A non-invasive blood glucose concentration detection method based on passive UHF RFID

Abstract

The application relates to the fields of Internet of Things and biomedical sensing technology, and discloses a non-invasive blood glucose concentration detection method based on passive UHF RFID, which comprises the following steps: fixing a passive UHF RFID tag on a preset monitoring part of the body surface of a living experimental animal, and placing the living experimental animal in a wireless signal reading area constructed by an RFID reader and an antenna thereof; and performing exogenous blood glucose regulation on the living experimental animal to make the blood glucose concentration in the living experimental animal produce dynamic changes covering the normal blood glucose concentration range. By constructing a standardized and systematic blood glucose detection process, from the early preparation of the experimental animal, exogenous blood glucose regulation, to the synchronous collection of signals and blood glucose reference values, and then to the preprocessing of the data set and the establishment of a quantitative model, each step is closely connected and the operation is standardized, random interference in the living experiment is effectively inhibited, and the reliability and stability of the detection data are guaranteed.

Description

Technical Field

[0001] This invention relates to the fields of Internet of Things and biomedical sensing technology, specifically a non-invasive blood glucose concentration detection method based on passive UHF RFID. Background Technology

[0002] Diabetes mellitus, a prevalent chronic metabolic disease worldwide, relies heavily on accurate and continuous monitoring of blood glucose levels for diagnosis and management. Timely and accurate blood glucose data is crucial for adjusting treatment plans and preventing complications in diabetic patients. Therefore, blood glucose monitoring technology has always been a key research focus in the field of biomedical sensing. Currently, the mainstream method for blood glucose monitoring in clinical and experimental studies is invasive blood sampling. This method involves collecting blood from the fingertip, tailbone, or other sites and using a blood glucose meter to measure blood glucose levels. While it can provide accurate reference values, it has significant inherent drawbacks: the blood sampling process causes physical trauma to the subject, easily leading to pain, tissue damage, and a potential risk of wound infection; furthermore, this method requires disposable consumables such as lancets and blood glucose test strips, resulting in high operating costs over long-term monitoring; and the invasive nature of the blood sampling procedure makes it difficult to achieve continuous, high-frequency blood glucose monitoring, failing to fully capture the dynamic changes in blood glucose levels and limiting the analysis and control of blood glucose trends.

[0003] To overcome the drawbacks of invasive testing, researchers have successively developed various non-invasive blood glucose monitoring technologies, including optical methods, impedance methods, and microwave methods. These technologies eliminate the need for blood sampling, enabling non-invasive blood glucose detection and significantly improving the convenience and acceptability of monitoring, thus becoming an important direction for the development of blood glucose detection technology. Among them, the optical method relies on the absorption and scattering characteristics of glucose to specific wavelengths of light to achieve blood glucose detection; the impedance method inverts blood glucose values ​​by detecting changes in tissue impedance caused by changes in glucose concentration in body fluids; and the microwave method uses the dielectric response of glucose to microwave signals to complete monitoring. However, existing non-invasive blood glucose monitoring technologies still have many problems to be solved in practical applications. The optical method is easily affected by factors such as skin pigmentation and tissue thickness at the detection site, resulting in insufficient detection accuracy and poor stability; the impedance method has strict requirements for the temperature and humidity of the detection environment and the electrode contact state, and is prone to signal drift, affecting the accuracy of the detection results; the microwave method requires complex hardware structures, is difficult to integrate, and has high equipment manufacturing costs, making it difficult to achieve widespread application. The aforementioned technical challenges mean that non-invasive blood glucose monitoring is still in a continuous optimization phase. Further breakthroughs are needed in terms of detection accuracy, system stability, and equipment cost to better meet the refined requirements of continuous blood glucose monitoring in clinical and experimental research.

[0004] In recent years, radio frequency identification (RFID) technology has seen its application in the biomedical sensing field deepen due to its advantages of being non-contact, passive, low-cost, and easy to integrate. Among these technologies, passive ultra-high frequency (UHF) RFID technology has become a potential solution for non-invasive biomarker detection due to its moderate signal transmission distance, lack of tag power requirements, and strong adaptability. The backscattered signal intensity of passive UHF RFID tags changes accordingly with variations in the dielectric properties of the surrounding medium. Changes in glucose concentration directly affect the complex dielectric constant of biological media, providing a physical basis for non-invasive blood glucose concentration detection based on passive UHF RFID technology. Currently, some studies have attempted to explore blood glucose detection using passive UHF RFID technology. However, most existing studies are limited to in vitro simulated solution environments with single components. They have not yet completed systematic experimental verification in real and complex in vivo physiological environments, nor have they constructed standardized detection procedures and accurate quantitative correlation models between signal and blood glucose concentration. They cannot solve practical problems such as physiological interference, signal acquisition synchronization, and data stability in in vivo experiments. As a result, the application of passive UHF RFID technology in the field of non-invasive blood glucose monitoring remains at the theoretical exploration stage. It has not formed a practical and reusable technical method, nor can it provide effective experimental evidence and methodological support for the translation of this technology into human clinical practice.

[0005] In summary, to address the problems of existing blood glucose concentration detection technologies, this paper proposes a non-invasive blood glucose concentration detection method based on passive UHF RFID. This method aims to achieve non-invasive, accurate, and stable blood glucose concentration detection while balancing hardware cost and ease of operation. It lays the methodological foundation for the transformation of this technology into non-invasive clinical blood glucose monitoring and is a technical challenge that urgently needs to be solved in this field. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a non-invasive blood glucose concentration detection method based on passive UHF RFID, which solves the problems mentioned in the background section.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a non-invasive blood glucose concentration detection method based on passive UHF RFID, comprising the following steps: S1. A passive UHF RFID tag is fixed to a preset monitoring site on the body surface of a live experimental animal, and the live experimental animal is placed in the wireless signal reading area constructed by the RFID reader and its antenna. S2. Exogenous blood glucose regulation is applied to the live experimental animal to induce dynamic changes in its blood glucose concentration that cover the normal blood glucose concentration range. S3. During the dynamic change of blood glucose concentration, the backscatter signal strength value of the passive UHF RFID tag is read at a preset working frequency by an RFID reader at fixed time intervals. S4. At the same time point as reading the RFID signal, the real-time blood glucose concentration value of the live experimental animal is obtained simultaneously using a traditional invasive blood glucose measurement method as a reference true value. S5. Collect the backscatter signal intensity values ​​and the corresponding blood glucose concentration reference values ​​for a complete monitoring period to form the original dataset; S6. Preprocess the original dataset by first removing outlier sample points, and then using cubic spline interpolation to obtain the corrected dataset. S7. Based on the regularized dataset, establish a four-parameter Logistic model to quantitatively correlate the backscattered signal intensity of the passive UHF RFID tag at the live monitoring site with the blood glucose concentration in the body, and convert the real-time backscattered signal intensity value into the corresponding blood glucose concentration estimate and display it.

[0008] Preferably, the live experimental animal in step S1 is a mouse, the preset monitoring site is the mouse tail, the antenna of the RFID reader is fixed directly above the mouse, and the radiation field of the antenna can effectively activate the passive UHF RFID tag on the mouse tail.

[0009] Preferably, the exogenous blood glucose regulation in step S2 is achieved by increasing the blood glucose concentration of mice through intraperitoneal injection of glucose solution and decreasing the blood glucose concentration of mice through intraperitoneal injection of insulin solution, and the blood glucose concentration of mice after regulation covers the concentration range of 1.6-7.9 mmol / L within 3 hours.

[0010] Preferably, in step S3, the RFID reader reads the backscatter signal strength multiple times at each observation point to obtain a sequence of signal strength values, and takes the statistical average of the sequence as the final recorded value of the backscatter signal strength at that test point.

[0011] Preferably, the number of consecutive reads at the test point is at least 300.

[0012] Preferably, the conventional invasive blood glucose measurement method in step S4 is the tail tip blood sampling method, in which tail tip blood is sampled immediately after a single RFID signal reading and the real-time blood glucose concentration is measured by a blood glucose meter.

[0013] Preferably, in step S7, the four-parameter Logistic model uses the backscattered signal intensity value RSSI as the independent variable and the predicted blood glucose concentration value Glucose as the dependent variable. The model parameters A, B, C, and D are determined by experimental data from a fitted dataset, and the blood glucose concentration fitting range established by the model is 1.6-9 mmol / L.

[0014] Preferably, before step S1, there is an experimental preparation step: debugging the RFID reader, antenna and passive UHF RFID tag, and setting data acquisition parameters; selecting healthy mice and fasting them for 12 hours to eliminate the interference of food metabolism on blood glucose concentration; and administering intraperitoneal injection anesthesia to the fasted mice, with the anesthesia dose being appropriate to keep the mice quiet and without obvious activity.

[0015] Preferably, the passive UHF RFID tag is an NXP-U8 / U9-9640 model passive tag, and the passive UHF RFID tag is completely attached to the monitoring part of the live experimental animal's body surface without gaps or looseness.

[0016] This invention provides a non-invasive blood glucose concentration detection method based on passive UHF RFID. It has the following beneficial effects: 1. This invention enables non-invasive blood glucose concentration detection. Compared with traditional invasive blood sampling methods, no invasive procedures are required on experimental animals throughout the process, fundamentally avoiding the pain and infection risks associated with blood sampling. It also eliminates the additional costs associated with the use of consumables, significantly improving the safety and convenience of the blood glucose monitoring process, and is more suitable for conducting long-term, continuous blood glucose monitoring research.

[0017] 2. This invention constructs a standardized and systematic blood glucose detection process, from the preliminary preparation of experimental animals and exogenous blood glucose regulation, to the synchronous acquisition of signals and blood glucose reference values, and then to the preprocessing of datasets and the establishment of quantitative models. Each step is closely connected and the operation is standardized, which effectively suppresses various random interferences in in vivo experiments, ensures the reliability and stability of detection data, and also provides a repeatable and referable technical paradigm for subsequent related research in this field.

[0018] 3. This invention relies on passive UHF RFID technology to build a detection system. The hardware used is inexpensive, readily available, and easy to integrate and build, without the need for complex supporting facilities. At the same time, it establishes a precise correlation between signal strength and blood glucose concentration through a four-parameter Logistic model, achieving efficient and accurate detection of blood glucose concentration. It has the advantages of low cost and high detection performance, and also lays a solid methodological foundation for the transformation of this technology into non-invasive blood glucose monitoring in human clinical practice. It has good research and application transformation value. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the overall structure of the non-invasive blood glucose concentration monitoring system of the present invention; Figure 2 This is a schematic diagram of the fitting curve of the four-parameter Logistic model between the backscattered signal intensity of the RFID tag and the blood glucose concentration in this invention. Figure 3 This diagram illustrates the fitting of the four-parameter Logistic model for the quantitative correlation between the backscattered signal intensity of the RFID tag and blood glucose concentration, and the verification using an independent test set.

[0020] In the image: 1. Antenna; 2. RFID reader; 3. Display terminal; 4. Mouse; 5. Electronic tag; 6. Absorbing material; 7. Tripod; 8. Mouse tail. Detailed Implementation

[0021] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Please see the appendix Figure 1 - Appendix Figure 3 This invention provides a non-invasive blood glucose concentration detection method based on passive UHF RFID.

[0023] The detection system used in this embodiment consists of two parts: hardware devices and a data processing unit. The hardware devices include an RFID reader, a circularly polarized antenna, a passive UHF RFID tag, a blood glucose meter, and matching blood collection consumables. The data processing unit is equipped with programs for data acquisition, preprocessing, model fitting, and result output, enabling simultaneous reception, analysis, and visualization of signal strength and blood glucose concentration data. Specifically, the RFID reader is a professional reader adapted to the UHF band, paired with a Laird-S9028PCR circularly polarized antenna. This antenna provides uniform signal coverage and strong stability for tag activation. The passive UHF RFID tag is the NXP-U8 / U9-9640 model. This passive wet tag has good adhesion to biological surfaces, high dielectric environment response sensitivity, and can accurately capture changes in the dielectric properties of the surrounding medium. The blood glucose meter is a high-precision blood glucose testing device commonly used in clinical settings, along with disposable consumables such as lancets and blood glucose test strips, ensuring the accuracy of invasive blood glucose reference value measurements.

[0024] I. Pre-experimental preparation Equipment debugging and parameter setting: First, connect the RFID reader to the circularly polarized antenna and check the continuity of the connection lines. After turning on the device, adjust the frequency band and acquisition parameters. Set the operating center frequency of the RFID reader to 922.625MHz, which is the optimal operating frequency band for passive UHF RFID, effectively reducing signal interference and improving signal acquisition stability. Set the signal acquisition interval to 5 minutes, and perform at least 300 consecutive reads at each test point to ensure that the signal strength value at each test point is statistically significant. Simultaneously, check the performance of the passive UHF RFID tags, discarding damaged or non-signal-feedback tags. Place qualified tags in the antenna signal coverage area and test the backscatter signal feedback status of the tags to ensure stable signal strength values ​​without significant drift. In addition, debug the blood glucose meter, perform zeroing, and check the expiration date of blood collection consumables to ensure that the invasive blood glucose measurement device is in normal working condition.

[0025] Animal selection and pretreatment: Healthy adult mice were selected as experimental models. The mice were free of underlying conditions such as diabetes and metabolic diseases, and were in good physiological condition. Mice were fasted for 12 hours before the experiment, and only drinking water was provided during this period. This eliminated the interference of food metabolism in the stomach on blood glucose concentration, ensuring that the mice's blood glucose levels were stable at the beginning of the experiment and avoiding fluctuations in blood glucose caused by food digestion and absorption that could affect the accuracy of the experimental data.

[0026] Anesthesia of laboratory animals: Mice were anesthetized via intraperitoneal injection after fasting. The anesthetic agent used was one conforming to animal experimental standards. The dosage was strictly controlled to maintain the mice's quiet state with minimal activity and stable physiological signs. Insufficient dosage would cause agitation, tag displacement, and signal acquisition interruption; excessive dosage would affect the mice's normal metabolism and interfere with natural blood glucose fluctuations. Therefore, the dosage needed to be accurately calculated based on the mice's weight to ensure adequate anesthesia. Simultaneously, the mice's respiration, heart rate, and other physiological signs were monitored in real time during the experiment, and any abnormalities were addressed promptly.

[0027] II. Experimental Model Construction RFID Tag Attachment: Anesthetized mice are placed on the experimental table. The mouse tail is cleaned to remove dirt and hair from the skin surface, ensuring it is smooth and dry to prevent impurities from affecting the tag's adhesion. A passive UHF RFID tag is then firmly attached to the middle of the mouse's dorsal tail. This area has thin skin, rich microcirculation, and rapid reflection of changes in blood glucose concentration in the skin's dielectric environment. Furthermore, the limited mobility in this area effectively prevents the tag from falling off or shifting during the experiment. During attachment, ensure complete adhesion between the tag and the mouse's tail skin, without gaps or looseness. After attachment, gently press the edges of the tag to enhance its adhesion stability.

[0028] Signal Reading Area Construction: Mice with RFID tags attached were placed in a pre-planned experimental area. The circularly polarized antenna of the RFID reader was fixed on a tripod directly above the mice. The vertical distance between the antenna and the mouse's tail was adjusted to ensure that the antenna's radiation field could effectively activate the passive UHF RFID tag on the mouse's tail, and that the tag's backscattered signal could be stably received by the antenna without significant signal attenuation. The experimental area was selected in an open environment free from electromagnetic interference, away from other wireless signal transmitting equipment, metal obstacles, etc., to avoid external electromagnetic interference affecting the RFID signal acquisition accuracy. At the same time, the temperature and humidity of the experimental area were kept constant to reduce the interference of environmental factors on the mice's physiological state and tag signal response.

[0029] III. Exogenous blood glucose regulation Exogenous blood glucose regulation was performed on mice placed in the signal reading area by alternating intraperitoneal injection of glucose solution and insulin solution to achieve dynamic regulation of blood glucose concentration in mice. The blood glucose concentration of mice underwent a complete dynamic change process from low to high and back to low within a monitoring period of about 3 hours. The range of blood glucose concentration change covered the normal human blood glucose concentration range of 1.6-7.9 mmol / L, and also covered the normal blood glucose concentration range of mice.

[0030] When injecting glucose solution, the dosage was precisely calculated based on the mouse's weight, gradually increasing the blood glucose concentration. Once the blood glucose concentration reached a preset peak, insulin solution was injected to gradually lower the blood glucose concentration through insulin's hypoglycemic effect. The injection process followed a slow injection principle to avoid sudden spikes and drops in blood glucose concentration caused by rapid drug injection, ensuring a smooth and continuous dynamic change in blood glucose concentration that better reflects physiological blood glucose fluctuations. Simultaneously, the mice's physiological signs were continuously monitored throughout the blood glucose regulation process to ensure they could tolerate the dynamic changes in blood glucose concentration without significant physiological abnormalities.

[0031] IV. Synchronous acquisition of signal and blood glucose concentration Throughout the entire monitoring period of dynamic changes in mouse blood glucose concentration, the backscatter signal intensity of RFID tags and the real-time blood glucose concentration of mice were synchronously collected at a pre-set 5-minute time interval. This ensured that the signal intensity value of each test point corresponded one-to-one with the blood glucose concentration reference value, providing an accurate paired dataset for subsequent data modeling.

[0032] RFID signal acquisition: At each test point, the backscatter signal of the mouse tail tag is read continuously at least 300 times by an RFID reader to obtain a set of signal strength (RSSI) value sequences. The reader automatically records all data in the sequence. Then, the data processing unit performs statistical analysis on the set of data and calculates its arithmetic mean. The average value is used as the final recorded value of the backscatter signal strength of the test point. By taking the average value after multiple readings, the random error in the signal acquisition process is effectively reduced, and the stability and reliability of the signal strength data are improved.

[0033] Blood glucose concentration reference value acquisition: After completing RFID signal acquisition at each test point, the real-time blood glucose concentration of the mouse is immediately obtained using the tail tip blood sampling method, which serves as the reference true value for the blood glucose concentration at that test point. During blood sampling, a small amount of blood is squeezed out by gently pricking the tip of the mouse's tail with a lancet. The blood glucose test strip is then brought into contact with the blood, and the blood glucose concentration value is quickly measured and recorded using a blood glucose meter. After the blood sampling operation is completed, simple hemostasis and disinfection are performed on the blood sampling site at the mouse's tail tip to avoid wound infection. The entire blood sampling process is rapid, ensuring high synchronization with the test point of RFID signal acquisition, and minimizing blood glucose concentration deviation caused by time difference.

[0034] V. Construction of the Original Dataset After a complete 3-hour monitoring cycle, the final recorded values ​​of backscattered signal intensity at all test points during the monitoring process, along with their corresponding true reference blood glucose concentrations, were collected. The two sets of data were paired according to the testing order to construct the original dataset. The original dataset underwent preliminary processing, with each data point labeled with its acquisition time, signal intensity value, and blood glucose concentration reference value. Invalid data points resulting from operational errors, equipment malfunctions, or other factors were removed to ensure the completeness and validity of the original dataset. Simultaneously, preliminary statistical analysis was performed on multiple signal intensity values ​​within the same blood glucose concentration range to observe the data distribution characteristics, providing a foundation for subsequent data preprocessing.

[0035] VI. Dataset Preprocessing To improve the accuracy and reliability of subsequent model fitting, the original dataset needs to be professionally preprocessed. First, outlier samples are removed, and then a normalized dataset is obtained through interpolation. The specific steps are as follows: Outlier Removal: Statistical methods were used to analyze the sample points in the original dataset to identify and remove outlier points that significantly interfered with the fitting effect. These outlier points mainly included abrupt changes in signal intensity caused by tag displacement due to mouse agitation, abnormal blood glucose concentration values ​​due to blood collection errors, and signal intensity drift caused by external electromagnetic interference. The removal process adhered to objective and rigorous principles, combining experimental operation records for comprehensive judgment to avoid mistakenly deleting valid data points and ensure that the remaining sample points truly reflected the intrinsic relationship between signal intensity and blood glucose concentration.

[0036] Cubic spline interpolation: Since the distribution of sample points in the original dataset within the blood glucose concentration range may be uneven, cubic spline interpolation is used to improve the smoothness and accuracy of the model fitting after outlier removal. With blood glucose concentration as the independent variable and signal intensity as the dependent variable, uniformly distributed interpolation points are inserted within the blood glucose concentration fitting range of 1.6-9 mmol / L. Adjacent sample points are fitted using a cubic spline function, and the estimated signal intensity corresponding to each interpolation point is calculated. This results in a well-formed dataset where blood glucose concentration and signal intensity values ​​are uniformly correlated, providing a high-quality data foundation for the subsequent establishment of a quantitative correlation model.

[0037] Model Fitting: Based on the preprocessed regularized dataset, using RFID tag backscattered signal intensity (RSSI) as the independent variable and blood glucose concentration as the dependent variable, a four-parameter logistic model was used for curve fitting within the blood glucose concentration range of 1.6-9 mmol / L to construct a quantitative correlation model between signal intensity and blood glucose concentration. The four-parameter logistic model is as follows:

[0038] Wherein, RSSI represents the backscattered signal intensity value, Glucose represents the predicted blood glucose concentration value, and A, B, C, and D are model parameters. By fitting the experimental data to a standardized dataset, the specific values ​​of the four model parameters are determined using the fitting algorithm of the data processing unit. This ensures that the fitted curve closely matches the experimental data points, accurately reflecting the nonlinear relationship between signal intensity and blood glucose concentration. The coefficient of determination R² of the four-parameter Logistic model is ≥0.97, and the root mean square error of the model's prediction for unknown blood glucose samples is ≤0.4 mmol / L.

[0039] VII. Establishment and Validation of Quantitative Correlation Model Model Validation: To evaluate the prediction accuracy, robustness, and generalization performance of the established four-parameter Logistic model, the regularized dataset was randomly divided into a modeling set and an independent test set. The modeling set was used to determine the model parameters, while the independent test set was used to validate the model's predictive ability for unknown samples. The coefficient of determination R0 of the model on the modeling set was calculated. 2 The model's prediction error was assessed based on the root mean square error on the independent test set, and according to the ISO-15197:2013 standard for accuracy of clinical blood glucose monitoring systems, to verify whether the model's prediction results met the clinically acceptable error range. If the model's coefficient of determination R... 2 A higher value, a smaller root mean square error, and a prediction error that fully meets clinical accuracy standards indicate that the established quantitative correlation model has good prediction accuracy and robustness, and can be used to convert real-time read signal intensity values ​​into corresponding blood glucose concentration estimates.

[0040] 8. Real-time blood glucose concentration detection and display Based on a validated four-parameter Logistic quantitative correlation model, a "blood glucose concentration-RSSI" correlation mapping system is established in the data processing unit. The backscattered signal intensity values ​​of the tags, read in real time by the RFID reader, are substituted into the quantitative correlation model, and the corresponding blood glucose concentration estimate is quickly obtained through model calculation. The data processing unit displays and records the blood glucose concentration estimate in real time, and the dynamic change process of blood glucose concentration can be visualized in various forms such as numerical values ​​and curves. Researchers can monitor the blood glucose concentration trend of the tested subjects in real time through the display interface, achieving non-invasive, real-time, and continuous monitoring of blood glucose concentration.

[0041] If long-term blood glucose monitoring is required, the hardware of this testing system can be miniaturized and integrated. Passive UHF RFID tags can be attached to preset monitoring sites on the body surface of the test subject. The RFID reader and antenna adopt a portable design to achieve mobile signal acquisition. The data processing unit can be connected to a mobile terminal to complete real-time data transmission, analysis and display, meeting the needs of long-term and continuous blood glucose monitoring.

[0042] To more intuitively evaluate the model's predictive performance and clinical conformity at specific blood glucose concentration points, the test set blood glucose reference values ​​and their corresponding model predicted values ​​and prediction errors are listed in the table below, and analyzed according to the clinical blood glucose monitoring system accuracy standard ISO-15197:2013. This standard requires that when the blood glucose concentration is <5.55 mmol / L, the detection error should meet ±0.83 mmol / L; when the blood glucose concentration is ≥5.55 mmol / L, the detection error should meet ±15%.

[0043]

[0044] Table 1 As shown in Table 1 above, the prediction errors of the model for all five independent test samples are within clinically acceptable ranges, fully meeting the requirements of the ISO-15197:2013 standard. This further confirms from a point-to-point accuracy perspective that the quantitative model established based on the method of this invention has high reliability and clinical applicability in its prediction results, providing solid experimental data support for the translational application of non-invasive blood glucose monitoring technology.

[0045] This invention addresses the limitations of existing non-invasive blood glucose monitoring technologies in terms of accuracy, stability, cost, and complexity. It proposes a non-invasive blood glucose concentration monitoring method and system based on passive ultra-high frequency RFID technology. This method utilizes the impedance detuning effect of RFID tag antennas in the dielectric environment of living tissue. By establishing a quantitative correlation model between the tag's backscattered signal intensity and in vivo blood glucose concentration, it achieves non-destructive, real-time, and continuous monitoring of blood glucose levels in experimental animals.

[0046] The advantages of this approach are its clear implementation principle, low hardware cost, and ease of system construction. It can effectively overcome the pain points of traditional invasive monitoring and the limitations of existing non-invasive technologies, and is particularly suitable for long-term, continuous blood glucose trend monitoring scenarios. At the same time, this method provides key experimental evidence and methodological groundwork for future non-invasive blood glucose monitoring applications in humans, and has significant research value and translational potential.

[0047] This invention enables highly sensitive, wireless, and passive monitoring of changes in in vivo blood glucose concentration. Through a passive RFID tag attached to the body surface, the system continuously acquires backscattered signals reflecting changes in tissue dielectric properties. Using a pre-established quantitative model, the signal intensity is converted into an estimated blood glucose concentration, allowing researchers or future users to monitor blood glucose dynamics in real time via a monitoring interface, thus improving the convenience and acceptability of monitoring. Compared to traditional finger-prick blood sampling methods, this invention is completely non-invasive and painless, avoiding the risk of infection and the burden of consumables. Compared to other non-invasive monitoring technologies, this system has advantages such as low cost, easy integration, and the ability to achieve continuous monitoring, providing a promising new technological path for health monitoring fields such as diabetes management.

Claims

1. A non-invasive blood glucose concentration detection method based on passive UHF RFID, characterized in that, Includes the following steps: S1. A passive UHF RFID tag is fixed to a preset monitoring site on the body surface of a live experimental animal, and the live experimental animal is placed in the wireless signal reading area constructed by the RFID reader and its antenna. S2. Exogenous blood glucose regulation is applied to the live experimental animal to induce dynamic changes in its blood glucose concentration that cover the normal blood glucose concentration range. S3. During the dynamic change of blood glucose concentration, the backscatter signal strength value of the passive UHF RFID tag is read at a preset working frequency by an RFID reader at fixed time intervals. S4. At the same time point as reading the RFID signal, the real-time blood glucose concentration value of the live experimental animal is obtained simultaneously using a traditional invasive blood glucose measurement method as a reference true value. S5. Collect the backscatter signal intensity values ​​and the corresponding blood glucose concentration reference values ​​for a complete monitoring period to form the original dataset; S6. Preprocess the original dataset by first removing outlier sample points, and then using cubic spline interpolation to obtain the corrected dataset. S7. Based on the regularized dataset, establish a four-parameter Logistic model to quantitatively correlate the backscattered signal intensity of the passive UHF RFID tag at the live monitoring site with the blood glucose concentration in the body, and convert the real-time backscattered signal intensity value into the corresponding blood glucose concentration estimate and display it.

2. The non-invasive blood glucose concentration detection method based on passive UHF RFID according to claim 1, characterized in that, The live experimental animal in step S1 is a mouse, the preset monitoring site is the mouse tail, the antenna of the RFID reader is fixed directly above the mouse, and the radiation field of the antenna can effectively activate the passive UHF RFID tag on the mouse tail.

3. The non-invasive blood glucose concentration detection method based on passive UHF RFID according to claim 1, characterized in that, The exogenous blood glucose regulation in step S2 involves increasing the blood glucose concentration of mice by intraperitoneal injection of glucose solution and decreasing the blood glucose concentration of mice by intraperitoneal injection of insulin solution, and the blood glucose concentration of mice after regulation covers the concentration range of 1.6-7.9 mmol / L within 3 hours.

4. The non-invasive blood glucose concentration detection method based on passive UHF RFID according to claim 1, characterized in that, In step S3, the RFID reader reads the backscatter signal strength multiple times at each test point to obtain a sequence of signal strength values, and takes the statistical average of the sequence as the final recorded value of the backscatter signal strength at that test point.

5. The non-invasive blood glucose concentration detection method based on passive UHF RFID according to claim 4, characterized in that, Each test point is read continuously at least 300 times.

6. The non-invasive blood glucose concentration detection method based on passive UHF RFID according to claim 1, characterized in that, The traditional invasive blood glucose measurement method in step S4 is the tail tip blood sampling method, which involves immediately sampling the tail tip after completing a single RFID signal reading and measuring the real-time blood glucose concentration using a blood glucose meter.

7. The non-invasive blood glucose concentration detection method based on passive UHF RFID according to claim 1, characterized in that, The four-parameter Logistic model in step S7 uses the backscattered signal intensity value RSSI as the independent variable and the blood glucose concentration prediction value Glucose as the dependent variable. The model parameters A, B, C, and D are determined by experimental data from a fitted and integrated dataset, and the blood glucose concentration fitting range established by the model is 1.6-9 mmol / L.

8. The non-invasive blood glucose concentration detection method based on passive UHF RFID according to claim 1, characterized in that, Before step S1, the experiment preparation steps are also included: debugging the RFID reader, antenna and passive UHF RFID tag, and setting data acquisition parameters; selecting healthy mice and fasting them for 12 hours to eliminate the interference of food metabolism on blood glucose concentration; and administering intraperitoneal injection anesthesia to the fasted mice, with the anesthesia dose being appropriate to keep the mice quiet and without obvious activity.

9. The non-invasive blood glucose concentration detection method based on passive UHF RFID according to claim 1, characterized in that, The passive UHF RFID tag is an NXP-U8 / U9-9640 model passive tag. The passive UHF RFID tag fits perfectly against the monitoring area on the surface of the live experimental animal, without gaps or looseness.

Images