interest. For example, in some embodiments, the analyte concentration using the corrected system gain may be calculated using Equation (25) above.

[0086] Figure 12 shows another exemplary method 1200 of fabricating a continuous analyte monitoring device according to embodiments provided herein. Referring to Figure 12, in block 1202, an analyte sensor is operably coupled to a host. For example, the analyte sensor can be inserted through the host's skin such that the analyte sensor contacts the interstitial fluid region. Exemplary analytes that can be detected include glucose, cholesterol, lactate, uric acid, alcohol, or the like.

[0087] In block 1204, the analyte sensor is used to continuously record an analyte signal, such as at a regular sampling rate. For example, the analyte signal can be recorded every 1 minute, 2 minutes, 3 minutes, 5 minutes, 10 minutes, or any other increment (e.g., sensed, measured, sampled, and / or stored). In some embodiments, the analyte signal can be collected as part of a clinical trial using a plurality of analyte sensors and a plurality of hosts. A reference analyte concentration can also be recorded (e.g., periodically during a continuous analyte monitoring process). For example, a reference analyte device, such as a blood glucose meter, can be used to record the reference analyte concentration. The reference analyte concentration can be recorded at any suitable increment, such as hourly, daily, every other day, or the like.

[0088] In block 1206, the analyte sensor signal is paired with the reference analyte concentration. For example, an analyte signal measured after a reference analyte concentration is recorded can be paired with the recorded reference analyte concentration until another reference analyte concentration is recorded. Other pairing can be used.

[0089] FIG. 13 is a graph 1300 of an exemplary CGM response current, such as Iw-Ib, paired with a reference glucose value from a blood glucose meter (e.g., a Contour Next One® blood glucose meter) according to embodiments provided herein. As shown in FIG. 13, Iw-Ib tracks the reference glucose value. In some embodiments, a glucose signal measured 3-5 minutes before a new reference glucose value is measured is paired with the new reference glucose value rather than the previously recorded reference glucose value.

[0090] FIG. 14 is a graph 1400 of an exemplary reference glucose value versus a glucose current, such as Iw-Ib, during a CGM process according to embodiments provided herein. As shown in FIG. 14, the glucose current signal accurately tracks the reference glucose value, and the R 2 value is 0.83.

[0091] At block 1208, a relative analyte error ΔA / A is calculated. For example, the analyte error is equivalent to a signal error ΔSignal / Signal and can be calculated using the above equation (23).

[0092] At block 1210, sensor progress information is aggregated and used to calculate sensor progress parameters by referring analyte data points of interest across the dataset to previously measured analyte data points. For example, the sensor progress parameters can be calculated for each analyte signal of interest by calculating a ratio, difference, or other relationship between the analyte signal of interest and the analyte signal measured prior to the analyte signal of interest.

[0093] In block 1212, at least one of the relative analyte error referenced to the reference analyte concentration and the relative gain error referenced to the reference gain is set as a target for statistical analysis, and statistical analysis is performed by setting sensor progression parameters as input variables to obtain a gain function. In some embodiments, the statistical analysis may include multivariate regression. Other statistical analysis methods may be used. As described above, a gain function can be obtained using dozens, hundreds, or thousands of sensor progression parameters.

[0094] In block 1214, the obtained gain function is recorded as a factory calibration component for storage in a continuous analyte monitoring device. The gain function includes the selected sensor progression parameters and their weighting coefficients. As previously described with reference to FIG. 2, more than one gain function can be determined for an analyte data set (e.g., 2, 3, 4, or more). In some embodiments, the gain function can be stored in the memory 712 of the CGM device 700, or in another continuous analyte monitoring device such as a CGM device. The gain function may be based on an electrochemical signal, an optical signal, or the like.

[0095] The foregoing description discloses only exemplary embodiments. Modifications of the above-disclosed apparatus and methods within the scope of this disclosure will be readily apparent to those skilled in the art.

[0096] Appendix The gain functions, coefficients, sensor progression parameters, and / or cross terms listed below are merely exemplary. Other gain functions, coefficients, sensor progression parameters, and / or cross terms may be employed.

[0097] Gain function 1 (3 - 20 hours) Gain Function1 = 3.091118 + 1.471262 * w27min - 0.984106 * w2h - 2.665682 * b3min - 2.254352 * b18min + 2.314656 * b1h - 0.0149674 * w6mG - 0.0227495 * w24mG + 1.117723 * w27mSS1 + 0.951549 * w3hSS1 + 0.0020214 * w3hGSS1 + 0.536988 * w3m2h - 0.97541 * w9m2h + 0.326197 * w15m1h + 0.699317 * w24m2h - 0.397502 * w27m2h + 0.0287526 * b3mG - 0.01240012 * b6mG - 0.43543 * b3hSS1 - 0.0025204 * b30mGSS1 - 0.841776 * w27mb6mSS1 - 1.268666 * w3hb9mSS1...

[0098] Gain Function 2 (12 - 45 hours) Gain Function2 = 6.333784 + 2.7006532 * w9min - 0.4503101 * b1h - 0.0004858 * w9mGSS1 + 9.35e - 5 * w12hGSS1 + 0.0930561 * w3m3h - 0.0721993 * w3m12h + 0.292186 * w9m1h - 0.2503538 * w12m2h - 0.231265 * w21m2h + 0.508004 * w24m2h - 0.2291636 * w30m1h - 0.0584004 * w30m3h + 0.0409812 * w30m4h - 0.0041765 * b6mG + 0.0039682 * b12hG - 0.1346056 * b6hSS1 + 0.0010892 * b1hGSS1 + 0.0009736 * b10hGSS1 + 0.1471952 * w8hb4hSS1 - 0.0364746 * w12hb4hSS1...

[0099] Gain Function 3 (40 - 167 hours) Gain Function3 = 0.979878 - 0.637773 * w9min + 0.0001337 * w2hGSS1 + 0.0001573 * w12hGSS1 + 0.167157 * w3m3h - 0.325834 * w3m12h + 0.311399 * w6m12h + 0.378304 * w12m1h - 0.308914 * w12m3h - 0.022367 * w18m10h - 0.377033 * w27m1h + 0.167712 * w27m3h - 0.390158 * b15mSS1 + 0.356742 * b8hSS1 - 0.0008352 * b9mGSS1 + 0.0008576 * b4hGSS1 - 0.204995 * w3mb2h - 0.59261 * w6mb1h + 0.0020452 * Gw12hb12h - 0.296086 * w12hb12mSS1 + 0.289866 * w12hb12hSS1...

[0100] Sensor progress parameters and cross-stream of Gain Function 1 At the in-situ calibration point, Gain i = BGM cal-i / (Iw - Ib) cal-i where i = 1, 2,.....10, etc.

[0101] S / S1 = Gain i / Gain1, the ratio of individual Gain i and Gain1 = BGM cal-1 / (Iw - Ib) cal-1

[0102] G raw = 0.85 * (Iw - Ib)t * Gain i , if Gain i > 12; elseif Gain i > 8, 0.9 * (Iw - Ib)t * Gain i ; else (Iw - Ib)t * Gain i

[0103] w27min = (Iw - Ib) t / (Iw - Ib) t-27min , the ratio of Iw - Ib at time t and Iw - Ib at t - 27 minutes

[0104] w2h=(Iw - Ib) t / (Iw - Ib) t-2hour 、the ratio of Iw - Ib at time t to Iw - Ib at time t - 2 hours

[0105] b3min = Ib t / Ib t-3min 、the ratio of Ib at time t to Ib at time t - 3 minutes

[0106] b18min = Ib t / Ib t-18min

[0107] b1h = Ib t / Ib t-1hour

[0108] w6mG = G raw *(Iw - Ib) t / (Iw - Ib) t-6min

[0109] w24mG = G raw *(Iw - Ib) t / (Iw - Ib) t-24min

[0110] w27mSS1=(S / S1)*(Iw - Ib) t / (Iw - Ib) t-27min

[0111] w3hSS1=(S / S1)*(Iw - Ib) t / (Iw - Ib) t-3hour

[0112] w3hGSS1 = G raw *(S / S1)*(Iw - Ib) t / (Iw - Ib) t-3hour

[0113] w3m2h = w_3min / w_2h = 〔(Iw - Ib) t / (Iw - Ib) t-3min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-2hour〕=(Iw - Ib) t-2hour / (Iw - Ib) t-3min

[0114] w9m2h = w_9min / w_2h = 〔(Iw - Ib) t / (Iw - Ib) t-9min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-2hour 〕=(Iw - Ib) t-2hour / (Iw - Ib) t-9min

[0115] w15m1h = w_15min / w_1h = 〔(Iw - Ib) t / (Iw - Ib) t-15min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-1hour 〕=(Iw - Ib) t-1hour / (Iw - Ib) t-15min

[0116] w24m2h = w_24min / w_2h = 〔(Iw - Ib) t / (Iw - Ib) t-24min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-2hour 〕=(Iw - Ib) t-2hour / (Iw - Ib) t-24min

[0117] w27m2h = w_27min / w_2h = 〔(Iw - Ib) t / (Iw - Ib) t-27min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-2hour 〕=(Iw - Ib) t-2hour / (Iw - Ib) t-27min

[0118] b3mG = G raw *Ib t / Ib t-3min

[0119] b6mG = G raw *Ib t / Ib t-6min

[0120] b3hSS1 = (S / S1) * Ib t / Ib t-3hour

[0121] b30mGSS1 = G raw * (S / S1) * Ib t / Ib t-30min

[0122] w27mb6mSS1 = (S / S1) * w_27min / b_6min = (S / S1) * 〔(Iw - Ib) t / (Iw - Ib) t-27min 〕 / 〔Ib t / Ib t-6min 〕

[0123] w3hb9mSS1 = (S / S1) * w_3h / b_9min = (S / S1) * 〔(Iw - Ib) t / (Iw - Ib) t-3hour 〕 / 〔Ib t / Ib t-9min 〕

[0124] Sensor progression parameters and cross stream of gain function 2 At the in - situ calibration point, Gain i = BGM cal-i / (Iw - Ib) cal-i where i = 1, 2,.....10 etc.

[0125] S / S1 = Gain i / Gain1, individual Gain i Ratio of Gain and Gain1 = BGM cal-1 / (Iw - Ib) cal-1

[0126] G raw = 0.85 * (Iw - Ib)t * Gain i , if Gaini > 12; elseif Gain i > 8, 0.9 * (Iw - Ib)t * Gain i ; else (Iw - Ib)t * Gain i

[0127] w9min = (Iw - Ib) t / (Iw - Ib) t-9min 、the ratio of Iw - Ib at time t to Iw - Ib at time t - 9 minutes

[0128] b1h = Ib t / Ib t-1hour 、the ratio of Ib at time t to Ib at time t - 1 hour

[0129] w9mGSS1 = G raw *(S / S1)*(Iw - Ib) t / (Iw - Ib) t-9min

[0130] w12hGSS1 = G raw *(S / S1)*(Iw - Ib) t / (Iw - Ib) t-12hour

[0131] w3m3h = w_3min / w_3h = 〔(Iw - Ib) t / (Iw - Ib) t-3min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-3hour 〕=(Iw - Ib) t-3hour / (Iw - Ib) t-3min

[0132] w3m12h = w_3min / w_12h = 〔(Iw - Ib) t / (Iw - Ib) t-3min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-12hour 〕=(Iw - Ib) t-12hour / (Iw - Ib) t-3min

[0133] w9m1h = w_9min / w_1h = 〔(Iw - Ib) t / (Iw - Ib) t-9min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-1hour 〕=(Iw - Ib) t-1hour / (Iw - Ib) t-9min

[0134] w12m2h = w_12min / w_2h = 〔(Iw - Ib) t / (Iw - Ib) t-12min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-2hour 〕=(Iw - Ib) t-2hour / (Iw - Ib) t-12min

[0135] w21m2h = w_21min / w_2h = 〔(Iw - Ib) t / (Iw - Ib) t-21min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-2hour 〕=(Iw - Ib) t-2hour / (Iw - Ib) t-21min

[0136] w24m2h = w_24min / w_2h = 〔(Iw - Ib) t / (Iw - Ib) t-24min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-2hour 〕=(Iw - Ib) t-2hour / (Iw - Ib) t-24min

[0137] w30m1h = w_30min / w_1h = 〔(Iw - Ib) t / (Iw - Ib) t-30min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-1hour 〕=(Iw - Ib) t-1hour / (Iw - Ib) t-30min

[0138] w30m3h = w_30min / w_3h = 〔(Iw - Ib) t / (Iw - Ib) t-30min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-3hour 〕=(Iw - Ib) t-3hour / (Iw - Ib) t-30min

[0139] w30m4h = w_30min / w_4h = 〔(Iw - Ib) t / (Iw - Ib) t-30min〕 / 〔(Iw - Ib) t / (Iw - Ib) t-4hour 〕 = (Iw - Ib) t-4hour / (Iw - Ib) t-30min

[0140] b6mG = G raw * Ib t / Ib t-6min

[0141] b12hG = G raw * Ib t / Ib t-12hour

[0142] b6hSS1 = (S / S1) * Ib t / Ib t-6hour

[0143] b1hGSS1 = G raw *(S / S1) * Ib t / Ib t-1hour

[0144] b10hGSS1 = G raw *(S / S1) * Ib t / Ib t-10hour

[0145] w8hb4hSS1 = (S / S1) * w_8h / b_4h = (S / S1) * 〔(Iw - Ib) t / (Iw - Ib) t-8hour 〕 / 〔Ib t / Ib t-4hour 〕

[0146] w12hb4hSS1 = (S / S1) * w_12h / b_4h = (S / S1) * 〔(Iw - Ib) t / (Iw - Ib) t-12hour 〕 / 〔Ib t / Ib t-4hour 〕

[0147] Gain function 3 sensor progress parameters and cross stream At the in - situ calibration point, Gain i = BGM cal-i / (Iw - Ib)cal-i where i = 1, 2,.....10, etc.

[0148] S / S1 = Gain i / Gain1, the ratio of each Gain i to Gain1 = BGM cal-1 / (Iw - Ib) cal-1

[0149] G raw = 0.85*(Iw - Ib)t*Gain i , if Gain i > 12; elseif Gain i > 8, 0.9*(Iw - Ib)t*Gain i ; else (Iw - Ib)t*Gain i

[0150] w9min=(Iw - Ib) t / (Iw - Ib) t-9min , the ratio of Iw - Ib at time t to Iw - Ib at time t - 9 minutes

[0151] w2hGSS1 = G raw *(S / S1)*(Iw - Ib) t / (Iw - Ib) t-2hour

[0152] w12hGSS1 = G raw *(S / S1)*(Iw - Ib) t / (Iw - Ib) t-12hour

[0153] w3m3h = w_3min / w_3h = 〔(Iw - Ib) t / (Iw - Ib) t-3min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-3hour 〕=(Iw - Ib) t-3hour / (Iw - Ib) t-3min

[0154] w3m12h = w_3min / w_12h = 〔(Iw - Ib) t / (Iw - Ib) t-3min 〕 / 〔(Iw - Ib)t / (Iw - Ib) t-12hour 〕 = (Iw - Ib) t-12hour / (Iw - Ib) t-3min

[0155] w6m12h = w_6min / w_12h = 〔(Iw - Ib) t / (Iw - Ib) t-6min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-12hour 〕 = (Iw - Ib) t-12hour / (Iw - Ib) t-6min

[0156] w12m1h = w_12min / w_1h = 〔(Iw - Ib) t / (Iw - Ib) t-12min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-1hour 〕 = (Iw - Ib) t-1hour / (Iw - Ib) t-12min

[0157] w12m3h = w_12min / w_3h = 〔(Iw - Ib) t / (Iw - Ib) t-12min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-3hour 〕 = (Iw - Ib) t-3hour / (Iw - Ib) t-12min

[0158] w18m10h = w_18min / w_10h = 〔(Iw - Ib) t / (Iw - Ib) t-18min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-10hour 〕 = (Iw - Ib) t-10hour / (Iw - Ib) t-18min

[0159] w27m1h = w_27min / w_1h = 〔(Iw - Ib) t / (Iw - Ib) t-27min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-1hour 〕 = (Iw - Ib) t-1hour / (Iw - Ib) t-27min

[0160] w27m3h = w_27min / w_3h = 〔(Iw - Ib) t / (Iw - Ib) t-27min 〕 / 〔(Iw - Ib) t / (Iw - Ib) t-3hour 〕=(Iw - Ib) t-3hour / (Iw - Ib) t-27min

[0161] b15mSS1=(S / S1)*Ib t / Ib t-15min

[0162] b8hSS1=(S / S1)*Ib t / Ib t-8hour

[0163] b9mGSS1 = G raw *(S / S1)*Ib t / Ib t-9min

[0164] b4hGSS1 = G raw *(S / S1)*Ib t / Ib t-4hour

[0165] w3mb2h = w_3min / b_2h = 〔(Iw - Ib) t / (Iw - Ib) t-3min 〕 / 〔Ib t / Ib t-2hour 〕

[0166] w6mb1h = w_6min / b_1h = 〔(Iw - Ib) t / (Iw - Ib) t-6min 〕 / 〔Ib t / Ib t-1hour 〕

[0167] Gw12hb12h = G raw *w_12h / b_12h = G raw *〔(Iw - Ib) t / (Iw - Ib) t-12hour 〕 / 〔Ib t / Ib t-12hour 〕

[0168] w12hb12mSS1=(S / S1)*w_12h / b_12min=(S / S1)*〔(Iw - Ib) t / (Iw - Ib) t-12hour 〕 / 〔Ib t / Ib t-12min 〕

[0169] w12hb12hSS1=(S / S1)*w_12h / b_12h=(S / S1)*〔(Iw - Ib) t / (Iw - Ib) t-12hour 〕 / 〔Ib t / Ib t-12hour 〕

Claims

1. A continuous analyte monitoring (CAM) device including a wearable portion comprising a biosensor, a processor, and a memory, wherein the biosensor comprises a counter electrode, a reference electrode, a background electrode, and a working electrode having a chemical composition, and the chemical composition is configured to oxidize an analyte of interest to generate an analyte signal from interstitial fluid, the memory is connected to the processor, and the memory includes a predetermined gain function for an analyte signal of interest based on an analyte signal measured prior to the analyte signal of interest, when the memory is executed by the processor, the CAM device performs the following processes by computer program code recorded in the memory: Process a: applying a constant voltage to the working electrode having the chemical composition to generate a continuous current from the working electrode; Process b: sensing a working electrode current signal from the working electrode; Process c: sensing a background electrode current signal from the background electrode; Process d: generating a gain function value from the predetermined gain function using the working electrode current signal at the point of interest, the background electrode current signal at the point of interest, a portion of the working electrode current signal, and a portion of the background electrode current signal; Process e: correcting the system gain using the gain function value generated from the predetermined gain function; Process f: determining the analyte concentration of the working electrode current signal at the point of interest by applying the system gain to the working electrode current signal at the point of interest; and Process g: transmitting the analyte concentration to the user A configured analyte monitoring (CAM) device.

2. The CAM device according to claim 1, wherein the predetermined gain function is a factor determined by multivariate regression.

3. The CAM device according to claim 1, wherein the gain function value is based on a working electrode current signal measured at least 1 hour prior to the working electrode current signal at the point of interest.

4. The CAM device according to claim 1, wherein the gain function value is based on a working electrode current signal measured 6 to 12 hours prior to the working electrode current signal at the point of interest.

5. The CAM device according to claim 1, wherein the predetermined gain function uses a ratio of the working electrode signal at the point of interest to a previously measured working electrode.

6. The CAM device further comprises a transmitter circuit connected to the processor, The CAM device according to claim 1, wherein the process g of transmitting the analyte concentration to the user in the CAM device includes the process of transmitting the analyte concentration to the portable user device using the transmitter circuit.

7. The CAM device further includes a display for displaying information to the user, The CAM device according to claim 1, wherein the process g of transmitting the analyte concentration to the user in the CAM device includes the process of displaying the analyte concentration to the user using the display.

8. A part of the working electrode current signal is measured before the working electrode current signal at the point of interest, A part of the background electrode current signal is measured before the background electrode current signal at the point of interest The CAM device according to claim 1.

9. The CAM device according to claim 1, wherein the process d of generating the gain function value includes the process of determining by a ratio based on the working electrode current signal at the point of interest and a part of the working electrode current signal.

10. The CAM device according to claim 1, wherein the process d of generating the gain function value includes the process of determining by the difference between the working electrode current signal at the point of interest and the background electrode current signal at the point of interest.

Images