Electrochemical sensor parameter extraction method and system based on multi-dimensional steady-state analysis

Abstract

The application relates to the technical field of electrochemical sensor parameter extraction, and provides an electrochemical sensor parameter extraction method and system based on multidimensional steady-state analysis. The method receives an original electrochemical response signal sequence output by a continuous glucose monitoring device, performs polarization steady-state fitting processing on an initial segment of the filtered signal sequence, predicts a steady-state response value, identifies a steady-state response interval in the signal sequence, establishes a corresponding relationship between a concentration gradient and a signal interval, infers a reasonable value range and performs data compensation on a steady-state segment corresponding to a concentration point that is not effectively identified, and obtains a complete steady-state segment set; each segment in the steady-state segment set is quantitatively scored, and high-quality steady-state segment data that meets a preset screening condition in a comprehensive score is screened out in the steady-state segment set; curve fitting operation is performed based on the high-quality steady-state segment data obtained through screening, and a sensitivity parameter of the electrochemical sensor and a corresponding quality evaluation report are output.

Description

Technical Field

[0001] This application relates to the field of electrochemical sensor parameter extraction technology, and in particular to an electrochemical sensor parameter extraction method and system based on multidimensional steady-state analysis. Background Technology

[0002] Continuous glucose monitoring (CGM) technology uses subcutaneous electrochemical sensors to monitor interstitial fluid glucose concentration in real time. Sensitivity parameters are key indicators characterizing the sensor's core performance. During sensor research and development, standardized testing of numerous sensors is necessary to extract sensitivity parameters. Existing parameter extraction methods suffer from the following drawbacks: First, they are susceptible to noise and transient interference, resulting in low accuracy in identifying steady-state response ranges; second, they cannot effectively address response delays caused by electrode polarization, leading to high testing time costs; third, they employ fixed time windows or simple threshold segmentation, failing to adapt to signal characteristics under different testing conditions and concentration gradients; and fourth, they lack unified evaluation standards, resulting in significant deviations and poor consistency in sensitivity parameters obtained through different operating methods.

[0003] Therefore, a method is urgently needed to solve at least one of the above problems. Summary of the Invention

[0004] This application provides a method and system for extracting electrochemical sensor parameters based on multidimensional steady-state analysis. It aims to address the challenge of continuous glucose monitoring technology, which uses subcutaneous electrochemical sensors to monitor interstitial fluid glucose concentration in real time. Sensitivity parameters are key indicators characterizing the sensor's core performance. During sensor research and development, the process of extracting sensitivity parameters requires standardized testing of a large number of sensors.

[0005] In a first aspect, embodiments of this application provide a method for extracting electrochemical sensor parameters based on multi-dimensional steady-state analysis, the method comprising: Receive the raw electrochemical response signal sequence output by the continuous glucose monitoring device; perform time-domain filtering, edge-preserving filtering and smoothing filtering on the raw electrochemical response signal sequence in sequence to obtain the filtered signal sequence; Polarization steady-state fitting is performed on the initial segment of the filtered signal sequence to predict the steady-state response value. Based on the steady-state response value, the corresponding signal change characteristics are determined, and a quality assessment report of the electrochemical sensor is output according to the sensitivity parameters. The steady-state response interval in the signal sequence is identified based on the signal change characteristics. The identified steady-state response interval is numerically segmented to establish the correspondence between the concentration gradient and the signal interval. For the steady-state segments corresponding to concentration points that are not effectively identified, the reasonable value range is inferred using the continuity characteristics of the concentration response and data compensation is performed to obtain a complete set of steady-state segments. Each segment in the steady-state segment set is quantitatively scored, and high-quality steady-state segment data that meet the preset screening conditions are selected from the steady-state segment set. Based on the selected high-quality steady-state segment data, curve fitting operation is performed to extract the sensitivity parameters of the electrochemical sensor. Based on the sensitivity parameters, a quality assessment report of the electrochemical sensor is output.

[0006] In some embodiments, performing time-domain filtering, edge-preserving filtering, and smoothing filtering sequentially on the original electrochemical response signal sequence to obtain a filtered signal sequence includes: performing an upsampling operation on the original electrochemical response signal sequence to unify the overall data sampling frequency; performing time-domain filtering on the signal sequence after unifying the sampling frequency to remove pulse interference signals; performing edge-preserving filtering on the time-domain filtered signal sequence to retain signal mutation characteristics; performing smoothing filtering on the edge-preserving filtered signal sequence to remove high-frequency random noise; and outputting the processed filtered signal sequence.

[0007] In some embodiments, performing polarization steady-state fitting processing on the initial segment of the filtered signal sequence to predict the steady-state response value includes: extracting continuous signal data from the initial period of the filtered signal sequence; performing fitting operations on the extracted initial period data using a multi-exponential decay mode to solve for the decay parameter and steady-state offset; and determining the solved steady-state offset as the predicted steady-state response value.

[0008] In some embodiments, determining the corresponding signal change characteristics based on the steady-state response value, outputting a quality assessment report of the electrochemical sensor according to the sensitivity parameters, and identifying the steady-state response interval in the signal sequence based on the signal change characteristics include: setting a maximum signal change threshold and a signal variance threshold based on the steady-state response value; traversing the filtered signal sequence using a fixed-size sliding window; calculating the maximum signal change and signal variance within each sliding window; determining the sliding window that simultaneously satisfies both threshold constraints as a steady-state window; and merging all continuously distributed steady-state windows to obtain a complete steady-state response interval.

[0009] In some embodiments, the step of numerically segmenting the identified steady-state response interval and establishing the correspondence between the concentration gradient and the signal interval includes: calculating the average value of all signal sampling points within a single steady-state response interval; calling a pre-trained current concentration mapping model to match the average value of each steady-state response interval to the glucose concentration gradient point; and establishing a one-to-one binding association between the concentration gradient point and the steady-state signal interval.

[0010] In some embodiments, the step of inferring a reasonable range of values ​​and performing data compensation based on the continuity characteristics of the concentration response to obtain a complete set of steady-state segments for the steady-state segments corresponding to concentration points that are not effectively identified includes: retrieving all glucose concentration gradient points in the unmatched steady-state response intervals; retrieving the average signal values ​​of adjacent and matched concentration points on both sides of the unmatched point; deducing the steady-state signal values ​​of the unmatched points based on the continuous change law of the concentration response; and supplementing the deduced values ​​into the corresponding missing steady-state segment data to form a complete set of steady-state segments.

[0011] In some embodiments, the step of quantitatively scoring each segment in the steady-state segment set and selecting high-quality steady-state segment data whose comprehensive scores meet preset screening conditions includes: independently scoring each segment in the steady-state segment set from three dimensions, including signal stability, segment consistency, and interval length; configuring fixed weight coefficients for the individual scores of the three dimensions, multiplying the individual scores by the corresponding weight coefficients and summing them to obtain the comprehensive score of the segment, and selecting segments with comprehensive scores higher than the preset score standard as high-quality steady-state segment data.

[0012] In some embodiments, the step of performing curve fitting operation based on the high-quality steady-state segment data obtained by screening to extract the sensitivity parameters of the electrochemical sensor includes: extracting the average steady-state current value corresponding to each glucose concentration point in the high-quality steady-state segment data, and constructing a fitting dataset with the glucose concentration and the corresponding average steady-state current value; performing a linear fitting operation on the fitting dataset, and determining the proportional coefficient of the fitting straight line as the sensitivity parameters of the electrochemical sensor.

[0013] In some embodiments, the step of outputting a quality assessment report for the electrochemical sensor based on the sensitivity parameters includes: comparing the extracted sensitivity parameters with a preset qualified parameter range to determine parameter compliance; generating data quality evaluation content by combining the comprehensive scoring results of each steady-state segment; and integrating the sensitivity parameters, parameter compliance determination results, and data quality evaluation content to generate a standardized document and output it externally.

[0014] Secondly, this application provides an electrochemical sensor parameter extraction system based on multi-dimensional steady-state analysis, the system comprising: The sequence receiving unit is used to receive the raw electrochemical response signal sequence output by the continuous glucose monitoring device; and to perform time-domain filtering, edge-preserving filtering and smoothing filtering on the raw electrochemical response signal sequence in sequence to obtain the filtered signal sequence. The polarization fitting unit is used to perform polarization steady-state fitting on the initial segment of the filtered signal sequence to predict the steady-state response value. The steady-state identification unit is used to determine the corresponding signal change characteristics based on the steady-state response value, output a quality assessment report of the electrochemical sensor according to the sensitivity parameters, and identify the steady-state response interval in the signal sequence according to the signal change characteristics. The segmented compensation unit is used to numerically segment the identified steady-state response interval, establish the correspondence between the concentration gradient and the signal interval, and for the steady-state segment corresponding to the concentration point that was not effectively identified, infer the reasonable value range and perform data compensation using the continuity characteristics of the concentration response to obtain a complete set of steady-state segments. The quantitative scoring unit is used to quantitatively score each segment in the steady-state segment set, and to select high-quality steady-state segment data that meet the preset screening conditions in the steady-state segment set; to perform curve fitting operation based on the selected high-quality steady-state segment data, and to extract the sensitivity parameters of the electrochemical sensor; and to output a quality assessment report of the electrochemical sensor based on the sensitivity parameters.

[0015] This application uses polarization steady-state fitting to predict steady-state response values ​​in advance, effectively solving the response delay problem caused by electrode polarization and significantly reducing testing time costs. It employs multi-level filtering combined with adaptive steady-state identification technology to greatly improve the accuracy of steady-state response range identification. The introduction of intelligent segmentation and data compensation mechanisms adapts to signal changes under different testing conditions and concentration gradients, addressing the problem of a single segmentation strategy. High-quality data is screened through multi-dimensional quantitative scoring, establishing a unified parameter extraction evaluation standard, significantly improving the consistency and reliability of sensitivity parameter extraction. Finally, it achieves fully automated processing from raw signal reception to parameter output, greatly improving the testing efficiency of continuous glucose monitoring sensors.

[0016] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic flowchart illustrating the steps of an electrochemical sensor parameter extraction method based on multidimensional steady-state analysis provided in an embodiment of this application; Figure 2 This is a schematic diagram illustrating the principle of an electrochemical sensor parameter extraction method based on multidimensional steady-state analysis provided in an embodiment of this application; Figure 3This is a flowchart illustrating the overall architecture of an electrochemical sensor parameter extraction method based on multidimensional steady-state analysis, provided in one embodiment of this application. Figure 4 This is a schematic block diagram of an electrochemical sensor parameter extraction system based on multidimensional steady-state analysis provided in an embodiment of this application; Figure 5 This is a schematic block diagram of the structure of a computer device provided in an embodiment of this application.

[0019] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it necessarily have to be performed in the order described. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change depending on the actual situation.

[0022] It should be understood that, in order to clearly describe the technical solutions of the embodiments of the present invention, the terms "first" and "second" are used in the embodiments of the present invention to distinguish identical or similar items with essentially the same function and effect. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and the terms "first" and "second" are not necessarily different.

[0023] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of the application. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0024] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0025] Continuous glucose monitoring (CGM) technology uses subcutaneous electrochemical sensors to monitor interstitial fluid glucose concentration in real time. Sensitivity parameters are key indicators characterizing the sensor's core performance. During sensor research and development, standardized testing of numerous sensors is necessary to extract sensitivity parameters. Existing parameter extraction methods suffer from the following drawbacks: First, they are susceptible to noise and transient interference, resulting in low accuracy in identifying steady-state response ranges; second, they cannot effectively address response delays caused by electrode polarization, leading to high testing time costs; third, they employ fixed time windows or simple threshold segmentation, failing to adapt to signal characteristics under different testing conditions and concentration gradients; and fourth, they lack unified evaluation standards, resulting in significant deviations and poor consistency in sensitivity parameters obtained through different operating methods.

[0026] Please refer to Figure 1 This invention relates to the field of electrochemical sensor parameter extraction technology, specifically a method for extracting sensitivity parameters for a continuous glucose monitoring sensor. This method is executed by a computer device, which can be deployed on any hardware platform with data processing capabilities, such as a single server, server cluster, handheld terminal, laptop, wearable device, or robot. All user data involved in this method is collected and processed only with the explicit authorization of the relevant users and in compliance with national data security laws and regulations, strictly protecting user privacy.

[0027] like Figure 1 As shown, the electrochemical sensor parameter extraction method based on multi-dimensional steady-state analysis provided by this invention is generally divided into three core steps: signal preprocessing and filtering step S101, polarization fitting and steady-state segmentation compensation step S102, and multi-dimensional scoring and parameter extraction output step S103. Figure 2 As shown, the application scenario of this method includes three parts: a continuous glucose monitoring sensor, a portable testing device, and a computer processing terminal. The continuous glucose monitoring sensor is implanted in the subcutaneous tissue of the human body to collect electrochemical signals corresponding to the glucose concentration in the interstitial fluid in real time. The portable testing device receives the raw signal output by the sensor and performs preliminary analog-to-digital conversion. The computer processing terminal receives the raw electrochemical signal through wired or wireless means, executes all the data processing procedures described in this invention, and finally outputs sensitivity parameters and quality assessment reports.

[0028] The provided method for extracting electrochemical sensor parameters based on multidimensional steady-state analysis includes steps S101 to S103. Details are as follows: Step S101. Receive the original electrochemical response signal sequence output by the continuous glucose monitoring device; perform time-domain filtering, edge-preserving filtering and smoothing filtering on the original electrochemical response signal sequence in sequence to obtain the filtered signal sequence.

[0029] Specifically, the core objective of this step is to eliminate various interferences in the raw electrochemical signal, providing high-quality foundational data for subsequent processing. The raw electrochemical response signal sequence is output from a continuous glucose monitoring device and typically contains pulse interference, high-frequency random noise, and signal abrupt changes caused by factors such as sensor electrode contact, human movement, and the electromagnetic environment. This step employs a three-stage cascaded filtering architecture, processing the signal sequentially in the order of time-domain filtering, edge-preserving filtering, and smoothing filtering, effectively removing noise while preserving the signal's effective characteristics to the greatest extent possible.

[0030] First, the raw electrochemical response signal sequence output from the continuous glucose monitoring device is received. This sequence is a discrete-time series, with each sampling point containing a timestamp and the corresponding current response value. Then, a uniform preprocessing operation is performed on the raw signal sequence, converting the device output data from different sampling frequencies into a signal sequence of a standard frequency to ensure consistency in subsequent processing. Next, a three-stage filtering process is executed sequentially: the first stage, time-domain filtering, removes sudden pulse interference signals, which typically manifest as anomalous jumps in one or a few sampling points; the second stage, edge-preserving filtering, retains the true abrupt edges caused by changes in glucose concentration, preventing subsequent smoothing filtering from eliminating effective signal features; the third stage, smoothing filtering, removes residual high-frequency random noise, making the signal curve smoother and facilitating subsequent steady-state identification and fitting calculations. Finally, the filtered signal sequence is output.

[0031] Step S102. Perform polarization steady-state fitting processing on the initial segment of the filtered signal sequence to predict the steady-state response value; determine the corresponding signal change characteristics based on the steady-state response value, output the quality assessment report of the electrochemical sensor according to the sensitivity parameters, identify the steady-state response interval in the signal sequence according to the signal change characteristics; perform numerical segmentation on the identified steady-state response interval, establish the correspondence between the concentration gradient and the signal interval, and for the steady-state segments corresponding to concentration points that are not effectively identified, infer the reasonable value range using the continuity characteristics of the concentration response and perform data compensation to obtain a complete set of steady-state segments.

[0032] Specifically, this step is one of the core innovations of this invention, primarily addressing the problems of long testing times and inaccurate steady-state identification caused by electrode polarization in existing technologies. After an electrochemical sensor is implanted in the human body or comes into contact with a test solution, polarization occurs on the electrode surface, causing the response signal to take a considerable amount of time to reach a true steady state. Traditional methods require waiting for the signal to fully stabilize before parameter extraction, significantly increasing testing time costs. This step, by performing polarization steady-state fitting on the initial segment of the signal, predicts the final steady-state response value in advance, allowing subsequent processing to proceed without waiting for the signal to fully stabilize, thus significantly shortening the testing time.

[0033] First, polarization steady-state fitting is performed on the initial segment of the filtered signal sequence. Since the electrode polarization process follows a multi-exponential decay law, a multi-exponential decay model is used to fit the initial segment of the signal, obtaining the model parameters and predicting the steady-state response value. Then, based on the predicted steady-state response value, the corresponding signal change characteristics are determined, and an adaptive steady-state identification threshold is set. A sliding window is used to traverse the entire filtered signal sequence, identifying all steady-state response intervals based on the signal change characteristics within the window.

[0034] All identified steady-state response intervals are numerically segmented, and characteristic values ​​for each interval are calculated. These characteristic values ​​are then matched with a pre-defined glucose concentration gradient to establish a one-to-one correspondence between the concentration gradient and the signal interval. Due to noise interference or fluctuations in test conditions, some steady-state segments corresponding to certain concentration points may not be effectively identified. In such cases, the continuity between glucose concentration and the response signal is utilized to infer the reasonable range of values ​​for the missing concentration points based on the signal values ​​of adjacent identified concentration points. Data compensation is then performed to ultimately obtain a complete set of steady-state segments.

[0035] Step S103. Quantify and score each segment in the steady-state segment set, and select high-quality steady-state segment data that meet the preset screening conditions in the steady-state segment set; perform curve fitting operation based on the selected high-quality steady-state segment data to extract the sensitivity parameters of the electrochemical sensor; output the quality assessment report of the electrochemical sensor according to the sensitivity parameters.

[0036] Specifically, the core objective of this step is to establish a unified parameter extraction evaluation standard to ensure high consistency and reliability of sensitivity parameters obtained from different batches and by different operators. Traditional methods lack an effective data quality evaluation mechanism and directly use all identified steady-state data for fitting, making them susceptible to the influence of low-quality data and resulting in significant deviations in parameter extraction results. This step uses multi-dimensional quantitative scoring to screen out high-quality steady-state data, and then performs curve fitting based on this data, significantly improving the accuracy of parameter extraction.

[0037] First, each segment in the steady-state segment set is quantitatively scored across three dimensions: signal stability, segment consistency, and interval length. Signal stability reflects the degree of signal fluctuation within the steady-state segment; lower fluctuation results in a higher score. Segment consistency reflects whether the signal change trend of the segment matches the concentration response pattern with adjacent segments; better consistency results in a higher score. Interval length reflects the duration of the steady-state segment; longer duration indicates more reliable data and a higher score. Weighting coefficients are assigned to each of the three dimensions, and a weighted sum is obtained for each segment to obtain a comprehensive score. Segments with comprehensive scores higher than a preset standard are selected as high-quality steady-state segment data.

[0038] Curve fitting was performed based on the high-quality steady-state data obtained through screening. Since the response signal of the continuous glucose monitoring sensor has a linear relationship with glucose concentration, a linear fitting method was used, with glucose concentration as the abscissa and the corresponding steady-state current response value as the ordinate, to obtain a concentration-current response curve. The slope of this curve is the sensitivity parameter of the electrochemical sensor, describing the quantitative relationship between the sensor's output electrical signal and the glucose concentration.

[0039] Finally, a quality assessment report is generated and output based on the extracted sensitivity parameters. The report includes the specific values ​​of the sensitivity parameters, the determination results of whether the parameters meet the preset qualification standards, the comprehensive score of each steady-state segment, and the overall data quality evaluation. The quality assessment report provides comprehensive and objective data support for sensor research and development, production, and quality control.

[0040] In some embodiments, such as Figure 3 As shown, the technical solution of the present invention includes the following core modules, which work together to achieve automated extraction from raw electrochemical signals to sensitivity parameters.

[0041] Signal preprocessing module: configured to receive the raw electrochemical response signal, perform upsampling and multi-stage filtering to unify the data frequency and eliminate noise interference, and output the filtered signal sequence.

[0042] ; in: Represents the time-domain filtering operator. This indicates the edge-preserving filter operator. This represents the smoothing filter operator.

[0043] Polarization steady-state fitting module: configured to fit the response signal after electrode polarization, predict the steady-state response value, reduce testing time cost, and improve recognition accuracy.

[0044] ; in, , For attenuation parameters, The steady-state offset is... Number of index terms.

[0045] Steady-state identification module: configured to adaptively identify steady-state intervals based on signal change characteristics, and adopt differentiated identification strategies for different concentration ranges.

[0046] ; in For indicator functions, The maximum change within the window. The variance within the window. and This is an adaptive threshold parameter.

[0047] Intelligent segmentation module: Configured to automatically segment steady-state segments based on signal numerical characteristics, establish the correspondence between concentration gradient and signal interval, and when steady-state segments of certain concentration points are not effectively identified, the system performs intelligent compensation based on the parameters of the identified concentration points, and infers the reasonable value range of the missing points by utilizing the continuity characteristics of the concentration response.

[0048] ; in, For the j-th concentration point, denoted as the corresponding signal interval, and N[] represents the pre-trained current-concentration correspondence model.

[0049] Multi-dimensional scoring module: Configured to quantitatively score each segment from multiple dimensions such as stability, consistency, and reliability, and output a comprehensive score result. For the first... Each steady-state segment Define the comprehensive scoring function: ; in, For stability rating, For consistency scoring, Rate the length. ;, , Weighting coefficients Fitting calculation module: Configured to perform curve fitting calculation based on steady-state data after score filtering, and output sensitivity parameters.

[0050] ; in, For steady-state current response, This refers to the glucose concentration. This is a sensitivity function.

[0051] The data processing flow includes: (1) Receive the raw electrochemical response signal sequence output by the CGM device; (2) Perform multi-stage filtering on the original signal to eliminate high-frequency noise and pulse interference; (3) Perform polarization steady-state fitting on the filtered initial segment signal to predict the steady-state response value and reduce test time cost; (4) Adaptive identification of steady-state response intervals based on signal characteristics; (5) Perform intelligent compensation and numerical segmentation on the identified steady-state segments to establish a correspondence with the concentration gradient; (6) Perform multi-dimensional scoring on each segment to filter high-quality data; (7) Perform curve fitting based on the filtered data to extract sensitivity parameters; (8) Output sensitivity parameters and quality assessment report.

[0052] In some embodiments, performing time-domain filtering, edge-preserving filtering, and smoothing filtering sequentially on the original electrochemical response signal sequence to obtain a filtered signal sequence includes: performing an upsampling operation on the original electrochemical response signal sequence to unify the overall data sampling frequency; performing time-domain filtering on the signal sequence after unifying the sampling frequency to remove pulse interference signals; performing edge-preserving filtering on the time-domain filtered signal sequence to retain signal mutation characteristics; performing smoothing filtering on the edge-preserving filtered signal sequence to remove high-frequency random noise; and outputting the processed filtered signal sequence.

[0053] This embodiment details the specific operation flow of the multi-stage signal filtering process in step S101, including: 1. Upsampling at a uniform frequency: The raw electrochemical response signal sequence output from the continuous glucose monitoring device is received. Different devices may have different sampling frequencies; for example, some devices use a sampling frequency of 0.5Hz, while others use 1Hz. A linear upsampling operation is performed on the raw signal sequence to uniformly convert all signal sequences to a standard sampling frequency of 1Hz, i.e., one sampling point per second. During the upsampling process, a linear interpolation method is used to calculate the values ​​of the newly added sampling points to ensure signal continuity.

[0054] 2. Time-Domain Filtering for Impulse Interference Removal: Median time-domain filtering is performed on the signal sequence after uniform sampling frequency, with the filtering window size set to 5 sampling points. For each sampling point, the values ​​of the two sampling points before and after it (a total of 5 sampling points) are taken. These 5 values ​​are then sorted by size, and the median value is taken as the filtered value for that sampling point. Median filtering can effectively remove impulse interference from single or a few sampling points without excessively smoothing the signal's edge features.

[0055] 3. Edge-preserving filtering retains abrupt changes: Bilateral edge-preserving filtering is performed on the time-domain filtered signal sequence. Bilateral filtering considers both spatial proximity and numerical similarity of the signal. For regions with large signal numerical changes (i.e., signal abrupt change edges), the filter weight is reduced to preserve edge features; for regions with small signal numerical changes, the filter weight is increased to remove noise. The filter parameters are set as follows: spatial domain standard deviation σs = 2, numerical domain standard deviation σr = 0.1 times the full-scale signal range.

[0056] 4. Smoothing Filtering to Remove High-Frequency Noise: A moving average smoothing filter is applied to the edge-preserving filtered signal sequence, with the filter window size set to 3 sampling points. For each sampling point, the average value of the three sampling points (the point itself and the one sampling point before and after it) is taken as the smoothed value for that sampling point. Moving average filtering effectively removes residual high-frequency random noise, making the signal curve smoother.

[0057] 5. Output the filtered signal sequence: Output the signal sequence after the above four-stage processing as the filtered signal sequence for use in subsequent steps.

[0058] In some embodiments, performing polarization steady-state fitting processing on the initial segment of the filtered signal sequence to predict the steady-state response value includes: extracting continuous signal data from the initial period of the filtered signal sequence; performing fitting operations on the extracted initial period data using a multi-exponential decay mode to solve for the decay parameter and steady-state offset; and determining the solved steady-state offset as the predicted steady-state response value.

[0059] This embodiment details the specific operation process of polarization steady-state fitting in step S102, including: 1. Initial Segment Data Extraction: Extract continuous signal data from the initial period of the filtered signal sequence, with a extraction length of 60 sampling points, corresponding to 60 seconds of signal data. This length covers the main stages of the electrode polarization process without introducing excessive steady-state data that could affect the fitting accuracy.

[0060] 2. Fitting with a Multi-Exponential Decay Model: A multi-exponential decay model was used to fit the initial 60-second data segment. This model accurately describes the polarization process of the electrochemical sensor, containing two decay terms with different time constants and a steady-state offset term. The least squares method was used for fitting, with 1000 iterations and a convergence accuracy of 1e-6.

[0061] 3. Solving for model parameters: After the fitting operation is completed, two attenuation parameters and one steady-state offset are obtained. The two attenuation parameters correspond to the time constants of the fast polarization process and the slow polarization process, respectively, and the steady-state offset is the response value when the signal finally reaches a steady state.

[0062] 4. Determine the predicted steady-state response value: The steady-state offset obtained from the solution is determined as the predicted steady-state response value. Using this method, the final steady-state response value can be accurately predicted with only 60 seconds of initial data, without waiting for the signal to fully stabilize (which typically takes 5-10 minutes), significantly reducing testing time costs.

[0063] In some embodiments, determining the corresponding signal change characteristics based on the steady-state response value, outputting a quality assessment report of the electrochemical sensor according to the sensitivity parameters, and identifying the steady-state response interval in the signal sequence based on the signal change characteristics include: setting a maximum signal change threshold and a signal variance threshold based on the steady-state response value; traversing the filtered signal sequence using a fixed-size sliding window; calculating the maximum signal change and signal variance within each sliding window; determining the sliding window that simultaneously satisfies both threshold constraints as a steady-state window; and merging all continuously distributed steady-state windows to obtain a complete steady-state response interval.

[0064] This embodiment details the specific operation process of adaptive steady-state response interval identification in step S102, including: 1. Set adaptive thresholds: Based on the predicted steady-state response value, set the maximum signal change threshold and the signal variance threshold. The maximum signal change threshold is set to 1% of the predicted steady-state response value, and the signal variance threshold is set to 0.01% of the square of the predicted steady-state response value. Using adaptive thresholds can adapt to the signal characteristics of sensors with different sensitivity, avoiding recognition errors caused by fixed thresholds.

[0065] 2. Sliding window traversal of the signal: A fixed-size sliding window is used to traverse the entire filtered signal sequence. The sliding window size is set to 10 sampling points, corresponding to 10 seconds of signal data, and the sliding step size is set to 1 sampling point.

[0066] 3. Calculate window eigenvalues: For each sliding window, calculate the maximum change and signal variance of all signal sampling points within the window. The maximum change is the difference between the maximum and minimum signal values ​​within the window, and the signal variance is the variance of all signal sampling points within the window.

[0067] 4. Determining a steady-state window: A sliding window that simultaneously meets the following two conditions is determined to be a steady-state window: the maximum change in the signal within the window is less than the maximum change threshold, and the variance of the signal within the window is less than the variance threshold.

[0068] 5. Merging Steady-State Windows: Merge all continuously distributed steady-state windows to obtain a complete steady-state response interval. If the interval between two steady-state windows is less than or equal to 5 sampling points, then these two steady-state windows are merged into a single complete steady-state response interval.

[0069] In some embodiments, the step of numerically segmenting the identified steady-state response interval and establishing the correspondence between the concentration gradient and the signal interval includes: calculating the average value of all signal sampling points within a single steady-state response interval; calling a pre-trained current concentration mapping model to match the average value of each steady-state response interval to the glucose concentration gradient point; and establishing a one-to-one binding association between the concentration gradient point and the steady-state signal interval.

[0070] This embodiment details the specific operational process for establishing the correspondence between numerical segments and concentrations in step S102, including: 1. Calculate the average value of the steady-state interval: For each identified steady-state response interval, calculate the average value of all signal sampling points within the interval, which is used as the characteristic value of the steady-state interval.

[0071] 2. Call the pre-trained model: Call the pre-trained current concentration mapping model. This model was trained using a large amount of historical test data. The input is the steady-state current response value, and the output is the corresponding glucose concentration value. The model uses a linear regression model, and the training data includes test results from different batches of sensors with different sensitivities.

[0072] 3. Matching Concentration Gradient Points: Input the average value of each steady-state response interval into the current concentration mapping model to obtain the corresponding predicted glucose concentration value. Match this predicted value with pre-set glucose concentration gradient points to find the closest concentration gradient point.

[0073] 4. Establish binding associations: Establish a one-to-one binding association between each glucose concentration gradient point and its corresponding steady-state response interval. If multiple steady-state intervals match the same concentration gradient point, select the steady-state interval with the highest comprehensive score as the corresponding data for that concentration point.

[0074] In some embodiments, the step of inferring a reasonable range of values ​​and performing data compensation based on the continuity characteristics of the concentration response to obtain a complete set of steady-state segments for the steady-state segments corresponding to concentration points that are not effectively identified includes: retrieving all glucose concentration gradient points in the unmatched steady-state response intervals; retrieving the average signal values ​​of adjacent and matched concentration points on both sides of the unmatched point; deducing the steady-state signal values ​​of the unmatched points based on the continuous change law of the concentration response; and supplementing the deduced values ​​into the corresponding missing steady-state segment data to form a complete set of steady-state segments.

[0075] This embodiment details the specific operation process for compensating for missing concentration point data in step S102, including: 1. Retrieve missing concentration points: Traverse all pre-defined glucose concentration gradient points, retrieve concentration points that do not match any steady-state response interval, and mark them as missing concentration points.

[0076] 2. Retrieve adjacent data: For each missing concentration point, retrieve the average signal value of the two adjacent concentration points on either side that have been matched. If the missing concentration point is located at the beginning of the concentration gradient, retrieve the data of the two adjacent concentration points to its right; if it is located at the end, retrieve the data of the two adjacent concentration points to its left.

[0077] 3. Deducing Steady-State Signal Values: Based on the continuous variation law of concentration response, the steady-state signal values ​​of missing concentration points are derived using a linear interpolation method. The calculation formula is: Missing point signal value = Left point signal value + (Missing point concentration - Left point concentration) × (Right point signal value - Left point signal value) / (Right point concentration - Left point concentration).

[0078] 4. Supplementing data to form a set: The steady-state signal values ​​obtained from the deduction are used as the steady-state segment data corresponding to the missing concentration point and supplemented into the steady-state segment set, ultimately forming a complete steady-state segment set containing all concentration gradient points.

[0079] In some embodiments, the step of quantitatively scoring each segment in the steady-state segment set and selecting high-quality steady-state segment data whose comprehensive scores meet preset screening conditions includes: independently scoring each segment in the steady-state segment set from three dimensions, including signal stability, segment consistency, and interval length; configuring fixed weight coefficients for the individual scores of the three dimensions, multiplying the individual scores by the corresponding weight coefficients and summing them to obtain the comprehensive score of the segment, and selecting segments with comprehensive scores higher than the preset score standard as high-quality steady-state segment data.

[0080] This embodiment details the specific operational process of multi-dimensional quantitative scoring and data filtering in step S103, including: 1. Calculate individual scores: Each segment within the steady-state segment set is scored independently from three dimensions: signal stability, segment consistency, and interval length. 2. Stability score: The calculation formula is 1 - (signal variance within the segment / total signal variance), and the score range is from 0 to 1. The larger the value, the more stable the signal.

[0081] 3. Consistency score: The calculation formula is 1-|(average value of segmented signal - linear interpolation of average value of adjacent segmented signal) / linear interpolation of average value of adjacent segmented signal|. The score range is from 0 to 1, and the larger the value, the better the consistency.

[0082] 4. Length score: The calculation formula is min(segment length / preset standard length, 1). The preset standard length is set to 30 sampling points. The score range is 0 to 1. The larger the value, the longer the interval and the more reliable it is.

[0083] 5. Configure weighting coefficients: Fixed weighting coefficients are assigned to the individual scores of the three dimensions, with stability score weighted at 0.5, consistency score at 0.3, and length score at 0.2. These weighting coefficients are optimized using a large amount of historical test data to balance the impact of each dimension on data quality.

[0084] 6. Calculate the overall score: Multiply the three individual scores of each segment by their corresponding weight coefficients and sum them up to obtain the overall score for that segment. The overall score ranges from 0 to 1.

[0085] 7. Filter high-quality data: Set a preset score standard of 0.7, and filter out segments with a comprehensive score higher than 0.7 as high-quality steady-state data. Segments with a comprehensive score lower than 0.7 are judged as low-quality data and will not be included in subsequent curve fitting calculations.

[0086] In some embodiments, the step of performing curve fitting operation based on the high-quality steady-state segment data obtained by screening to extract the sensitivity parameters of the electrochemical sensor includes: extracting the average steady-state current value corresponding to each glucose concentration point in the high-quality steady-state segment data, and constructing a fitting dataset with the glucose concentration and the corresponding average steady-state current value; performing a linear fitting operation on the fitting dataset, and determining the proportional coefficient of the fitting straight line as the sensitivity parameters of the electrochemical sensor.

[0087] This embodiment details the specific operation process for extracting sensitivity parameters in step S103, including: 1. Construct a fitting dataset: Extract the average steady-state current value corresponding to each glucose concentration point in the high-quality steady-state data, and construct a fitting dataset with glucose concentration as the x-axis and the corresponding average steady-state current value as the y-axis.

[0088] 2. Perform linear fitting: Perform a least-squares linear fitting operation on the fitted dataset to obtain the intercept and slope of the fitted line. The goal of linear fitting is to minimize the sum of the squares of the perpendicular distances from all data points to the fitted line.

[0089] 3. Determine the sensitivity parameter: The slope of the fitted straight line is determined as the sensitivity parameter of the electrochemical sensor, with units of nA / (mmol / L). This parameter represents the change in sensor output current when the glucose concentration changes by 1 mmol / L, and is a key indicator characterizing the sensor's core performance.

[0090] In some embodiments, the step of outputting a quality assessment report for the electrochemical sensor based on the sensitivity parameters includes: comparing the extracted sensitivity parameters with a preset qualified parameter range to determine parameter compliance; generating data quality evaluation content by combining the comprehensive scoring results of each steady-state segment; and integrating the sensitivity parameters, parameter compliance determination results, and data quality evaluation content to generate a standardized document and output it externally.

[0091] This embodiment details the specific operational process for outputting the quality assessment report in step S103, including: 1. Parameter Compliance Determination: The extracted sensitivity parameters are compared with a preset acceptable parameter range. The preset acceptable parameter range is determined based on the sensor's design specifications, for example, set to 80-120 nA / (mmol / L). If the sensitivity parameter is within the acceptable range, it is deemed acceptable; otherwise, it is deemed unacceptable.

[0092] 2. Generate Data Quality Assessment: Generate data quality assessment content by combining the comprehensive scores of each steady-state segment. Calculate the average comprehensive score for all high-quality steady-state segments, and categorize data quality into three levels based on the average score: average score ≥ 0.9 is excellent, 0.8 ≤ average score < 0.9 is good, and 0.7 ≤ average score < 0.8 is acceptable. Simultaneously, list the numbers and comprehensive scores of all low-quality segments, explaining the reasons for their exclusion.

[0093] 3. Generate and output report: Integrate sensitivity parameters, parameter compliance judgment results, data quality evaluation content, and detailed data from each concentration point to generate a standardized PDF quality assessment report. The report also includes basic information about this test, such as test time, sensor number, and test equipment model. The generated report is stored in a local database and sent to the designated terminal device via a network interface.

[0094] Please see Figure 4 As shown, Figure 4 This is a schematic diagram of the structure of an electrochemical sensor parameter extraction system 200 based on multidimensional steady-state analysis provided in this application embodiment. The electrochemical sensor parameter extraction system 200 based on multidimensional steady-state analysis is used to execute the steps of the electrochemical sensor parameter extraction method based on multidimensional steady-state analysis shown in the above embodiments. The electrochemical sensor parameter extraction system 200 based on multidimensional steady-state analysis can be a single server or a server cluster, or it can be a terminal, such as a handheld terminal, a laptop computer, a wearable device, or a robot.

[0095] like Figure 4 As shown, the electrochemical sensor parameter extraction system 200 based on multidimensional steady-state analysis includes: The sequence receiving unit 201 is used to receive the original electrochemical response signal sequence output by the continuous glucose monitoring device; and to perform time-domain filtering, edge-preserving filtering and smoothing filtering on the original electrochemical response signal sequence in sequence to obtain the filtered signal sequence. The polarization fitting unit 202 is used to perform polarization steady-state fitting processing on the initial segment of the filtered signal sequence to predict the steady-state response value. The steady-state identification unit 203 is used to determine the corresponding signal change characteristics based on the steady-state response value, output a quality assessment report of the electrochemical sensor according to the sensitivity parameter, and identify the steady-state response interval in the signal sequence according to the signal change characteristics. The segmented compensation unit 204 is used to numerically segment the identified steady-state response interval, establish the correspondence between the concentration gradient and the signal interval, and for the steady-state segment corresponding to the concentration point that was not effectively identified, infer the reasonable value range and perform data compensation using the continuity characteristics of the concentration response to obtain a complete set of steady-state segments. The quantification scoring unit 205 is used to quantify and score each segment in the steady-state segment set, and to select high-quality steady-state segment data that meet the preset screening conditions in the steady-state segment set; to perform curve fitting operation based on the selected high-quality steady-state segment data, and to extract the sensitivity parameters of the electrochemical sensor; and to output a quality assessment report of the electrochemical sensor according to the sensitivity parameters.

[0096] In some embodiments, performing time-domain filtering, edge-preserving filtering, and smoothing filtering sequentially on the original electrochemical response signal sequence to obtain a filtered signal sequence includes: performing an upsampling operation on the original electrochemical response signal sequence to unify the overall data sampling frequency; performing time-domain filtering on the signal sequence after unifying the sampling frequency to remove pulse interference signals; performing edge-preserving filtering on the time-domain filtered signal sequence to retain signal mutation characteristics; performing smoothing filtering on the edge-preserving filtered signal sequence to remove high-frequency random noise; and outputting the processed filtered signal sequence.

[0097] In some embodiments, performing polarization steady-state fitting processing on the initial segment of the filtered signal sequence to predict the steady-state response value includes: extracting continuous signal data from the initial period of the filtered signal sequence; performing fitting operations on the extracted initial period data using a multi-exponential decay mode to solve for the decay parameter and steady-state offset; and determining the solved steady-state offset as the predicted steady-state response value.

[0098] In some embodiments, determining the corresponding signal change characteristics based on the steady-state response value, outputting a quality assessment report of the electrochemical sensor according to the sensitivity parameters, and identifying the steady-state response interval in the signal sequence based on the signal change characteristics include: setting a maximum signal change threshold and a signal variance threshold based on the steady-state response value; traversing the filtered signal sequence using a fixed-size sliding window; calculating the maximum signal change and signal variance within each sliding window; determining the sliding window that simultaneously satisfies both threshold constraints as a steady-state window; and merging all continuously distributed steady-state windows to obtain a complete steady-state response interval.

[0099] In some embodiments, the step of numerically segmenting the identified steady-state response interval and establishing the correspondence between the concentration gradient and the signal interval includes: calculating the average value of all signal sampling points within a single steady-state response interval; calling a pre-trained current concentration mapping model to match the average value of each steady-state response interval to the glucose concentration gradient point; and establishing a one-to-one binding association between the concentration gradient point and the steady-state signal interval.

[0100] In some embodiments, the step of inferring a reasonable range of values ​​and performing data compensation based on the continuity characteristics of the concentration response to obtain a complete set of steady-state segments for the steady-state segments corresponding to concentration points that are not effectively identified includes: retrieving all glucose concentration gradient points in the unmatched steady-state response intervals; retrieving the average signal values ​​of adjacent and matched concentration points on both sides of the unmatched point; deducing the steady-state signal values ​​of the unmatched points based on the continuous change law of the concentration response; and supplementing the deduced values ​​into the corresponding missing steady-state segment data to form a complete set of steady-state segments.

[0101] In some embodiments, the step of quantitatively scoring each segment in the steady-state segment set and selecting high-quality steady-state segment data whose comprehensive scores meet preset screening conditions includes: independently scoring each segment in the steady-state segment set from three dimensions, including signal stability, segment consistency, and interval length; configuring fixed weight coefficients for the individual scores of the three dimensions, multiplying the individual scores by the corresponding weight coefficients and summing them to obtain the comprehensive score of the segment, and selecting segments with comprehensive scores higher than the preset score standard as high-quality steady-state segment data.

[0102] In some embodiments, the step of performing curve fitting operation based on the high-quality steady-state segment data obtained by screening to extract the sensitivity parameters of the electrochemical sensor includes: extracting the average steady-state current value corresponding to each glucose concentration point in the high-quality steady-state segment data, and constructing a fitting dataset with the glucose concentration and the corresponding average steady-state current value; performing a linear fitting operation on the fitting dataset, and determining the proportional coefficient of the fitting straight line as the sensitivity parameters of the electrochemical sensor.

[0103] In some embodiments, the step of outputting a quality assessment report for the electrochemical sensor based on the sensitivity parameters includes: comparing the extracted sensitivity parameters with a preset qualified parameter range to determine parameter compliance; generating data quality evaluation content by combining the comprehensive scoring results of each steady-state segment; and integrating the sensitivity parameters, parameter compliance determination results, and data quality evaluation content to generate a standardized document and output it externally.

[0104] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the electrochemical sensor parameter extraction system and its modules based on multidimensional steady-state analysis described above can be found in the corresponding contents of the various embodiments of the electrochemical sensor parameter extraction method based on multidimensional steady-state analysis, and will not be repeated here.

[0105] The aforementioned method for extracting electrochemical sensor parameters based on multidimensional steady-state analysis can be implemented as a computer program, which can be used in various applications such as... Figure 4 It runs on the device shown.

[0106] Please see Figure 5 , Figure 5 This is a schematic block diagram of the structure of a computer device provided in an embodiment of this application. The computer device includes a processor, a memory, and a network interface connected via a device bus, wherein the memory may include a storage medium and internal memory.

[0107] The storage medium can store operating devices and computer programs. The computer program includes program instructions that, when executed, cause the processor to perform any electrochemical sensor parameter extraction method based on multidimensional steady-state analysis.

[0108] The processor provides computing and control capabilities, supporting the operation of the entire computer device.

[0109] Internal memory provides an environment for the execution of computer programs in non-volatile storage media. When executed by a processor, the computer program enables the processor to perform any electrochemical sensor parameter extraction method based on multidimensional steady-state analysis.

[0110] This network interface is used for network communication, such as sending assigned tasks. Those skilled in the art will understand that... Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the terminal to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0111] It should be understood that the processor can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, a general-purpose processor can be a microprocessor or any conventional processor.

[0112] In one embodiment, the processor is configured to run a computer program stored in memory to perform the following steps: Receive the raw electrochemical response signal sequence output by the continuous glucose monitoring device; perform time-domain filtering, edge-preserving filtering and smoothing filtering on the raw electrochemical response signal sequence in sequence to obtain the filtered signal sequence; Polarization steady-state fitting is performed on the initial segment of the filtered signal sequence to predict the steady-state response value. Based on the steady-state response value, the corresponding signal change characteristics are determined, and a quality assessment report of the electrochemical sensor is output according to the sensitivity parameters. The steady-state response interval in the signal sequence is identified based on the signal change characteristics. The identified steady-state response interval is numerically segmented to establish the correspondence between the concentration gradient and the signal interval. For the steady-state segments corresponding to concentration points that are not effectively identified, the reasonable value range is inferred using the continuity characteristics of the concentration response and data compensation is performed to obtain a complete set of steady-state segments. Each segment in the steady-state segment set is quantitatively scored, and high-quality steady-state segment data that meet the preset screening conditions are selected from the steady-state segment set. Based on the selected high-quality steady-state segment data, curve fitting operation is performed to extract the sensitivity parameters of the electrochemical sensor. Based on the sensitivity parameters, a quality assessment report of the electrochemical sensor is output.

[0113] In some embodiments, performing time-domain filtering, edge-preserving filtering, and smoothing filtering sequentially on the original electrochemical response signal sequence to obtain a filtered signal sequence includes: performing an upsampling operation on the original electrochemical response signal sequence to unify the overall data sampling frequency; performing time-domain filtering on the signal sequence after unifying the sampling frequency to remove pulse interference signals; performing edge-preserving filtering on the time-domain filtered signal sequence to retain signal mutation characteristics; performing smoothing filtering on the edge-preserving filtered signal sequence to remove high-frequency random noise; and outputting the processed filtered signal sequence.

[0114] In some embodiments, performing polarization steady-state fitting processing on the initial segment of the filtered signal sequence to predict the steady-state response value includes: extracting continuous signal data from the initial period of the filtered signal sequence; performing fitting operations on the extracted initial period data using a multi-exponential decay mode to solve for the decay parameter and steady-state offset; and determining the solved steady-state offset as the predicted steady-state response value.

[0115] In some embodiments, determining the corresponding signal change characteristics based on the steady-state response value, outputting a quality assessment report of the electrochemical sensor according to the sensitivity parameters, and identifying the steady-state response interval in the signal sequence based on the signal change characteristics include: setting a maximum signal change threshold and a signal variance threshold based on the steady-state response value; traversing the filtered signal sequence using a fixed-size sliding window; calculating the maximum signal change and signal variance within each sliding window; determining the sliding window that simultaneously satisfies both threshold constraints as a steady-state window; and merging all continuously distributed steady-state windows to obtain a complete steady-state response interval.

[0116] In some embodiments, the step of numerically segmenting the identified steady-state response interval and establishing the correspondence between the concentration gradient and the signal interval includes: calculating the average value of all signal sampling points within a single steady-state response interval; calling a pre-trained current concentration mapping model to match the average value of each steady-state response interval to the glucose concentration gradient point; and establishing a one-to-one binding association between the concentration gradient point and the steady-state signal interval.

[0117] In some embodiments, the step of inferring a reasonable range of values ​​and performing data compensation based on the continuity characteristics of the concentration response to obtain a complete set of steady-state segments for the steady-state segments corresponding to concentration points that are not effectively identified includes: retrieving all glucose concentration gradient points in the unmatched steady-state response intervals; retrieving the average signal values ​​of adjacent and matched concentration points on both sides of the unmatched point; deducing the steady-state signal values ​​of the unmatched points based on the continuous change law of the concentration response; and supplementing the deduced values ​​into the corresponding missing steady-state segment data to form a complete set of steady-state segments.

[0118] In some embodiments, the step of quantitatively scoring each segment in the steady-state segment set and selecting high-quality steady-state segment data whose comprehensive scores meet preset screening conditions includes: independently scoring each segment in the steady-state segment set from three dimensions, including signal stability, segment consistency, and interval length; configuring fixed weight coefficients for the individual scores of the three dimensions, multiplying the individual scores by the corresponding weight coefficients and summing them to obtain the comprehensive score of the segment, and selecting segments with comprehensive scores higher than the preset score standard as high-quality steady-state segment data.

[0119] In some embodiments, the step of performing curve fitting operation based on the high-quality steady-state segment data obtained by screening to extract the sensitivity parameters of the electrochemical sensor includes: extracting the average steady-state current value corresponding to each glucose concentration point in the high-quality steady-state segment data, and constructing a fitting dataset with the glucose concentration and the corresponding average steady-state current value; performing a linear fitting operation on the fitting dataset, and determining the proportional coefficient of the fitting straight line as the sensitivity parameters of the electrochemical sensor.

[0120] In some embodiments, the step of outputting a quality assessment report for the electrochemical sensor based on the sensitivity parameters includes: comparing the extracted sensitivity parameters with a preset qualified parameter range to determine parameter compliance; generating data quality evaluation content by combining the comprehensive scoring results of each steady-state segment; and integrating the sensitivity parameters, parameter compliance determination results, and data quality evaluation content to generate a standardized document and output it externally.

[0121] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to implement the steps of the electrochemical sensor parameter extraction method based on multidimensional steady-state analysis provided in any embodiment of this application.

[0122] The computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiments, such as the hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, SmartMedia Card (SMC), Secure Digital (SD) card, or Flash Card equipped on the computer device.

[0123] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for extracting electrochemical sensor parameters based on multi-dimensional steady-state analysis, characterized in that, include: Receive the raw electrochemical response signal sequence output by the continuous glucose monitoring device; perform time-domain filtering, edge-preserving filtering and smoothing filtering on the raw electrochemical response signal sequence in sequence to obtain the filtered signal sequence; Polarization steady-state fitting is performed on the initial segment of the filtered signal sequence to predict the steady-state response value; based on the steady-state response value, the corresponding signal change characteristics are determined, and a differentiated identification strategy is adopted according to the concentration gradient range to identify the steady-state response interval in the signal sequence based on the signal change characteristics. The identified steady-state response intervals are numerically segmented, and the correspondence between concentration gradients and signal intervals is established. For steady-state segments corresponding to concentration points that are not effectively identified, the reasonable range of values ​​is inferred using the continuity characteristics of the concentration response, and data compensation is performed to obtain a complete set of steady-state segments. Each segment in the steady-state segment set is quantitatively scored, and high-quality steady-state segment data that meet the preset screening conditions are selected from the steady-state segment set. Based on the selected high-quality steady-state segment data, curve fitting operation is performed to extract the sensitivity parameters of the electrochemical sensor. Based on the sensitivity parameters, a quality assessment report of the electrochemical sensor is output.

2. The method according to claim 1, characterized in that, The original electrochemical response signal sequence is sequentially subjected to time-domain filtering, edge-preserving filtering, and smoothing filtering to obtain a filtered signal sequence, including: The original electrochemical response signal sequence is upsampled to unify the overall data sampling frequency. The signal sequence after unifying the sampling frequency is then subjected to time-domain filtering to remove pulse interference signals. The signal sequence after time-domain filtering is then subjected to edge-preserving filtering to retain signal mutation characteristics. The signal sequence after edge-preserving filtering is then subjected to smoothing filtering to remove high-frequency random noise. Output the filtered signal sequence after processing.

3. The method according to claim 1, characterized in that, The process of performing polarization steady-state fitting on the initial segment of the filtered signal sequence to predict the steady-state response value includes: Extract continuous signal data from the initial period of the filtered signal sequence; The data from the initial period were fitted using a multi-exponential decay model to obtain the decay parameters and steady-state offset. The steady-state offset obtained from the solution is determined as the predicted steady-state response value.

4. The method according to claim 1, characterized in that, The process of determining the corresponding signal change characteristics based on the steady-state response value, outputting a quality assessment report for the electrochemical sensor based on the sensitivity parameters, and identifying the steady-state response interval in the signal sequence based on the signal change characteristics includes: Based on the steady-state response value, the maximum change threshold and the signal variance threshold are set, and a fixed-size sliding window is used to traverse the filtered signal sequence. Calculate the maximum change in signal and the signal variance within each sliding window, and determine the sliding window that simultaneously satisfies both threshold constraints as a steady-state window. Merge all continuously distributed steady-state windows to obtain the complete steady-state response interval.

5. The method according to claim 1, characterized in that, The step of numerically segmenting the identified steady-state response interval and establishing the correspondence between the concentration gradient and the signal interval includes: Calculate the average value of all signal sampling points within a single steady-state response interval, call the pre-trained current concentration mapping model, and match the average value of each steady-state response interval to the corresponding glucose concentration gradient point. Establish a one-to-one binding relationship between concentration gradient points and steady-state signal intervals.

6. The method according to claim 5, characterized in that, For the steady-state segments corresponding to concentration points that were not effectively identified, the reasonable value range was inferred using the continuity characteristics of the concentration response, and data compensation was performed to obtain a complete set of steady-state segments, including: Retrieve all glucose concentration gradient points in the unmatched steady-state response interval, and retrieve the average signal of the adjacent and matched concentration points on both sides of the unmatched point. Based on the continuous change law of concentration response, the steady-state signal values ​​of the unmatched points are deduced, and the deduced values ​​are supplemented to form the corresponding missing steady-state segment data to form a complete set of steady-state segments.

7. The method according to claim 1, characterized in that, The process of quantifying and scoring each segment in the steady-state segment set, and then selecting high-quality steady-state segment data that meet preset screening criteria based on their comprehensive scores, includes: Each segment within the steady-state segment set is scored independently from three dimensions: signal stability, segment consistency, and interval length. Fixed weight coefficients are assigned to the individual scores of the three dimensions. The individual scores are multiplied by the corresponding weight coefficients and then summed to obtain the segmented comprehensive scores. Segments with comprehensive scores higher than the preset score standard are selected as high-quality steady-state segment data.

8. The method according to claim 1, characterized in that, The high-quality steady-state data obtained through screening is used to perform curve fitting operations to extract the sensitivity parameters of the electrochemical sensor, including: Extract the average steady-state current value corresponding to each glucose concentration point in the high-quality steady-state segment data, and construct the fitting dataset with glucose concentration and corresponding average steady-state current value. A linear fitting operation is performed on the fitted dataset, and the proportional coefficient of the fitted straight line is determined as the sensitivity parameter of the electrochemical sensor.

9. The method according to claim 1, characterized in that, The process of outputting a quality assessment report for the electrochemical sensor based on the sensitivity parameters includes: The extracted sensitivity parameters are compared with the preset qualified parameter range to determine the compliance of the parameters, and the data quality evaluation content is generated by combining the comprehensive scoring results of each steady-state segment. Integrate sensitivity parameters, parameter compliance judgment results, and data quality evaluation content to generate standardized documents and output them externally.

10. An electrochemical sensor parameter extraction system based on multidimensional steady-state analysis, used to implement the method as described in any one of claims 1-9, characterized in that, include: The sequence receiving unit is used to receive the raw electrochemical response signal sequence output by the continuous glucose monitoring device; and to perform time-domain filtering, edge-preserving filtering and smoothing filtering on the raw electrochemical response signal sequence in sequence to obtain the filtered signal sequence. The polarization fitting unit is used to perform polarization steady-state fitting on the initial segment of the filtered signal sequence to predict the steady-state response value. The steady-state identification unit is used to determine the corresponding signal change characteristics based on the steady-state response value, output a quality assessment report of the electrochemical sensor according to the sensitivity parameters, and identify the steady-state response interval in the signal sequence according to the signal change characteristics. The segmented compensation unit is used to numerically segment the identified steady-state response interval, establish the correspondence between the concentration gradient and the signal interval, and for the steady-state segment corresponding to the concentration point that was not effectively identified, infer the reasonable value range and perform data compensation using the continuity characteristics of the concentration response to obtain a complete set of steady-state segments. The quantitative scoring unit is used to quantitatively score each segment in the steady-state segment set, and to select high-quality steady-state segment data that meet the preset screening conditions in the steady-state segment set; to perform curve fitting operation based on the selected high-quality steady-state segment data, and to extract the sensitivity parameters of the electrochemical sensor; and to output a quality assessment report of the electrochemical sensor based on the sensitivity parameters.

Images