In-vivo sensitivity calibration method and device of biosensor and storage medium
By establishing a current signal trend correlation algorithm model within the biosensor and using machine learning to predict the sensor's in vivo sensitivity, the signal deviation problem caused by individual differences was solved, achieving high reliability and accurate calibration of the sensor.
Patent Information
- Application Number
- CN202510932388.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-07
- Publication Date
- 2025-11-21
AI Technical Summary
Existing biosensors exhibit inconsistent sensitivity in vitro and in vivo due to individual differences and batch-specific sterilization variations, leading to detection signal deviations and low reliability.
By establishing a correlation algorithm model based on the current signal trend of the sensor during the initial polarization stage in the human body, the in vivo sensitivity of the sensor is predicted using a machine learning model. The sensor is driven by a constant or waveform voltage to collect initial signal data, construct feature parameters and train the algorithm model, and output the prediction results.
This improves the reliability and accuracy of the sensor, reduces individual variability and uncontrollability, and ensures the accuracy of the sensor's signal detection within the body.
Smart Images
Figure CN120983034A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to a sensor calibration method, in particular to an in-vivo sensitivity calibration method, device and storage medium of a biosensor. BACKGROUND
[0002] In recent years, diabetes monitoring technology has developed rapidly, and continuous glucose monitoring (CGM) has been widely used. CGM devices can update blood glucose information every 1-5 minutes, providing continuous and comprehensive dynamic blood glucose data, showing blood glucose fluctuation status under different scenarios, discovering unobservable high and low blood glucose risks, and promoting individualized treatment. CGM devices are mainly composed of three parts, including a disposable wired sensor inserted into subcutaneous tissue, a transmitter, and a receiver (handheld device or smart phone) that displays glucose data. The CGM on the market is a minimally invasive device that implants a sensor subcutaneously to obtain interstitial fluid glucose concentration. Its main mechanism is electro-chemical method, glucose reacts with glucose oxidase / glucose dehydrogenase, and finally generates electrons corresponding to the glucose concentration. The electrical signal is processed to calculate the glucose concentration.
[0003] Currently, glucose sensors need to be factory calibrated after production, and the sensitivity of the sensor is determined based on the correlation between the in-vitro sensitivity and the in-vivo sensitivity of the sensor. The method used in the prior art is to factory calibrate the sensor sensitivity based on prior calibration parameters (such as in-vitro test information from batch-produced sensors), but in actual use, the in-vitro and in-vivo changes of each sensor sensitivity are not consistent, individual differences have a great influence, resulting in a certain deviation in the signal detected by the sensor, and the reliability is not high. In addition, there are also differences in sterilization batches, the influence of storage and transportation process on the sensitivity of the sensor, etc., which cause the difference and uncontrollability of the in-vivo sensitivity of the sensor. SUMMARY
[0004] To solve the above problems, the present application provides an in-vivo sensitivity calibration method for a biosensor, and the specific technical solution is as follows: An in-vivo sensitivity calibration method for a biosensor, an algorithm model is established according to the correlation between the sensitivity of the sensor in the human body and the current signal trend of the sensor in the initial polarization stage in the human body, and the algorithm model is used to predict the in-vivo sensitivity of the sensor; wherein the voltage of the sensor is controlled to drive the sensor in the initialization stage.
[0005] Preferably, the voltage includes a constant voltage or a waveform voltage.
[0006] Preferably, the constant voltage includes a working voltage or a preset voltage of the sensor.
[0007] Preferably, the waveform voltage includes any one or a combination of a sine wave voltage, a square wave voltage, a triangular wave voltage or a trapezoidal wave voltage.
[0008] Preferably, the method comprises the following steps: sensor initialization driving mode: controlling the voltage on the sensor to drive the sensor; data collection: collecting early signal data after the sensor is implanted and production information of the sensor; creating characteristic parameters: constructing characteristic parameters according to the collected data samples, the characteristic parameters including initial signal trend characteristics and time sequence characteristics; establishing an algorithm model: establishing and training an algorithm model according to the created characteristic parameters and the collected data samples; and outputting a prediction result: obtaining the in-vivo sensitivity of the sensor.
[0009] Preferably, the production information of the sensor includes one or more of initial parameters of the sensor, production batches, production process information and sterilization information.
[0010] Preferably, the algorithm model includes a machine learning model or a deep learning model.
[0011] Preferably, the obtaining of the correlation includes: collecting signal data of a polarization stage of the sensor after being implanted into the human body, creating characteristic parameters representing the current signal of the stage, and performing correlation analysis on the sensitivity of the sensor.
[0012] An apparatus of a method for calibrating in-vivo sensitivity of a biosensor, the apparatus comprising: a processor, a memory and a program; the program is stored in the memory, and the processor invokes the program stored in the memory to execute steps of the method for calibrating in-vivo sensitivity of a biosensor.
[0013] A computer readable storage medium configured to store a program, the program being configured to execute steps of the method for calibrating in-vivo sensitivity of a biosensor.
[0014] Compared with the prior art, the present application has the following beneficial effects: The method for calibrating in-vivo sensitivity of a biosensor provided by the present application calibrates the sensitivity of the sensor based on initial signal data in the body, which can effectively improve the reliability and accuracy of the sensor and effectively solve and avoid problems caused by consistency and individual differences of the sensor. BRIEF DESCRIPTION OF DRAWINGS
[0015] Figure 1 is a flowchart of the present application; Figure 2 is a linear correlation diagram of sensor sensitivity and in-vivo signal characteristics; Figure 3 is a model prediction result comparison diagram; Figure 4 is a schematic diagram of a constant voltage; Figure 5 is a schematic diagram of a square wave voltage; Figure 6 is a schematic diagram of a triangular wave voltage; Figure 7 is a schematic diagram of a sinusoidal wave voltage; Figure 8 is a schematic diagram of a trapezoidal wave voltage. DETAILED DESCRIPTION
[0016] The application will be further described with reference to the accompanying drawings.
[0017] Through the analysis of in-vivo test data, it is found that the sensitivity of the sensor in the human body has a high correlation with the current signal trend in the initial polarization stage in the body.
[0018] Through the collected polarization stage signal data of the sensor after being implanted in the human body, characteristic parameters representing the current signal in this stage are created, and the correlation of the sensor sensitivity is analyzed. The analysis result proves that there is a certain correlation between these signal characteristics and the sensitivity of the sensor. The analysis result can be seen in the following chart.
[0019]
[0020] MANOVA (Multivariate Analysis of Variance) can be used to analyze the influence of multiple independent variables on dependent variables, or the relationship between multiple independent variables.
[0021] In the Multivariate Tests table, Pillai's Trace, Wilks' Lambda, Hotelling's Trace, and Roy's Largest Root are four multivariate statistics used to test group differences. The most commonly used statistic is Wilks' Lambda, which tests P When P < 0.05, the group difference of the independent variable has statistical significance.
[0022] Figure 2 is a linear correlation graph (Pearson correlation) between the characteristic variables of the algorithm model, in which characteristic 1 to characteristic n (here n = 15), including but not limited to, sensor signal trend related characteristics, sensor parameter information related characteristics, etc.
[0023] Pearson correlation is usually represented by the symbol r, and its value ranges from -1 to 1, indicating the degree of linear relationship between two variables. The calculation formula is as follows:
[0024] When the r value is close to 1, it indicates that there is a strong positive correlation between the two variables; when the r value is close to -1, it indicates that there is a strong negative correlation between the two variables; when the r value is close to 0, it indicates that there is a weak correlation or no relationship between the two variables.
[0025] The in-vivo sensitivity calibration method of a biosensor is used to calibrate the sensitivity change of the sensor after being implanted in the human body.
[0026] The in-vivo sensitivity of the sensor is predicted based on a machine learning model. The learning model can include but is not limited to various regression algorithms such as linear regression, logistic regression, support vector machine, decision tree, random forest, neural network, etc.
[0027] The method mainly includes five steps: sensor initialization, data collection, creation of feature parameters, model establishment and training, and result prediction.
[0028] As shown in Figure 1 The in-vivo sensitivity calibration method of a biosensor drives the sensor by controlling the voltage of the sensor during the initialization stage, stimulates the sensor to generate more initialization information, collects the corresponding current information at the same time, and establishes an algorithm model according to the correlation between the sensitivity of the sensor in the human body and the current signal trend of the sensor in the initial polarization stage in the human body. The algorithm model is used to predict the in-vivo sensitivity of the sensor.
[0029] Driving the sensor by voltage during the initialization stage can promote the initialization of the sensor faster, so that the sensor can enter the stable working stage faster.
[0030] Specifically, the following steps are included: Sensor initialization driving mode: control the voltage on the sensor to drive the sensor; Data collection: collect the early signal data after the sensor is implanted and the production information of the sensor; Creating feature parameters: constructing feature parameters according to the collected data samples, the feature parameters including initial signal trend features and time series features; Establishing an algorithm model: establishing and training an algorithm model according to the created feature parameters and collected data samples; Output prediction results: get the in-vivo sensitivity of the sensor.
[0031] The production information of the sensor includes initial parameters of the sensor, factory calibration parameters, production test data, production batch, production process information, and sterilization information.
[0032] The algorithm model includes a multi-dimensional information model, a multi-modal model, a machine learning model, or a deep learning model.
[0033] The acquisition of correlations includes: collecting polarization phase signal data of the sensor after it is implanted in the human body, creating characteristic parameters representing the current signal in this phase, and performing correlation analysis on the sensor sensitivity.
[0034] The final output prediction result is the in vivo sensitivity of the sensor.
[0035] Figure 3 This is an example graph showing the algorithm's prediction results based on internal data. The horizontal axis represents the true sensitivity, and the vertical axis represents the predicted sensitivity. The red line is the 45° ideal line (i.e., the predicted sensitivity equals the true sensitivity), and the two gray lines are tolerance lines. Ideally, the predicted sensitivity is equal to or as close as possible to the true sensitivity. Therefore, the data points on the graph will fall on the red line or be as close as possible to the red line, or be distributed as much as possible within the two gray tolerance lines. Figure 3 As can be seen, the majority of the results predicted by the algorithm are distributed near the red dot and within the two gray tolerance lines, proving the accuracy of the prediction sensitivity and demonstrating that the algorithm has achieved excellent results.
[0036] Traditional sensor factory calibration is based on obtaining prior calibration parameters from sensor data produced in vitro to calibrate batches of sensors. This method has high requirements for product consistency. If the consistency of the batch is poor, the measured glucose value will be unreliable. In contrast, this application estimates and calibrates the sensor sensitivity based on in vivo data of individual sensors, thereby improving the reliability and accuracy of the sensor, reducing the uncontrollability of the sensor, and enabling factory calibration of individual products.
[0037] like Figure 4 to Figure 8 As shown, the voltage includes constant voltage or waveform voltage. Constant voltage includes the sensor operating voltage or preset voltage. Waveform voltage includes any one or more combinations of sine wave voltage, square wave voltage, triangular wave voltage, or trapezoidal wave voltage, with different waveforms used sequentially according to time.
[0038] An apparatus for in vivo sensitivity calibration of a biosensor, the apparatus comprising: a processor, a memory, and a program; the program is stored in the memory, and the processor invokes the program stored in the memory to execute the steps of the in vivo sensitivity calibration method for a biosensor. The apparatus itself can be a terminal device, such as a sensor, transmitter, wearable device, etc., or it can be a mobile phone used in conjunction with the device, or a cloud server.
[0039] The memory and the processor are electrically connected directly or indirectly to realize the transmission or interaction of data. For example, the elements can be electrically connected through one or more communication buses or signal lines, such as through a bus connection. The memory stores computer execution instructions for realizing the data access control method, including at least one software function module stored in the memory in the form of software or firmware. The processor executes various function applications and data processing by running the software program and the module stored in the memory.
[0040] The memory can be, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), etc. The memory is used to store programs, and the processor executes the programs after receiving execution instructions.
[0041] The processor can be an integrated circuit chip with signal processing capability. The processor described above can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc. The processor can realize or execute the disclosed methods, steps and logic block diagrams in the embodiments of the present application. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.
[0042] The embodiments of the present application are described with reference to the flowcharts according to the method, terminal device (system), and computer program product of the embodiments of the present application. The computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to produce a machine, so that the instructions executed by the computer or other programmable data processing terminal device produce a device for realizing the functions specified in the flowcharts and / or.
[0043] A computer readable storage medium is configured to store a program, and the program is configured to execute the steps of the in-vivo sensitivity calibration method of the biosensor.
[0044] Computer readable storage media as used herein includes random access memory (RAM), internal memory, read only memory (ROM), erasable programmable ROM, electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disks, CD-ROMs, or any other form of storage medium known in the art.
[0045] The technical principles of the present application are described above in connection with specific embodiments. These descriptions are only for explaining the principles of the present application and cannot be interpreted as limiting the scope of protection of the present application in any way. Based on the explanations herein, other specific embodiments of the present application can be conceived by those skilled in the art without any creative effort, and these embodiments will all fall within the scope of protection of the claims of the present application.
Claims
1. A method of in vivo sensitivity calibration of a biosensor, characterized by, An algorithm model is established according to the correlation between the sensitivity of the sensor in the human body and the current signal trend of the sensor in the initial polarization stage in the human body, and the algorithm model is used to predict the in-vivo sensitivity of the sensor. The initialization stage is achieved by controlling the voltage of the sensor to drive the sensor.
2. The method of claim 1, wherein the calibration is performed in vivo. The voltage includes a constant voltage or a waveform voltage.
3. The method of claim 2, wherein the calibration is performed in vivo. The constant voltage includes a working voltage of the sensor or a preset voltage.
4. The method of claim 2, wherein the calibration is performed in vivo. The waveform voltage includes any one or a combination of a sine wave voltage, a square wave voltage, a triangular wave voltage, or a trapezoidal wave voltage.
5. The method of in vivo sensitivity calibration of a biosensor according to any one of claims 1 to 4, characterized in that, The method comprises the following steps: Sensor initialization; Data collection: collecting early signal data after the sensor is implanted and production information of the sensor; Creating characteristic parameters: constructing characteristic parameters according to the collected data samples, the characteristic parameters including initial signal trend characteristics and time sequence characteristics; Establishing an algorithm model: establishing and training an algorithm model according to the created characteristic parameters and the collected data samples; Outputting a prediction result: obtaining the in-vivo sensitivity of the sensor.
6. The method of claim 5, wherein the calibration is performed in vivo. The production information of the sensor includes one or more of initial parameters of the sensor, production batches, production process information, and sterilization information.
7. The in vivo sensitivity calibration method for a biosensor according to claim 5, characterized in that, The algorithm model includes a machine learning model or a deep learning model.
8. The method of claim 1, wherein the calibration is performed in vivo. The correlation is obtained by collecting polarization stage signal data of the sensor after being implanted into the human body, creating characteristic parameters representing the current signal in this stage, and performing correlation analysis on the sensitivity of the sensor.
9. An apparatus for calibrating the in-vivo sensitivity of a biosensor, the apparatus comprising: a processor, a memory, and a program; The program is stored in the memory, and the processor calls the program stored in the memory to execute the steps of the method for calibrating the in-vivo sensitivity of a biosensor according to claim 1.
10. A computer-readable storage medium, characterized in that, The computer readable storage medium is configured to store a program configured to perform the steps of the method for calibrating the in-vivo sensitivity of a biosensor according to claim 1.