Low-concentration CEA concentration determination method and system based on microwave biosensor

Abstract

The invention discloses a low-concentration CEA concentration determination method and system based on a microwave biosensor, and the method comprises the steps: designing a sensor structure by using full-wave electromagnetic simulation software, carrying out performance simulation, and determining the structure and size of the sensor; according to the determined structure and size of the sensor, a sensor object is prepared by using photoetching and wet etching methods, and an SMA joint is assembled on the sensor object; the vector network analyzer is connected with the sensor real object through the SMA connector, and the performance of the sensor real object is verified; the method comprises the following steps: firstly, dropwise adding an antibody to a sensor real object meeting requirements for incubation, after the antibody is incubated for several minutes, dropwise adding a plurality of groups of low-concentration CEA solutions to the sensor real object with the incubated antibody for incubation respectively, obtaining sensor response data, and establishing a response relationship between the CEA solution concentration and sensor parameters; according to the response relation, the concentration of the CEA solution is rapidly measured, the influences of the two parameters of frequency and amplitude are integrated, and the sensitivity and accuracy of low-concentration detection are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of microwave biosensor detection, and in particular to a method and system for determining low-concentration CEA concentration based on a microwave biosensor. Background Art

[0002] Cancer, also known as malignant tumor, is a disease that seriously threatens human health and life. Cancer is the second leading cause of death after cardiovascular disease. Early diagnosis and intervention of cancer is an important way to improve the survival rate of patients. Therefore, establishing a rapid, accurate and sensitive detection of tumor cells is of great significance for the detection, diagnosis and treatment of cancer. Among them, the detection of tumor markers can be used for auxiliary diagnosis of cancer, distinguishing between malignant and benign conditions, and observing the treatment effect of cancer patients. It plays an important role and has become one of the most widely used detection items in clinical examinations. In the past, the detection of carcinoembryonic antigen (CEA) was used as a specific marker for early clinical diagnosis of lung cancer, colon cancer, etc. Generally speaking, when the content of CEA in human serum exceeds 5ng / mL, it is associated with many types of cancer. The determination of CEA is also of great value in the diagnosis, disease monitoring and efficacy evaluation of gastric cancer, lung cancer, colorectal cancer, breast cancer and other malignant tumors. Therefore, the quantitative and selective detection of CEA is very important for clinical diagnosis.

[0003] In recent years, a variety of immunoassays for CEA have been reported, such as ELISA (enzyme-linked immunosorbent assay), fluorescent immunoassay, and electrochemical immunoassay, and some methods have been applied to commercially available kits. The above immunoassays modify antibodies on solid supports, so that the analytes to be tested can be effectively captured in these assays. ELISA is a rapid, sensitive, and highly reproducible detection method that can be used to detect antigens or antibodies, and can detect substances at extremely low concentrations; it can efficiently detect a variety of antigens, and can also detect a variety of antibodies. Fluorescent immunoassays have high sensitivity and strong specificity. Electrochemical immunoassays have high sensitivity, stable markers, long-lasting luminescence, and no radiation hazards. However, these methods are complicated to operate, require a large amount of reagents, expensive samples, are time-consuming, and have a high possibility of damage, and require professional operators, which is not conducive to rapid detection by non-professionals. Therefore, there is an urgent need for a sensor that is easy to operate, low-cost, and has a fast response speed to quantitatively detect the concentration of CEA.

[0004] Microwave detection technology can achieve non-invasive detection and real-time detection. Microwave sensors are formed by etching different patterns on some substrates. They are mainly divided into three layers. The upper layer is a pattern layer composed of microstrip transmission lines, the middle is a substrate, and the lower layer is copper as grounding. Microwave sensors have the advantages of low cost and small size. When the object under test produces a small disturbance to the microwave, such as changes in position, shape, dielectric constant, etc., the microwave signal will change. Using the perturbation method, the impact of these small changes on the microwave signal can be analyzed, thereby extracting physical quantity information related to the object under test. When liquid is dripped into the sensitive area of ​​the sensor, the electromagnetic field distribution on the surface of the sensor will change, and the response of the sensor will also change. It is specifically manifested in the changes in frequency and amplitude. The properties of the object under test can be inferred based on the changes in these microwave parameters.

[0005] In the prior art, the invention patent with patent publication number CN112964765A discloses an electrochemical immunosensor for detecting CEA and its preparation and application. The invention uses a new direct electrochemical immunosensor and uses a trimetal of AuPtRu to modify hollow nano-organic skeletons (HNMs) for signal amplification. At the same time, combined with the advantages of rG0-TEPA having a large specific surface area and a large number of amino groups, it can be used as a base material to capture more proteins. However, the preparation process of the composite material in the invention is relatively complicated.

[0006] In 2021, the paper "Using Antibody Modified Terahertz Metamaterial Biosensor to Detect Concentration of Carcinoembryonic Antigen" was published in the IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS. It proposed an antibody-modified terahertz metamaterial biosensor for detecting the concentration of carcinoembryonic antigen. The surface CEA concentration is analyzed by changing the resonant frequency of the sensor, but it has not been applied to actual blood testing.

[0007] The invention patent with the patent publication number CN116068188A discloses a HEMT sensor for detecting carcinoembryonic antigen and its preparation method. The invention adopts a curved gate enhanced structure, epitaxy a layer of p-GaN on the AlGaN / GaN heterojunction, and forms a pn junction with the 2DEG under the gate, so that the AlGaN conduction band is raised above the Fermi level, thereby depleting the 2DEG, making the device in a closed state when not working, and realizing power consumption reduction. Multiple curved gates are connected to reduce the transconductance, increase the coverage area of ​​the electric field on the 2DEG channel, and the current near the drain, thereby improving the responsiveness and resolution of the device, but the sensor preparation process is complicated.

[0008] The paper "Resonant Wireless Biosensor for Quantitative Analysis of Carcinoembryonic Antigen" published by Li Yingyin et al. changes the capacitance of the interdigital electrodes according to the different concentrations of CEA antigen, causing the bandwidth value to change, and reads the sensor signal through the mutual inductance coupling of the inductor coil to achieve rapid detection of the concentration of CEA antigen. However, this method uses the change in the capacitance of the interdigital electrodes to cause the change in bandwidth to detect the CEA concentration. The detection range is suitable for the higher concentration range of 1, 10, 100, and 1000 ng / ml, and is not suitable for low concentration detection. Summary of the invention

[0009] The technical problem to be solved by the present invention is to provide a method for measuring low-concentration CEA concentration based on a microwave biosensor.

[0010] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0011] A method for measuring low-concentration CEA concentration based on a microwave biosensor, comprising:

[0012] Use full-wave electromagnetic simulation software to design the sensor structure and perform performance simulation to determine the sensor structure and size;

[0013] According to the determined sensor structure and size, the sensor is prepared by using photolithography and wet etching, and an SMA connector is assembled on the sensor.

[0014] Connect the vector network analyzer to the actual sensor through the SMA connector to detect the response of the actual sensor and verify the performance of the actual sensor.

[0015] First, add antibodies to the sensors that meet the requirements for incubation. After the antibodies have been incubated for several minutes, add multiple groups of low-concentration CEA solutions to the sensors that have been incubated with antibodies, obtain sensor response data, and establish a response relationship between the CEA solution concentration and the sensor parameters.

[0016] According to the established response relationship between the concentration of CEA solution and sensor parameters, the concentration of CEA solution was quickly determined.

[0017] In one embodiment of the present invention, when designing the sensor structure, the quality factor is used as an indicator to evaluate the resolution of the sensor; on the basis of ensuring the resolution, the sensor structure and size are optimized according to the dielectric constant of the CEA solution with the goal of improving the sensitivity of the resonant frequency and the S21 amplitude response.

[0018] In one embodiment of the present invention, the sensor structure includes: a copper base plate 100, an intermediate dielectric substrate 200, an open resonant ring 300 and interdigital electrodes 400; the intermediate dielectric substrate 200 is arranged on the copper base plate 100, and the open resonant ring 300 and the interdigital electrodes 400 are arranged on the intermediate dielectric substrate 200; and the structure of the open resonant ring 300 and the interdigital electrodes 400 constitutes a resonant circuit, which is equivalent to a total capacitor and a total inductor connected in parallel.

[0019] In one embodiment of the present invention, the open resonant ring 300 includes an upper branch 310 of the resonant ring, a lower branch 320 of the resonant ring and a side wall 330 of the resonant ring; the two side walls 330 of the resonant ring respectively connect the two ends of the upper branch 310 of the resonant ring and the lower branch 320 of the resonant ring to form a rectangular shape; and the upper branch 310 of the resonant ring is provided with an opening, and the interdigital electrode 400 is connected to the open resonant ring 300 at the opening; the two side walls 330 of the resonant ring are respectively connected with microstrip lines, the microstrip line on one side thereof serves as an input signal terminal 331, and the microstrip line on the other side serves as an output signal terminal 332; the test area of ​​the sensor is at the interdigital electrode 400 of the concentrated electric field, the signal is input from the input signal terminal 331, the signal generates a resonant signal through the middle resonant circuit, and then outputs through the output signal terminal 332.

[0020] In one embodiment of the present invention, the interdigital electrode 400 includes a first interdigital electrode 410 and a second interdigital electrode 420; the first interdigital electrode 410 is connected to the branch 310 on the resonant ring near the input signal end 331, and the second interdigital electrode 420 is connected to the branch 310 on the resonant ring near the output signal end 332; wherein,

[0021] The first interdigitated electrode 410 includes a first longitudinal interdigitated finger 411, a first transverse interdigitated finger 412 and a second transverse interdigitated finger 413; the roots of the first transverse interdigitated finger 412 and the second transverse interdigitated finger 413 are connected to the middle of the first longitudinal interdigitated finger 411 to form a first resonant interval region;

[0022] The second interdigitated electrode 420 includes a second longitudinal interdigitated finger 421, a third transverse interdigitated finger 422 and a fourth transverse interdigitated finger 423; the roots of the third transverse interdigitated finger 422 and the fourth transverse interdigitated finger 423 are connected to the second longitudinal interdigitated finger 421 to form a second resonant interval region; and the third transverse interdigitated finger 422 is connected to the end of the second longitudinal interdigitated finger 421;

[0023] The second transverse interdigital finger 413 is located in the second resonance interval region, and the fourth transverse interdigital finger 423 is located in the first resonance interval region, forming a first newly added resonance region, a second newly added resonance region, and a third newly added resonance region, and the three newly added resonance regions are connected, and are also connected to the resonance region of the open resonant ring 300;

[0024] In the first resonance interval region and the second resonance interval region, the finger spacing (G) between two adjacent lateral interdigitated fingers is equal.

[0025] In one embodiment of the present invention, the length L1 of the input signal terminal (331) and the input signal terminal (331) ranges from 3.2 to 1.8 mm, and the width W1 ranges from 1.12 to 1.68 mm; the length L2 of the resonance ring side wall (330) ranges from 3.6 to 5.4 mm, and the width W2 ranges from 0.8 to 1.2 mm; the length L3 of the resonance ring lower branch (320) ranges from 7.2 to 10.9 mm, and the width W3 ranges from 0.8 to 1.2 mm; the length L4 of the resonance ring upper branch (310) ranges from 2 to 3 mm, and the width W4 ranges from 0.8 to 1.2 mm;

[0026] The length L5 of the first longitudinal finger 411 and the second longitudinal finger 421 ranges from 3.6 to 5.4 mm, and the width W5 ranges from 0.8 to 1.2 mm; the length L6 of the first transverse finger 412, the second transverse finger 413, the third transverse finger 422 and the fourth transverse finger 423 ranges from 1.2 to 1.8 mm, and the width W6 ranges from 0.56 to 0.84 mm; the finger spacing G ranges from 0.36 to 0.54 mm.

[0027] In one embodiment of the present invention, based on the acquired sensor response data, a linear response relationship between the CEA solution concentration and the resonant frequency and S21 amplitude is established by applying a data fitting method; wherein,

[0028] The linear response relationship between CEA solution concentration and resonant frequency is:

[0029] y=3.483E-5*x+1.485;

[0030] The linear response relationship between CEA solution concentration and S21 amplitude is:

[0031] y=1.029E-4*x+4.345;

[0032] In the formula, E-5 represents 10 to the negative fifth power, and E-4 represents 10 to the negative fourth power.

[0033] In one embodiment of the present invention, the acquired sensor response data is used as a data set; a machine learning method is applied to establish a response relationship between the CEA solution concentration and the resonant frequency and the S21 amplitude as a CEA solution concentration detection model; the data set is used to train the CEA solution concentration detection model, the sensor response data is input, and the CEA solution concentration is output; the trained CEA solution concentration detection model is used for online detection;

[0034] The vector network analyzer is connected to the computing and display terminal, and the trained CEA solution concentration detection model is deployed on the computing and display terminal. At the same time, a data processing platform is built in the computing and display terminal; the data processing platform is enabled to collect and save sensor response data in real time, and to call and display the detection results of the trained CEA solution concentration detection model.

[0035] In one embodiment of the present invention, the low concentration of CEA solution ranges from 0.15 ng / mL to 2.4 ng / mL.

[0036] The present invention also provides a low-concentration CEA concentration determination system based on a microwave biosensor, which uses the above-mentioned low-concentration CEA concentration determination method based on a microwave biosensor, comprising:

[0037] The structural simulation module is used to draw the sensor structure using full-wave electromagnetic simulation software, and to perform performance simulation to determine the sensor structure and size;

[0038] The physical verification module is used to prepare the physical sensor by photolithography and wet etching according to the determined sensor structure and size, and assemble the SMA connector to the physical sensor; connect the vector network analyzer to the physical sensor through the SMA connector, detect the response of the physical sensor, and verify the performance of the physical sensor;

[0039] The response relationship establishment module first drips antibodies onto the sensor that meets the requirements for incubation. After the antibodies have been incubated for several minutes, multiple groups of low-concentration CEA solutions are dripped onto the sensor that has been incubated with antibodies for incubation, and the sensor response data is obtained to establish the response relationship between the CEA solution concentration and the sensor parameters.

[0040] The online concentration determination module performs rapid determination of the concentration of the CEA solution based on the established response relationship between the concentration of the CEA solution and the sensor parameters.

[0041] Compared with the prior art, the beneficial effects of the present invention are as follows: when designing the sensor, the present invention optimizes the sensor structure and size according to the dielectric constant of the CEA solution, with the goal of improving the sensitivity of the resonant frequency and the S21 amplitude response, and uses the quality factor as an indicator to evaluate the resolution of the sensor. The sensor is simulated first, and after reaching the required resolution, the sensor is physically manufactured, and then measured again to verify the measurement performance of the sensor.

[0042] In the sensor structure, the open resonant ring is combined with the interdigital electrode to increase the resonance area and improve the concentration response sensitivity. At the same time, the linear relationship between the CEA solution concentration and the frequency and S21 amplitude response is obtained by combining data fitting. Alternatively, a CEA solution concentration detection model is established based on the actual measured data and combined with machine learning to achieve low-concentration CEA solution measurement.

[0043] Compared with the fluorescent immunoassay method, the sensor preparation of the present invention is simple and the test is convenient. The test process only requires dropping the antibody antigen on the sensor without other operations. The CEA concentration is judged by the change of the sensor frequency and S21 amplitude. The sensor has a fast response speed. After testing, the sensor time should be 108ms. The present invention combines the influence of the two parameters of frequency and amplitude to significantly improve the sensitivity, accuracy and response time of low-concentration detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 The present invention is a flowchart of a method for measuring low-concentration CEA concentration based on a microwave biosensor according to an embodiment of the present invention.

[0045] Figure 2 Schematic diagram of the sensor structure of an embodiment of the present invention.

[0046] 3(a) and (b) are schematic diagrams of relatively sparse and dense electric field distribution of interdigital electrodes according to an embodiment of the present invention.

[0047] Figure 4 Schematic diagram of the S-parameter simulation results of the sensor according to the embodiment of the present invention.

[0048] Figure 5 It is a schematic diagram of simulation results of the response change of the sensor S21 according to an embodiment of the present invention when the dielectric constant of the object to be measured changes.

[0049] Figure 6 It is a schematic diagram of simulation results of the response change of the sensor S21 according to an embodiment of the present invention as the loss tangent of the object to be measured is changed.

[0050] Figure 7 Schematic diagram of sensor testing according to an embodiment of the present invention.

[0051] Figure 8 Schematic diagram of comparison between actual measurement and simulation of the sensor according to an embodiment of the present invention.

[0052] Fig. 9 Schematic diagram of test results of CEA solutions of different concentrations according to an embodiment of the present invention.

[0053] Fig.10 Schematic diagram of the fitting results of the CEA solution concentration and the resonance frequency according to an embodiment of the present invention.

[0054] Fig.11 Schematic diagram of the fitting results of the CEA solution concentration and the resonance amplitude according to an embodiment of the present invention.

[0055] Fig.12 A schematic diagram of building a machine learning algorithm model in an embodiment of the present invention.

[0056] Fig.13 This is a comparison result between the CEA concentration value in the blood predicted by the machine learning model of an embodiment of the present invention and the CEA concentration value actually tested in the hospital.

[0057] Fig.14 Schematic diagram of a sensor real-time monitoring system according to an embodiment of the present invention.

[0058] Fig.15 The block diagram of a low-concentration CEA concentration determination system based on a microwave biosensor according to an embodiment of the present invention. DETAILED DESCRIPTION

[0059] In order to facilitate those skilled in the art to understand the technical solution of the present invention, the technical solution of the present invention is further described in conjunction with the accompanying drawings of the specification.

[0060] The terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0061] See also Figure 1 As shown, the present invention provides a method for measuring low-concentration CEA concentration based on a microwave biosensor, comprising:

[0062] S10, using full-wave electromagnetic simulation software to design the sensor structure and perform performance simulation to determine the sensor structure and size.

[0063] In one embodiment of the present invention, the structure of the sensor is drawn using HFSS full-wave electromagnetic simulation software. The sensor is a dual-port microwave sensor. After the initial structure simulation is obtained, the dimensions of the sensor are optimized to obtain the optimal structural parameters.

[0064] In one embodiment of the present invention, different solutions have different dielectric properties, and their dielectric constants are different. From the perspective of microwaves, objects with different dielectric constants dripped onto the sensor will cause the dielectric constant of the sensor area to change, and then the frequency and S21 amplitude value of the sensor will change. Therefore, when designing the sensor structure, the quality factor is used as an indicator to evaluate the resolution of the sensor. On the basis of ensuring the resolution, according to the dielectric constant of the CEA solution, the sensor structure and size are optimized with the goal of improving the sensitivity of the resonant frequency and the S21 amplitude response.

[0065] See also Figure 1 As shown in FIG3 , in this embodiment, the sensor structure includes: a copper base plate 100, an intermediate dielectric substrate 200, an open resonant ring 300, and interdigital electrodes 400. The intermediate dielectric substrate 200 is disposed on the copper base plate 100, and the open resonant ring 300 and the interdigital electrodes 400 are disposed on the intermediate dielectric substrate 200. Furthermore, the structure of the open resonant ring 300 and the interdigital electrodes 400 constitutes a resonant circuit 3400, which is equivalent to a total capacitor and a total inductor connected in parallel.

[0066] In this embodiment, the open resonant ring 300 includes an upper branch 310 of the resonant ring, a lower branch 320 of the resonant ring, and a side wall 330 of the resonant ring. The two side walls 330 of the resonant ring respectively connect the two ends of the upper branch 310 of the resonant ring and the lower branch 320 of the resonant ring to form a rectangular shape. The upper branch 310 of the resonant ring is provided with an opening, and the interdigital electrode 400 is connected to the open resonant ring 300 at the opening. The two side walls 330 of the resonant ring are respectively connected with microstrip lines, the microstrip line on one side thereof serves as an input signal terminal 331, and the microstrip line on the other side serves as an output signal terminal 332; the test area of ​​the sensor is at the interdigital electrode 400 of the concentrated electric field, the signal is input from the input signal terminal 331, the signal generates a resonant signal through the middle resonant circuit, and then outputs through the output signal terminal 332.

[0067] In this embodiment, the interdigitated electrode 400 includes a first interdigitated electrode 410 and a second interdigitated electrode 420 . The first interdigitated electrode 410 is connected to the branch 310 on the resonant ring near the input signal end 331 , and the second interdigitated electrode 420 is connected to the branch 310 on the resonant ring near the output signal end 332 .

[0068] In this embodiment, the first interdigitated electrode 410 includes a first longitudinal interdigitated finger 411, a first transverse interdigitated finger 412 and a second transverse interdigitated finger 413, and the roots of the first transverse interdigitated finger 412 and the second transverse interdigitated finger 413 are connected to the middle of the first longitudinal interdigitated finger 411 to form a first resonant spacing region.

[0069] In this embodiment, the second interdigitated electrode 420 includes a second longitudinal interdigitated finger 421, a third transverse interdigitated finger 422, and a fourth transverse interdigitated finger 423. The roots of the third transverse interdigitated finger 422 and the fourth transverse interdigitated finger 423 are connected to the second longitudinal interdigitated finger 421 to form a second resonant spacing region, and the third transverse interdigitated finger 422 is connected to the end of the second longitudinal interdigitated finger 421.

[0070] In this embodiment, the second transverse interdigital finger 413 is located in the second resonance interval region, and the fourth transverse interdigital finger 423 is located in the first resonance interval region, forming a first newly added resonance region A, a second newly added resonance region B, and a third newly added resonance region C, and the three newly added resonance regions are connected, and are also connected to the resonance region D of the open resonant ring 300, and in the first resonance interval region and the second resonance interval region, the finger spacing G between two adjacent transverse interdigital fingers is equal. The addition of three resonance regions makes the resonance signal more sensitive.

[0071] In one embodiment of the present invention, on the basis of the open resonant ring 300, the forked electrode 400 structure is introduced to further constrain the electric field to form a resonant unit, and the test area of ​​the sensor is the forked electrode 400 where the electric field is concentrated. The finger spacing G between the lateral forks and the finger width of the forked electrode 400 have a significant impact on the electric field distribution and sensitivity of the sensor. A smaller finger spacing G helps to enhance the local electric field, thereby improving the sensitivity of the sensor to the target substance. However, too small a finger spacing G will increase the complexity of the manufacturing process and may affect the mechanical stability of the sensor. Therefore, it is necessary to optimize the finger width and the finger spacing of the forked electrode 400 in the design so that a stronger electric field can be obtained under the premise of ensuring the feasibility of the process. The present invention compares the sizes of two forked electrodes 400 of different sizes, and the comparison diagrams are shown in Figures 3 (a) and 3 (b). It can be seen from the figure that the denser the forks, the better the electric field concentration.

[0072] In one embodiment of the present invention, the copper bottom plate 100 has a thickness of 0.07 mm. The intermediate dielectric substrate 200 is Teflon with a thickness of 0.508 mm. The interdigital electrode structure based on the open resonant ring belongs to the top layer, and the sensor structure is made of copper with a thickness of 0.07 mm.

[0073] In this embodiment, the length and width dimensions of the sensor structure are as follows: the length L1 of the input signal terminal 331 and the input signal terminal 331 ranges from 3.2 to 1.8 mm, and the width W1 ranges from 1.12 to 1.68 mm. The length L2 of the resonant ring sidewall 330 ranges from 3.6 to 5.4 mm, and the width W2 ranges from 0.8 to 1.2 mm. The length L3 of the resonant ring lower branch 320 ranges from 7.2 to 10.9 mm, and the width W3 ranges from 0.8 to 1.2 mm. The length L4 of the resonant ring upper branch 310 ranges from 2 to 3 mm, and the width W4 ranges from 0.8 to 1.2 mm.

[0074] In this embodiment, the length L5 of the first longitudinal interdigit 411 and the second longitudinal interdigit 421 is in the range of 3.6 to 5.4 mm, and the width W5 is in the range of 0.8 to 1.2 mm. The length L6 of the first transverse interdigit 412, the second transverse interdigit 413, the third transverse interdigit 422, and the fourth transverse interdigit 423 is in the range of 1.2 to 1.8 mm, and the width W6 is in the range of 0.56 to 0.84 mm. And the finger spacing G is in the range of 0.36 to 0.54 mm. Specifically, as shown in Table 1.

[0075] Table 1 Sensor dimensions

[0076]

[0077] In this embodiment, the simulation results of the sensor are as follows: Figure 4 As shown, the response curves of S11 and S21 are recorded respectively. It can be seen from the figure that the resonant frequency of the sensor is 4.271 GHz.

[0078] In one embodiment of the present invention, the structure of the split resonant ring 300 and the interdigital electrode 400 constitutes a resonant circuit, which can be equivalent to the overall capacitance C after the improved structure. r and the total inductance L r Its resonant frequency f0 depends on the total capacitance C r and the total inductance L r :

[0079]

[0080] When liquid is added, the overall capacitance is no longer the capacitance of the sensor itself, C r , the introduction of the solution's own capacitance will change the overall capacitance of the circuit, thereby changing the resonant frequency of the sensor.

[0081] In this embodiment, the resonance signal has different resonance frequencies and S21 amplitudes under the disturbance of liquids with different dielectric constants. The present invention studies the influence of the dielectric constant ε and dielectric loss tangent tanδ of the measured object on the resonance frequency and transmission coefficient of the sensor, where the transmission coefficient can be understood as the S21 amplitude. The simulation results are shown in Figure 5 and 6 As shown in the figure, when the loss tangent is constant, the resonant frequency of the sensor decreases as the dielectric constant increases. When the dielectric constant is constant, the transmission coefficient of the sensor increases as the loss tangent increases. The sensitivity s of an empty sensor (a sensor without any liquid added) can be expressed by simulating the change in frequency under a unit dielectric constant change, that is:

[0082]

[0083] In the formula, ΔF represents the frequency shift of the sensor, and Δε represents the change in the dielectric constant of the simulated object. In microwave systems, the interaction between microwaves and the medium is fundamental. The interaction is determined by the complex dielectric constant ε of the medium. r Decide that:

[0084] ε r =ε'-jε";

[0085] Among them, the real part ε' is caused by various displacement polarizations inside the material, representing the energy storage term of the material, and the imaginary part ε" is caused by the various relaxation polarizations caused by the inability of the conversion polarization inside the material to keep up with the high-frequency electric field changes, representing the loss of the material. In addition, ε' and ε" represent the dielectric constant and loss factor, respectively.

[0086] According to the previous simulation results, the resonant frequency and transmission coefficient of the sensor are determined by the dielectric constant of the sample to be tested. Different solutions have different dielectric constants, and the corresponding resonant frequency and transmission coefficient of the sensor are different. For microwave sensors, the quality factor Q is also a very important parameter. The quality factor Q factor of the present invention can be calculated by the following formula:

[0087]

[0088] In the formula, f 3dB2 With f 3dB1 They represent the highest frequency and the lowest frequency respectively when the amplitude is 3dB different from the resonant frequency. Figure 4 This is the HFSS simulation result of the sensor. The resonant frequency of the sensor is 4.271GHz, and the S21 amplitude is -53.27262dB. The highest and lowest frequencies with a 3dB difference from the center frequency amplitude are 4.264GHz and 4.278GHz respectively, so the Q value of the invented sensor is 305. The higher the Q value of the sensor, the higher the resolution of the sensor.

[0089] S20, according to the determined sensor structure and size, use photolithography and wet etching methods to prepare the sensor object, and assemble the SMA connector to the sensor object.

[0090] In one embodiment of the present invention, at least two SMA connectors are respectively welded to the input signal terminal 331 and the output signal terminal 332 .

[0091] S30, connect the vector network analyzer to the actual sensor through the SMA connector to detect the response of the actual sensor and verify the performance of the actual sensor.

[0092] In one embodiment of the present invention, Figure 7 As shown in the figure, after connecting the vector network analyzer, the actual test results are compared with the HFSS simulation results. Figure 8 The simulation results are very close to the measured results, which proves the practicality and reliability of the sensor.

[0093] S40, first dripping antibodies onto the sensor object that meets the requirements for incubation, and after the antibodies have been incubated for several minutes, dripping multiple groups of low-concentration CEA solutions onto the sensor object that has been incubated with antibodies, respectively, to obtain sensor response data, and establish a response relationship between the CEA solution concentration and the sensor parameters.

[0094] In one embodiment of the present invention, a solution is dripped onto the interdigital portion of the interdigital electrode 400 for an experiment. First, use a pipette to drip 2uL of antibody onto the sensor for incubation. After the antibody has been incubated for ten minutes, 2uL of CEA solution with concentrations of 0.15ng / mL, 0.3ng / mL, 0.6ng / mL, 1.2ng / mL, and 2.4ng / mL are dripped onto the sensor incubated with the antibody and incubated for ten minutes to observe the response of the sensor. In this process, after testing each concentration of CEA, the sensor should be washed two to three times with PBS solution before testing the next group to prevent the remaining liquid from the previous group from affecting the experimental results of the next group. Each concentration of CEA solution was tested five times to avoid accidental errors. The sensor response data results of the five different concentrations of CEA solutions are shown as follows. Fig. 9 As shown in the figure, it can be seen that with the increase of CEA concentration, the frequency of the sensor gradually increases and the S21 amplitude value also gradually increases.

[0095] In one embodiment of the invention, based on the acquired sensor response data, a linear response relationship between the CEA solution concentration and the resonant frequency and the S21 amplitude is established by applying data fitting. The linear response relationship between the CEA solution concentration and the resonant frequency is:

[0096] y=3.483E-5*x+1.485;

[0097] The linear response relationship between CEA solution concentration and S21 amplitude is:

[0098] y=1.029E-4*x+4.345;

[0099] In the formula, E-5 represents 10 to the negative fifth power, and E-4 represents 10 to the negative fourth power.

[0100] like Fig.10 As shown in Figure 2, the concentration of CEA solution and the sensor resonance frequency have a good fit of up to 0.999. Fig.11 As shown, the concentration of CEA solution is fitted with the S21 amplitude value, and the fitting degree is as high as 0.997.

[0101] In another embodiment of the present invention, the acquired sensor response data is used as a data set; a machine learning method is used to establish a response relationship between the CEA solution concentration and the resonant frequency and S21 amplitude as a CEA solution concentration detection model; the data set is used to train the CEA solution concentration detection model, the sensor response data is input, and the CEA solution concentration is output; and the trained CEA solution concentration detection model is used for online detection.

[0102] In this embodiment, the machine learning model has certain advantages over the traditional linear fitting method. The traditional linear fitting method is usually based on a simple relationship between two values, such as the relationship between frequency and concentration, to establish a linear formula. This method may be more direct when the data features are simple and the relationship is clear, but when the data features are complex, it is difficult to accurately reflect the nonlinear relationship. The machine learning model can extract multiple feature dimensions from the test data, such as frequency and amplitude. The combination of these two dimensions can provide richer information than a single variable fitting, thereby achieving more accurate concentration prediction. In contrast, the multi-feature fitting of the machine learning model can capture complex nonlinear relationships more flexibly, with higher adaptability and accuracy. In addition, the machine learning model can be continuously optimized after training, becoming more accurate as the amount of data increases, and has good generalization capabilities.

[0103] In one embodiment of the present invention, the machine learning model used in this embodiment is a convolutional neural network. Fig.12 As shown, 80% of the sensor response data tested is used for training. The training process includes first normalizing the data and then extracting the eigenvalues. The eigenvalues ​​are mainly the lowest point of the sensor response curve and the 5 values ​​before and after the lowest point. Then the remaining 20% ​​of the sensor response data is normalized and input into the training model to predict the concentration of the CEA solution. The color depth of the solution in the output result represents the concentration of the CEA solution. The prediction results are used to check the quality of the model.

[0104] In one embodiment of the present invention, the same steps as step S40 are used, except that serum is used instead of different concentrations of CEA solution for testing. This experiment tested five blood samples. The data obtained from the test blood was input into the data model established in step 4 to predict the concentration of CEA in the blood. The prediction results are as follows: Fig.13 As shown in the figure, the bar graph shows the CEA concentration values ​​of five people in the hospital. At the same time, the dotted line data with error bars reflects the CEA concentration predicted by the sensor test combined with the machine learning algorithm. Although there is a certain degree of deviation between the predicted results and the actual values, these deviations are within an acceptable range.

[0105] In one embodiment of the present invention, the vector network analyzer is connected to a computing display terminal, and the trained CEA solution concentration detection model is deployed on the computing display terminal, such as connecting to a computer, and at the same time, a data processing platform is built in the computing display terminal using LabView software to form a sensor system. The data processing platform can collect and save sensor response data in real time, and call and display the detection results of the trained CEA solution concentration detection model. Similarly, the measurement point results obtained by the data fitting method can also be called. The sensor system can assist doctors in making diagnoses, such as Fig.14 As shown in the figure, it provides doctors with intuitive and accurate diagnostic auxiliary information. Such a system design not only improves the accuracy and efficiency of diagnosis, but also provides a new solution for data analysis in the medical field.

[0106] S50, performing a rapid determination of the concentration of the CEA solution according to the established response relationship between the concentration of the CEA solution and the sensor parameters.

[0107] See also Fig.15 As shown, the present invention also provides a low-concentration CEA concentration determination system based on a microwave biosensor, which uses the above-mentioned low-concentration CEA concentration determination method based on a microwave biosensor, including:

[0108] The structure simulation module 10 is used to draw the sensor structure using full-wave electromagnetic simulation software, and to perform performance simulation to determine the sensor structure and size.

[0109] The physical verification module 20 is used to prepare the physical sensor by photolithography and wet etching according to the determined sensor structure and size, and assemble the SMA connector to the physical sensor. Through the SMA connector, the vector network analyzer is connected to the physical sensor to detect the response of the physical sensor and verify the performance of the physical sensor.

[0110] The response relationship establishment module 30 first drips antibodies onto the sensor object that meets the requirements for incubation. After the antibodies have been incubated for several minutes, multiple groups of low-concentration CEA solutions are dripped onto the sensor object incubated with antibodies for incubation, and the sensor response data is obtained to establish the response relationship between the CEA solution concentration and the sensor parameters.

[0111] The online concentration determination module 40 performs a rapid determination of the concentration of the CEA solution according to the established response relationship between the concentration of the CEA solution and the sensor parameters.

[0112] It is obvious to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential features of the present invention. Therefore, the embodiments should be regarded as exemplary and non-limiting from any point of view, and the scope of the present invention is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention, and any reference numerals in the claims should not be regarded as limiting the claims involved.

[0113] The above-described embodiments merely represent implementation methods of the invention. The protection scope of the present invention is not limited to the above-described embodiments. For those skilled in the art, several modifications and improvements may be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A method for measuring low-concentration CEA concentration based on a microwave biosensor, characterized in that: include: Use full-wave electromagnetic simulation software to design the sensor structure and perform performance simulation to determine the sensor structure and size; According to the determined sensor structure and size, the sensor is prepared by using photolithography and wet etching, and an SMA connector is assembled on the sensor. Connect the vector network analyzer to the actual sensor through the SMA connector to detect the response of the actual sensor and verify the performance of the actual sensor. First, add antibodies to the sensors that meet the requirements for incubation. After the antibodies have been incubated for several minutes, add multiple groups of low-concentration CEA solutions to the sensors that have been incubated with antibodies, obtain sensor response data, and establish a response relationship between the CEA solution concentration and the sensor parameters. According to the established response relationship between the concentration of CEA solution and sensor parameters, the concentration of CEA solution was quickly determined.

2. The method for determining low-concentration CEA concentration based on a microwave biosensor according to claim 1, characterized in that: When designing the sensor structure, the quality factor is used as an indicator to evaluate the resolution of the sensor. On the basis of ensuring the resolution, the sensor structure and size are optimized according to the dielectric constant of the CEA solution with the goal of improving the sensitivity of the resonant frequency and S21 amplitude response.

3. The method for determining low-concentration CEA based on a microwave biosensor according to claim 2, characterized in that: The sensor structure comprises: a copper bottom plate (100), an intermediate dielectric substrate (200), an open resonant ring (300) and interdigital electrodes (400); the intermediate dielectric substrate (200) is arranged on the copper bottom plate (100), the open resonant ring (300) and the interdigital electrodes (400) are arranged on the intermediate dielectric substrate (200); and the structure of the open resonant ring (300) and the interdigital electrodes (400) constitutes a resonant circuit, which is equivalent to a total capacitance and a total inductance connected in parallel.

4. The method for determining low-concentration CEA based on a microwave biosensor according to claim 3, characterized in that: The open resonant ring (300) comprises an upper resonant ring branch (310), a lower resonant ring branch (320) and a resonant ring side wall (330); the two resonant ring side walls (330) respectively connect the two ends of the upper resonant ring branch (310) and the lower resonant ring branch (320) to form a rectangular shape; The branch (310) on the resonant ring is provided with an opening, and the interdigital electrode (400) is connected to the open resonant ring (300) at the opening; the two resonant ring side walls (330) are respectively connected with microstrip lines, the microstrip line on one side is used as an input signal end (331), and the microstrip line on the other side is used as an output signal end (332); the test area of ​​the sensor is at the interdigital electrode (400) of the concentrated electric field, the signal is input from the input signal end (331), the signal generates a resonant signal through the middle resonant circuit, and then outputs through the output signal end (332).

5. The method for determining low-concentration CEA concentration based on a microwave biosensor according to claim 4, characterized in that: The interdigital electrode (400) comprises a first interdigital electrode (410) and a second interdigital electrode (420); the first interdigital electrode (410) is connected to a branch (310) on the resonant ring close to an input signal end (331), and the second interdigital electrode (420) is connected to a branch (310) on the resonant ring close to an output signal end (332); wherein, The first interdigitated electrode (410) comprises a first longitudinal interdigitated finger (411), a first transverse interdigitated finger (412) and a second transverse interdigitated finger (413); the roots of the first transverse interdigitated finger (412) and the second transverse interdigitated finger (413) are connected to the middle of the first longitudinal interdigitated finger (411) to form a first resonant interval region; The second interdigitated electrode (420) comprises a second longitudinal interdigitated finger (421), a third transverse interdigitated finger (422) and a fourth transverse interdigitated finger (423); the roots of the third transverse interdigitated finger (422) and the fourth transverse interdigitated finger (423) are connected to the second longitudinal interdigitated finger (421) to form a second resonant interval region; and the third transverse interdigitated finger (422) is connected to the end of the second longitudinal interdigitated finger (421); The second transverse interdigitated finger (413) is located in the second resonance interval region, and the fourth transverse interdigitated finger (423) is located in the first resonance interval region, forming a first newly added resonance region, a second newly added resonance region, and a third newly added resonance region, and the three newly added resonance regions are connected and are also connected to the resonance region of the open resonance ring (300); In the first resonance interval region and the second resonance interval region, the finger spacing (G) between two adjacent lateral interdigitated fingers is equal.

6. The method for determining low-concentration CEA based on a microwave biosensor according to claim 5, characterized in that: The length L1 of the input signal end (331) and the input signal end (331) is in the range of 3.2 to 1.8 mm, and the width W1 is in the range of 1.12 to 1.68 mm; the length L2 of the resonance ring side wall (330) is in the range of 3.6 to 5.4 mm, and the width W2 is in the range of 0.8 to 1.2 mm; the length L3 of the resonance ring lower branch (320) is in the range of 7.2 to 10.9 mm, and the width W3 is in the range of 0.8 to 1.2 mm; the length L4 of the resonance ring upper branch (310) is in the range of 2 to 3 mm, and the width W4 is in the range of 0.8 to 1.2 mm; The length L5 of the first longitudinal interdigital finger (411) and the second longitudinal interdigital finger (421) ranges from 3.6 to 5.4 mm, and the width W5 ranges from 0.8 to 1.2 mm; the length L6 of the first transverse interdigital finger (412), the second transverse interdigital finger (413), the third transverse interdigital finger (422) and the fourth transverse interdigital finger (423) ranges from 1.2 to 1.8 mm, and the width W6 ranges from 0.56 to 0.84 mm; and the finger spacing (G) ranges from 0.36 to 0.54 mm.

7. The method for determining low-concentration CEA based on a microwave biosensor according to claim 1, characterized in that: According to the acquired sensor response data, the linear response relationship between CEA solution concentration and resonant frequency and S21 amplitude is established by using data fitting method; The linear response relationship between CEA solution concentration and resonant frequency is: y=3.483E-5*x+1.485; The linear response relationship between CEA solution concentration and S21 amplitude is: y=1.029E-4*x+4.345; In the formula, E-5 represents 10 to the negative fifth power, and E-4 represents 10 to the negative fourth power.

8. The method for determining low-concentration CEA based on a microwave biosensor according to claim 1, characterized in that: The acquired sensor response data is used as a data set; a machine learning method is used to establish a response relationship between the CEA solution concentration and the resonance frequency and S21 amplitude as a CEA solution concentration detection model; the data set is used to train the CEA solution concentration detection model, the sensor response data is input, and the CEA solution concentration is output; Use the trained CEA solution concentration detection model for online detection; Connect the vector network analyzer to the computing display terminal, deploy the trained CEA solution concentration detection model to the computing display terminal, and build a data processing platform in the computing display terminal; The data processing platform can collect and save sensor response data in real time, and call and display the detection results of the trained CEA solution concentration detection model.

9. The method for determining low-concentration CEA based on a microwave biosensor according to claim 1, characterized in that: The low concentration of CEA solution ranges from 0.15 ng / mL to 2.4 ng / mL.

10. A low-concentration CEA concentration determination system based on microwave biosensor, characterized in that: The method for determining low-concentration CEA concentration based on a microwave biosensor according to any one of claims 1 to 9 comprises: The structural simulation module is used to draw the sensor structure using full-wave electromagnetic simulation software, and to perform performance simulation to determine the sensor structure and size; The physical verification module is used to prepare the physical sensor by photolithography and wet etching according to the determined sensor structure and size, and assemble the SMA connector to the physical sensor; connect the vector network analyzer to the physical sensor through the SMA connector, detect the response of the physical sensor, and verify the performance of the physical sensor; The response relationship establishment module first drips antibodies onto the sensor that meets the requirements for incubation. After the antibodies have been incubated for several minutes, multiple groups of low-concentration CEA solutions are dripped onto the sensor that has been incubated with antibodies for incubation, and the sensor response data is obtained to establish the response relationship between the CEA solution concentration and the sensor parameters. The online concentration determination module performs rapid determination of the concentration of the CEA solution based on the established response relationship between the concentration of the CEA solution and the sensor parameters.

Images