Microwave biosensing real-time monitoring platform suitable for low-medium-concentration CEA in serum

Abstract

The invention discloses a microwave biosensing real-time monitoring platform suitable for low-medium-concentration CEA in serum, and relates to the technical field of microwave biosensing. Quantitative analysis is carried out on serum protein characteristic parameters through the interference evaluation unit, four interference coefficients are obtained, interference state values are calculated in a normalized mode, then interference grading judgment is carried out according to the interference state values to obtain interferent information, grading separation shielding is carried out through the interference separation unit, and the interference quality is improved. The problem that low-concentration CEA microwave response signals are covered due to the fact that serum matrix interference cannot be accurately recognized and eliminated is solved. A sample parameter acquisition module, a serum interference separation module, a microwave monitoring module and a microwave evaluation real-time output module are used for completing microwave sensing monitoring and real-time evaluation output on a serum sample subjected to interference optimization, so that the problems that the sensitivity and specificity of serum low-concentration CEA detection are insufficient and real-time monitoring cannot be realized in the prior art are solved; and real-time microwave biosensing detection is realized.

Description

Technical Field

[0001] This invention relates to the field of microwave biosensing technology, specifically to a microwave biosensing real-time monitoring platform suitable for low concentrations of CEA in serum. Background Technology

[0002] Low serum CEA (carcinoembryonic antigen) is a core broad-spectrum tumor marker in clinical practice. The reference threshold for serum CEA in healthy individuals is typically <5 ng / mL. Low-concentration CEA specifically refers to trace levels ranging from pg / mL to low ng / mL, often observed in early-stage malignant tumors, postoperative micrometastasis and recurrence, and the progression of precancerous lesions. It is a key indicator for early tumor screening and prognostic monitoring. Real-time monitoring using microwave biosensors can dynamically track changes in CEA concentrations after surgery and during chemotherapy, enabling earlier detection of abnormal concentration fluctuations and timely warnings of tumor recurrence and metastasis risks.

[0003] Currently, microwave biosensor real-time monitoring platforms demonstrate advantages such as high sensitivity and rapid response in the field of tumor marker detection. Furthermore, methods for determining low-concentration CEA concentrations based on microwave biosensors exist, achieving rapid concentration measurement by designing sensor structures and establishing the response relationship between CEA solution concentration and sensor parameters. However, serum samples are complex in composition; high-frequency interfering substances such as albumin and globulins, along with non-specific binding proteins, can mask the microwave response signal of low-concentration CEA, leading to decreased detection accuracy. Summary of the Invention

[0004] To address the technical problems raised in the background section, this invention is proposed. Embodiments of this invention provide a microwave biosensing real-time monitoring platform suitable for low concentrations of CEA in serum.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] The sample parameter acquisition module is used to acquire data from serum samples to obtain serum protein characteristic parameters, and then transmit the serum protein characteristic parameters to the serum interference separation module.

[0007] In a preferred embodiment of the present invention, serum protein characteristic parameters are obtained by data collection from serum samples, specifically as follows:

[0008] Serum protein characteristic parameters are obtained by acquiring data from serum samples through the installed sample acquisition sensor group; the sample acquisition sensor group includes, but is not limited to, a dual-frequency interdigital transducer, a planar microwave resonant sensor, a microwave signal generator, and a particle size inversion DSP module; the serum protein characteristic parameters include dual-frequency time-series microwave S-parameter data, microwave dielectric response data, dielectric property synchronization data, and multi-frequency microwave S-parameter data.

[0009] Dual-frequency time-series microwave S-parameter data consists of the transmission microwave time-domain signals of serum samples at low and high frequencies at a set frequency. Microwave dielectric response data is obtained by transmitting a wide-band swept microwave excitation covering the resonant frequency band of the microwave sensing unit, collecting microwave network scattering parameters, and obtaining the dielectric response time-domain sequence through inverse Fourier transform. Dielectric property synchronization data consists of the microwave attenuation coefficient sequence of serum samples within a set sampling period at low frequency, as well as the synchronously acquired resonant frequency offset sequence. Multi-frequency microwave S-parameter data consists of the microwave attenuation coefficients at low, medium, and high frequencies at a set frequency. Based on the Rayleigh scattering model, the microwave attenuation at each frequency is inverted to obtain the particle size range of various interfering proteins. The mean particle size is calculated based on the particle size range, and the ratio of the mass of interfering proteins in each particle size range to the total mass of interfering proteins is also calculated, i.e., the particle size mass ratio. The particle size mean and particle size mass ratio are combined to obtain the multi-frequency S-parameter data. An effective separation particle size range is set.

[0010] The serum interference separation module includes an interference assessment unit and an interference separation unit.

[0011] The interference assessment unit extracts serum protein characteristic parameters based on the sample dataset, then performs interference assessment analysis on the serum protein characteristic parameters to obtain interference information, and transmits it to the interference separation unit.

[0012] In a preferred embodiment of the present invention, interference assessment analysis is performed on serum protein characteristic parameters to obtain interfering information. The specific analysis process is as follows:

[0013] The characteristic parameters of serum proteins were identified to obtain dual-frequency time-series microwave S-parameter data, microwave dielectric response data, dielectric property synchronization data, and multi-frequency S-parameter data.

[0014] The dual-frequency attenuation difference coefficient was obtained by performing characteristic quantitative analysis on dual-frequency time-series microwave S-parameter data; the dielectric interference ratio coefficient was obtained by performing characteristic quantitative analysis on microwave dielectric response data; the coupling interference coefficient was obtained by performing characteristic quantitative analysis on dielectric characteristic synchronization data; and the particle size distribution interference coefficient was obtained by performing characteristic quantitative analysis on multi-frequency S-parameter data.

[0015] Furthermore, characteristic quantification analysis was performed on the dual-frequency time-series microwave S-parameter data to obtain the dual-frequency attenuation difference coefficient, which is as follows:

[0016] The average values ​​of the transmitted microwave signals acquired multiple times at low and high frequencies are obtained based on the dual-frequency time-series microwave S-parameter data. The average values ​​corresponding to low and high frequencies are denoted as low-frequency microwave time-domain signals and high-frequency microwave time-domain signals, respectively. The fractional decay energy corresponding to the dual frequencies of the serum sample is obtained by processing the low-frequency microwave time-domain signals and high-frequency microwave time-domain signals using the set Caputo fractional derivative.

[0017] The formula for calculating the Caputo fractional derivative is: ;

[0018] The formula for calculating attenuation energy is: ;

[0019] in, Microwave time domain signal The q-th order Caputo fractional derivative, where q is the given fractional order; The gamma function is defined as follows: e is the natural constant; Let be the first derivative of the microwave time-domain signal, and t be the real-time time variable of the microwave time-domain signal. For integration time; The fractional decay energy is T; the sampling duration is T.

[0020] The low-frequency fractional-order decay energy and the high-frequency fractional-order decay energy are obtained based on the energy calculation formula and are denoted as follows: Then, the two fractional-order decay energies are input into the set nonlinear difference coefficient calculation formula. The dual-frequency attenuation difference coefficient was calculated. ;in, These are the q-order fractional decay energies of the low-frequency and high-frequency reference signals of the set blank buffer sample, respectively.

[0021] Furthermore, a characteristic quantification analysis of the microwave dielectric response data was performed to obtain the dielectric interference proportion coefficient, which is as follows:

[0022] The dielectric response time-domain sequence is obtained based on microwave dielectric response data, and then substituted into the Gaussian radial basis kernel function to obtain the regenerating kernel function Hilbert space. The dielectric interference ratio coefficient is obtained by analyzing the dielectric response time-domain sequence and the regenerating kernel function Hilbert space based on the orthogonal projection of the set RKHS space algorithm.

[0023] Wherein, the Gaussian radial basis function is It is used to map low-dimensional signals to high-dimensional RKHS space. Let i be the vector of the i-th and j-th sequence sampling points in the dielectric response time-domain sequence. The set kernel width parameter; Let be the Euclidean norm; map the i-th and j-th sequence sampling points to vectors in a high-dimensional RKHS space, and denote them as high-dimensional vectors; form a linear space from all the mapped high-dimensional vectors, and label it as the Hilbert space, denoted as . Similarly, the subspace is obtained based on the time-domain sequence of the dielectric response of a pre-defined blank buffer sample. .

[0024] The RKHS space algorithm is based on the formula... The dielectric interference ratio was calculated. In the formula, It is a time-domain sequence of dielectric response. The reference dielectric response time-domain sequence of the blank buffer sample is given, where n is the sequence sampling point; for arrive Orthogonal projection operator; The norm of the RKHS space algorithm is calculated using the following formula: , This is the inner product of the RKHS algorithm.

[0025] Furthermore, the coupling interference coefficient is obtained by performing characteristic quantization analysis on the dielectric property synchronization data, specifically as follows:

[0026] Based on the dielectric property synchronization data, microwave attenuation coefficient sequences and resonant frequency offset sequences are obtained, and then an empirical distribution function transformation is performed to obtain a uniform distribution sequence, i.e., through the empirical distribution function... Analysis yielded a microwave attenuation uniform distribution sequence. and resonant uniform distribution sequence ;in, Microwave attenuation coefficient sequence The empirical distribution function; For the resonant frequency offset sequence The empirical distribution function, s is the sampling point of the dielectric property synchronization data, and S is the total number of sampling points; These are the independent variables of the set empirical distribution function, which can be arbitrarily selected based on the microwave attenuation coefficient and the resonant frequency offset; This is an indicator function that takes the value 1 when the condition inside the parentheses is true, and 0 otherwise.

[0027] The microwave attenuation uniform distribution sequence and the resonant uniform distribution sequence are input into the Gumbel Copula function. ;in, The joint distribution characteristics of the microwave attenuation uniform distribution value and the resonance uniform distribution value corresponding to the two uniform distribution sequences of the s-th sampling point; These are Copula dependency parameters.

[0028] The optimal copula dependency parameters are obtained by maximum likelihood estimation and analysis: based on the microwave attenuation uniform distribution sequence and the resonance uniform distribution sequence, the preset number of distribution sequence groups are corresponding to the measured { , Substituting each value into the Gumbel Copula function and multiplying all the results, we obtain the likelihood function for the Copula dependency parameters. Then the optimal dependency parameters are calculated. ;in, The number of preset distribution sequence groups.

[0029] Input the optimal dependency parameters into the coupling interference quantization formula The coupling interference coefficient is calculated. .

[0030] Furthermore, the particle size distribution interference coefficient is obtained by performing characteristic quantification analysis on the multi-frequency S-parameter data, specifically as follows:

[0031] Based on multi-frequency S-parameter data, the mean particle size and particle size-to-mass ratio were obtained, and the corresponding effective separation particle size range was acquired. , The minimum and maximum particle sizes were used, and a weighted analysis was performed on the particle size values ​​of each interfering protein and the effective separation particle size range. ,in, is the weighting adjustment factor for the r-th particle size interval; The set window size penalty coefficient; Let be the average particle size of the r-th particle size interval.

[0032] Calculate the weighted mean of particle size based on the weight adjustment factor. Obtain the particle size weighted average ;in, Let R be the percentage of particle size in the r-th particle size range; R is the total number of particle size ranges.

[0033] Calculation formula based on non-uniform disturbance Calculate the output particle size distribution interference coefficient ;in, The sensitivity coefficient is set.

[0034] The dual-frequency attenuation difference coefficient, dielectric interference ratio coefficient, coupling interference coefficient, and particle size distribution interference coefficient are normalized and mapped to the interval [0, 1]. Then, a preset comprehensive coupling formula is used. Calculate and output the disturbance state value gK; where, These correspond to the normalized values ​​of the dual-frequency attenuation difference coefficient, dielectric interference ratio coefficient, coupling interference coefficient, and particle size distribution interference coefficient, respectively. The coupling weight factor is set for each coefficient; the interference state value is divided into a low interference state interval and an interference separation state interval according to the preset state value segmentation value; if the interference state value of the current serum sample is in the low interference state interval, the interference information generated is interference-free information; if the interference state value is in the interference separation state interval, the interference information generated is an interference separation start signal.

[0035] The interference separation unit receives interference information and performs interference separation and shielding to obtain an optimized serum sample. Specifically, if the interference information of the current serum sample is identified as having no interference information, it is directly marked as an optimized serum sample; when the interference information is identified as an interference separation start signal, interference separation and shielding is performed.

[0036] The microwave monitoring module is used to perform microwave sensing monitoring on optimized serum samples to obtain microwave sensing parameters and transmit them to the microwave evaluation real-time output module.

[0037] The microwave assessment real-time output module is used to receive microwave sensing parameters and perform microwave state assessment to obtain a serum CEA status report.

[0038] Optionally, the microwave assessment real-time output module transmits the serum CEA status report to the human-computer interaction unit for display and storage; the human-computer interaction unit includes, but is not limited to, mobile terminals, touch screens, and host computers.

[0039] Compared with the prior art, the beneficial effects of the present invention are:

[0040] 1. This invention uses an interference assessment unit to quantitatively analyze serum protein characteristic parameters, obtain four interference coefficients, and normalize and calculate interference state values. Then, based on the interference state values, interference grading is performed to obtain interference information. Finally, an interference separation unit performs graded separation and shielding, solving the problem that the low-concentration CEA microwave response signal is masked because the serum matrix interference cannot be accurately identified and removed.

[0041] 2. This invention collects serum protein characteristic parameters through a sample parameter acquisition module, a serum interference separation module, a microwave monitoring module, and a microwave evaluation real-time output module. It completes microwave sensing monitoring and real-time evaluation output through the interference-optimized serum sample, solving the problem of insufficient sensitivity and specificity of existing detection of low-concentration CEA in serum, which makes real-time monitoring impossible, and realizing real-time microwave biosensing detection. Attached Figure Description

[0042] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. The following drawings are not drawn to scale according to the actual size, but are intended to illustrate the main idea of ​​the present invention.

[0043] Figure 1 This is a schematic diagram of a microwave biosensor real-time monitoring platform for low concentrations of CEA in serum, according to the present invention.

[0044] Figure 2 This is a schematic diagram illustrating the analysis principle of the interference assessment unit of a microwave biosensor real-time monitoring platform for low concentrations of CEA in serum, as described in this invention. Detailed Implementation

[0045] 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 also within the scope of protection of the present invention.

[0046] As indicated in this invention and the claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" do not specifically refer to the singular and may also include the plural. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of explicitly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0047] While this invention makes various references to certain modules in systems according to embodiments of the invention, any number of different modules can be used and run on user terminals and / or servers. The modules are merely illustrative, and different aspects of the systems and methods may use different modules.

[0048] This invention uses flowcharts to illustrate the operations performed by the system according to embodiments of the invention. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, various steps can be processed in reverse order or simultaneously, as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from them.

[0049] Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. It is obvious that the described embodiments are merely some embodiments of the present invention, and not all embodiments of the present invention, and it should be understood that the present invention is not limited to the exemplary embodiments described herein.

[0050] Please see Figure 1-2 As shown, a microwave biosensor real-time monitoring platform suitable for low concentrations of CEA in serum includes: a sample parameter acquisition module, a serum interference separation module, a microwave monitoring module, and a microwave evaluation real-time output module.

[0051] The sample parameter acquisition module collects data from serum samples to obtain serum protein characteristic parameters, and then transmits these parameters to the serum interference separation module. Specifically, the sample acquisition sensor group collects data from serum samples to obtain serum protein characteristic parameters. The sample acquisition sensor group includes, but is not limited to, a dual-frequency interdigital transducer, a planar microwave resonant sensor, a microwave signal generator, and a particle size inversion DSP module. The serum protein characteristic parameters include dual-frequency time-series microwave S-parameter data, microwave dielectric response data, dielectric property synchronization data, and multi-frequency S-parameter data.

[0052] The dual-frequency time-series microwave S-parameter data consists of the transmitted microwave time-domain signals of serum samples at low and high frequencies set at a predetermined frequency. For example, after a set microwave excitation signal from a low frequency of 50 MHz to a high frequency of 6 GHz is transmitted through an installed channel to sense the serum sample under test, the microwave sensor collects the time-domain and frequency-domain waveforms of the microwave radiation value to obtain the transmitted microwave signal, which reflects the propagation and attenuation characteristics of microwaves in the sample under test. The microwave dielectric response data is obtained by transmitting a wide-band swept-frequency microwave excitation covering the resonant frequency band of the microwave sensing unit, collecting the microwave network scattering parameters, and obtaining the dielectric response time-domain sequence through inverse Fourier transform. The dielectric property synchronization data consists of the microwave attenuation coefficient sequence of serum samples within a set sampling period at low frequency, as well as the synchronous sampling... The resonant frequency offset sequence reflects the continuous change in the dielectric properties of the sample; the multi-frequency microwave S-parameter data consists of microwave attenuation coefficients at low, medium, and high frequencies, for example, the S-parameter attenuation coefficients after passing through a serum sample with microwave excitation signals of 50MHz low frequency, 1GHz medium frequency, and 3GHz high frequency; the particle size ranges of various interfering proteins are obtained by inverting the microwave attenuation at each frequency based on the Rayleigh scattering model, and the average particle size is calculated based on the average particle size range, as well as the ratio of the mass of interfering proteins in each particle size range to the total mass of interfering proteins, i.e., the particle size mass ratio. The average particle size and the particle size mass ratio are combined to obtain the multi-frequency S-parameter data; the effective separation particle size range is defined and denoted as . , These are the minimum and maximum particle sizes, respectively. For example, the particle size of CEA is approximately 3 nm, and the effective separation particle size range is selected as [4 nm, 20 nm].

[0053] The serum interference separation module includes an interference assessment unit and an interference separation unit.

[0054] The interference assessment unit extracts serum protein characteristic parameters based on the sample dataset, then performs interference assessment analysis on the serum protein characteristic parameters to obtain interference information, and transmits it to the interference separation unit.

[0055] The interference assessment and analysis of serum protein characteristic parameters yielded information on interfering substances. The specific analysis procedure is as follows:

[0056] The characteristic parameters of serum proteins were identified to obtain dual-frequency time-series microwave S-parameter data, microwave dielectric response data, dielectric property synchronization data, and multi-frequency microwave S-parameter data.

[0057] Characteristic quantization analysis of dual-frequency time-series microwave S-parameter data yields the dual-frequency attenuation difference coefficient, which is as follows:

[0058] The average values ​​of the transmitted microwave signals acquired multiple times at low and high frequencies are obtained based on the dual-frequency time-series microwave S-parameter data. The average values ​​corresponding to low and high frequencies are denoted as low-frequency microwave time-domain signals and high-frequency microwave time-domain signals, respectively. The fractional decay energy corresponding to the dual frequencies of the serum sample is obtained by processing the low-frequency microwave time-domain signals and high-frequency microwave time-domain signals using the set Caputo fractional derivative.

[0059] The formula for calculating the Caputo fractional derivative is: ;

[0060] The formula for calculating attenuation energy is: ;

[0061] in, Microwave time domain signal The q-th order Caputo fractional derivative, where q is the set fractional order, and its value is calibrated based on the experiment, for example, 0.85 for serum media; The gamma function is defined as follows: e is the natural constant; Let be the first derivative of the microwave time-domain signal, and t be the real-time time variable of the microwave time-domain signal. For integration time; The fractional decay energy; T is the sampling duration, with a value range of... ;

[0062] The low-frequency fractional-order decay energy and the high-frequency fractional-order decay energy are obtained based on the energy calculation formula and are denoted as follows: Then, the two fractional-order decay energies are input into the set nonlinear difference coefficient calculation formula. The dual-frequency attenuation difference coefficient was calculated. ;in, These are the q-order fractional decay energies of the low-frequency and high-frequency reference signals of the set blank buffer sample, respectively.

[0063] Biological serum is a viscoelastic medium in which microwave propagation has memory characteristics. Integer-order attenuation models cannot accurately describe its frequency domain nonlinear attenuation characteristics. However, fractional-order calculus can accurately characterize the microwave propagation memory effect of viscoelastic media, and there are frequency domain differences in the fractional-order attenuation characteristics of interfering proteins and target CEA proteins.

[0064] The dielectric interference proportion coefficient is obtained by performing characteristic quantification analysis on microwave dielectric response data, specifically as follows:

[0065] The dielectric response time-domain sequence is obtained based on microwave dielectric response data, and then substituted into the Gaussian radial basis kernel function to obtain the regenerating kernel function Hilbert space. The dielectric interference ratio coefficient is obtained by analyzing the dielectric response time-domain sequence and the regenerating kernel function Hilbert space based on the orthogonal projection of the set RKHS space algorithm.

[0066] Wherein, the Gaussian radial basis function is It is used to map low-dimensional signals to high-dimensional RKHS space. Let i be the vector of the i-th and j-th sequence sampling points in the dielectric response time-domain sequence. The set kernel width parameter is based on experimental calibration, for example, calibrated to 0.8 for microwave dielectric response; Let be the Euclidean norm; map the i-th and j-th sequence sampling points to vectors in a high-dimensional RKHS space, and denote them as high-dimensional vectors; form a linear space from all the mapped high-dimensional vectors, and label it as the Hilbert space, denoted as . Similarly, the subspace is obtained based on the time-domain sequence of the dielectric response of a pre-defined blank buffer sample. .

[0067] The RKHS space algorithm is based on the formula... The dielectric interference ratio was calculated. In the formula, It is a time-domain sequence of dielectric response. The reference dielectric response time-domain sequence of the blank buffer sample is given, where n is the sequence sampling point; for arrive An orthogonal projection operator is used to extract pure nonspecific interference components from the total dielectric response time-domain sequence; The norm of the RKHS space algorithm is calculated using the following formula: , The inner product of the RKHS algorithm is given by denoted as ...

[0068] By using the RBF kernel function, the low-dimensional nonlinear dielectric response sequence is mapped to the high-dimensional RKHS space, transforming the originally nonlinear superposition of nonspecific interference and specific response into two orthogonal subspaces. , Orthogonal projection operator It can only extract components that match the non-specific response of interfering proteins, separating the specific response of the target CEA from environmental noise; the RKHS spatial algorithm formula is obtained through energy ratio. The weight of non-specific interference in the total dielectric response is quantified. The higher the value, the stronger the non-specific dielectric response of the interfering protein and the stronger its ability to mask low-concentration CEA-specific signals.

[0069] The coupling interference coefficient is obtained by performing characteristic quantization analysis on the dielectric property synchronization data, specifically as follows:

[0070] Based on the dielectric property synchronization data, microwave attenuation coefficient sequences and resonant frequency offset sequences were obtained, and then empirical distribution function transformation was performed to obtain uniform distribution sequences, i.e., through empirical distribution function... Analysis yielded a microwave attenuation uniform distribution sequence. and resonant uniform distribution sequence The value range is [0, 1]; where, Microwave attenuation coefficient sequence The empirical distribution function; For the resonant frequency offset sequence The empirical distribution function, s is the sampling point of the dielectric property synchronization data, and S is the total number of sampling points; These are the independent variables of the set empirical distribution function, which can be arbitrarily selected based on the microwave attenuation coefficient and the resonant frequency offset; This is an indicator function that takes the value 1 when the condition inside the parentheses is true, and 0 otherwise.

[0071] The microwave attenuation uniform distribution sequence and the resonant uniform distribution sequence are input into the Gumbel Copula function. ;in, The joint distribution characteristics of the microwave attenuation uniform distribution value and the resonance uniform distribution value corresponding to the two uniform distribution sequences of the s-th sampling point; The Copula dependency parameter has a range of values. The Gumbel Copula function is used to capture... and Tail dependence: When the concentration of interfering protein increases, the microwave attenuation coefficient and the resonant frequency shift both exhibit extremely high values, and the coupling between the two increases sharply. The θ parameter is used to quantify this nonlinear coupling characteristic.

[0072] The optimal copula dependency parameters are obtained by maximum likelihood estimation and analysis: based on the microwave attenuation uniform distribution sequence and the resonance uniform distribution sequence, the preset number of distribution sequence groups are corresponding to the measured { , Substituting each value into the Gumbel Copula function and multiplying all the results, we obtain the likelihood function for the Copula dependency parameters. Then the optimal dependency parameters are calculated. ;in, The preset number of distribution sequence groups, with a value range of... .

[0073] Input the optimal dependency parameters into the coupling interference quantization formula The coupling interference coefficient is calculated. .

[0074] The particle size distribution interference coefficient was obtained by performing characteristic quantification analysis on multi-frequency S-parameter data, specifically as follows:

[0075] Based on multi-frequency S-parameter data, the mean particle size and particle size-to-mass ratio were obtained, and the corresponding effective separation particle size range was acquired. The particle size values ​​of each interfering protein are weighted and analyzed in relation to the effective separation particle size range. ,in, is the weighting adjustment factor for the r-th particle size interval; The set superwindow penalty coefficient is based on experimental calibration and is set to 5. Let be the average particle size of the r-th particle size interval.

[0076] Calculate the weighted mean of particle size based on the weight adjustment factor. Obtain the particle size weighted average ;in, Let R be the percentage of particle size in the r-th particle size range; R is the total number of particle size ranges.

[0077] Calculation formula based on non-uniform disturbance Calculate the output particle size distribution interference coefficient ;in, The sensitivity coefficient is set, and its value range is [value range missing]. Calibration is based on experiments.

[0078] The dual-frequency attenuation difference coefficient, dielectric interference ratio coefficient, coupling interference coefficient, and particle size distribution interference coefficient are normalized and mapped to the interval [0, 1]. Then, a preset comprehensive coupling formula is used. Calculate and output the disturbance state value gK; where, These correspond to the normalized values ​​of the dual-frequency attenuation difference coefficient, dielectric interference ratio coefficient, coupling interference coefficient, and particle size distribution interference coefficient, respectively. The coupling weight factors assigned to each coefficient are determined based on experimental calibration, for example, based on clinical samples. The interference state value is divided into a low interference state interval and an interference separation state interval according to the preset state value segmentation value. If the interference state value of the current serum sample is in the low interference state interval, the interference information generated is interference-free information; if the interference state value is in the interference separation state interval, the interference information generated is an interference separation start signal.

[0079] The interference separation unit receives interference information and performs interference separation and shielding to obtain an optimized serum sample. Specifically, if the interference information of the current serum sample is identified as having no interference information, it is directly marked as an optimized serum sample. When the interference information is identified as an interference separation initiation signal, interference separation and shielding is performed, which includes, but is not limited to, microwave radiation directional separation mechanism, dual-frequency time-series microwave flow control hierarchical separation mechanism, and micro vortex capture separation mechanism. For example, the microwave radiation directional separation mechanism constructs a standing wave or traveling wave field in the microchannel and uses the principal gradient force of microwave attenuation radiation force to directionally migrate particles of different impedances and sizes to different regions of the channel, realizing the physical separation of interference proteins and target CEA. Large-diameter interference proteins are more strongly affected by microwave radiation force and can be directionally pushed to the waste liquid collection chamber, while small-diameter CEA enters the detection area along the main channel to form an optimized serum sample.

[0080] The microwave monitoring module performs microwave sensing monitoring on optimized serum samples to obtain microwave sensing parameters, and then transmits them to the microwave evaluation real-time output module.

[0081] The microwave assessment real-time output module receives microwave sensor parameters and performs microwave state assessment to obtain a serum CEA status report, which is then transmitted to the human-machine interface unit. The human-machine interface unit includes, but is not limited to, a mobile terminal, a touch screen display, and a host computer.

[0082] Furthermore, those skilled in the art will understand that aspects of the present invention can be described and illustrated through several patentable types or situations, including any new and useful combination of processes, machines, products, or substances, or any new and useful improvements thereof. Accordingly, aspects of the present invention can be implemented entirely by hardware, entirely by software (including firmware, resident software, microcode, etc.), or by a combination of hardware and software. All of the above hardware or software may be referred to as a "data block," "module," "engine," "unit," "component," or "system." Furthermore, aspects of the present invention may be embodied as a computer product located on one or more computer-readable media, the product comprising computer-readable program code.

[0083] Unless otherwise defined, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that terms such as those defined in a common dictionary shall be interpreted as having a meaning consistent with their meaning in the context of the relevant art, and not as having an idealized or highly formalized meaning, unless expressly defined herein.

[0084] The foregoing description is illustrative of the invention and should not be construed as limiting it. Although several exemplary embodiments of the invention have been described, those skilled in the art will readily understand that many modifications can be made to the exemplary embodiments without departing from the novel teachings and advantages of the invention. Therefore, all such modifications are intended to be included within the scope of the invention as defined in the claims. It should be understood that the foregoing description is illustrative of the invention and should not be construed as limiting it to the specific embodiments disclosed, and modifications to the disclosed embodiments and other embodiments are intended to be included within the scope of the appended claims. The invention is defined by the claims and their equivalents.

Claims

1. A microwave biosensor real-time monitoring platform suitable for low concentrations of CEA in serum, characterized in that, include: The sample parameter acquisition module is used to acquire data from serum samples to obtain serum protein characteristic parameters, and then transmit the serum protein characteristic parameters to the serum interference separation module; The serum interference separation module includes an interference assessment unit and an interference separation unit. The interference assessment unit extracts serum protein characteristic parameters based on the sample dataset, then performs interference assessment analysis on the serum protein characteristic parameters to obtain interference information, and transmits it to the interference separation unit. The interference separation unit receives the interference information and performs interference separation and shielding to obtain an optimized serum sample. The microwave monitoring module is used to perform microwave sensing monitoring on optimized serum samples to obtain microwave sensing parameters and transmit them to the microwave evaluation real-time output module. The microwave assessment real-time output module is used to receive microwave sensing parameters and perform microwave state assessment to obtain a serum CEA status report, and then output the serum CEA status report in real time.

2. The microwave biosensing real-time monitoring platform for low concentrations of CEA in serum according to claim 1, characterized in that, The serum protein characteristic parameters are obtained by data acquisition from the serum sample. Specifically, the serum protein characteristic parameters are obtained by data acquisition from the serum sample through the installed sample acquisition sensor group. The sample acquisition sensor group includes, but is not limited to, a dual-frequency interdigital transducer, a planar microwave resonant sensor, a microwave signal generator, and a particle size inversion DSP module. The serum protein characteristic parameters include dual-frequency time-series microwave S-parameter data, microwave dielectric response data, dielectric property synchronization data, and multi-frequency microwave S-parameter data.

3. The microwave biosensing real-time monitoring platform for low concentrations of CEA in serum according to claim 1, characterized in that, Interference assessment analysis was performed on the serum protein characteristic parameters to obtain interfering information. The specific analysis process is as follows: The characteristic parameters of serum proteins were identified to obtain dual-frequency time-series microwave S-parameter data, microwave dielectric response data, dielectric property synchronization data, and multi-frequency microwave S-parameter data. The dual-frequency attenuation difference coefficient was obtained by performing characteristic quantitative analysis on dual-frequency time-series microwave S-parameter data; the dielectric interference ratio coefficient was obtained by performing characteristic quantitative analysis on microwave dielectric response data; the coupling interference coefficient was obtained by performing characteristic quantitative analysis on dielectric characteristic synchronization data; and the particle size distribution interference coefficient was obtained by performing characteristic quantitative analysis on multi-frequency microwave S-parameter data. The dual-frequency attenuation difference coefficient, dielectric interference ratio coefficient, coupling interference coefficient, and particle size distribution interference coefficient are normalized and mapped to the interval [0, 1]. The interference state value is calculated and output using a preset comprehensive coupling formula. The interference state value is divided into a low interference state interval and an interference separation state interval according to a preset state value segmentation value. If the interference state value of the current serum sample is in the low interference state interval, the generated interference information is no interference information. If the interference state value is in the interference separation state interval, the generated interference information is an interference separation initiation signal.

4. A microwave biosensor real-time monitoring platform for low concentrations of CEA in serum according to claim 3, characterized in that, The dual-frequency attenuation difference coefficient is obtained by performing characteristic quantization analysis on the dual-frequency time-series microwave S-parameter data, specifically as follows: The average values ​​of the transmitted microwave S-parameter signals acquired multiple times at low and high frequencies are obtained based on dual-frequency time-series microwave S-parameter data. The average values ​​corresponding to low and high frequencies are denoted as low-frequency microwave time-domain signals and high-frequency microwave time-domain signals, respectively. The fractional decay energy corresponding to the dual frequencies of the serum sample is obtained by processing the low-frequency microwave time-domain signals and high-frequency microwave time-domain signals using the set Caputo fractional derivative. The low-frequency fractional attenuation energy and the high-frequency fractional attenuation energy are obtained based on the energy calculation formula; then the two fractional attenuation energies are input into the set nonlinear difference coefficient calculation formula to calculate the dual-frequency attenuation difference coefficient.

5. A microwave biosensor real-time monitoring platform for low concentrations of CEA in serum according to claim 4, characterized in that, The dielectric interference ratio is obtained by performing characteristic quantification analysis on the microwave dielectric response data. Specifically, the dielectric response time-domain sequence is obtained based on the microwave dielectric response data, and it is substituted into the Gaussian radial basis kernel function to obtain the regenerating kernel function Hilbert space. The dielectric interference ratio is obtained by analyzing the dielectric response time-domain sequence and the regenerating kernel function Hilbert space based on the orthogonal projection of the set RKHS space algorithm.

6. A microwave biosensor real-time monitoring platform for low concentrations of CEA in serum according to claim 5, characterized in that, The coupling interference coefficient is obtained by performing characteristic quantization analysis on the dielectric property synchronization data, specifically as follows: Based on the synchronous data of dielectric properties, the microwave attenuation coefficient sequence and the resonant frequency offset sequence are obtained, and the empirical distribution function transformation is performed to obtain the uniform distribution sequence. That is, the microwave attenuation uniform distribution sequence and the resonant uniform distribution sequence are obtained by analyzing the empirical distribution function. The microwave attenuation uniform distribution sequence and the resonance uniform distribution sequence are input into the Gumbel Copula function; which includes the joint distribution characteristics of the microwave attenuation uniform distribution value and the resonance uniform distribution value corresponding to the two uniform distribution sequences for each sample's microwave S-parameter sampling points, as well as the Copula dependency parameter; The optimal dependency parameters are obtained by solving the maximum likelihood estimation of the Copula dependency parameters; the optimal dependency parameters are then input into the coupling interference quantization formula to calculate the coupling interference coefficient.

7. A microwave biosensor real-time monitoring platform for low concentrations of CEA in serum according to claim 6, characterized in that, The particle size distribution interference coefficient is obtained by performing characteristic quantification analysis on the multi-frequency microwave S-parameter data, specifically as follows: Based on multi-frequency microwave S-parameter data, the mean particle size and particle size mass ratio are obtained, and the corresponding effective separation particle size intervals are acquired. The particle size values ​​of each interfering protein are weighted and analyzed with respect to the effective separation particle size intervals to obtain the weight adjustment factor of each particle size interval. The weighted mean of particle size is obtained by calculating the weighted mean of particle size based on the weight adjustment factor; The particle size distribution interference coefficient is calculated and output using the non-uniformity interference calculation formula.

8. A microwave biosensor real-time monitoring platform for low concentrations of CEA in serum according to claim 6, characterized in that, The optimal copula dependency parameters are obtained by performing maximum likelihood estimation. Specifically, the preset number of distribution sequence groups corresponds to the measured microwave attenuation uniform distribution sequence and the resonance uniform distribution sequence. The Gumbel Copula function is substituted sequentially, and all results are multiplied to obtain the likelihood function for the copula dependency parameters. The optimal solution of the copula dependency parameters obtained by maximum likelihood estimation is referred to as the optimal dependency parameters below.

9. A microwave biosensor real-time monitoring platform for low concentrations of CEA in serum according to claim 1, characterized in that, The interference separation unit receives interference information and performs interference separation and shielding to obtain an optimized serum sample. Specifically, if the interference information of the current serum sample is identified as having no interference information, it is directly marked as an optimized serum sample; when the interference information is identified as an interference separation start signal, interference separation and shielding is performed.

10. A microwave biosensor real-time monitoring platform for low concentrations of CEA in serum according to claim 2, characterized in that, The dual-frequency time-series microwave S-parameter data are the time-domain and frequency-domain signals of the S-parameters of serum samples at low and high frequencies of a set frequency. The microwave dielectric response data is obtained by transmitting a wideband swept microwave excitation covering the resonant frequency band of the microwave sensing unit, collecting microwave network scattering parameters, and obtaining the dielectric response time-domain sequence through inverse Fourier transform. The dielectric property synchronization data consists of the microwave S-parameter attenuation coefficient sequence and the resonant frequency offset sequence of serum samples within a set sampling period at low frequency. The multi-frequency microwave S-parameter data consists of the S-parameter attenuation coefficients at low, medium, and high frequencies of the set frequencies. Based on the Rayleigh scattering model, the microwave attenuation at each frequency is inverted to obtain the particle size range of various interfering proteins. The average particle size is calculated based on the particle size range, as well as the ratio of the mass of interfering proteins in each particle size range to the total mass of interfering proteins, i.e., the particle size mass ratio. The average particle size and the particle size mass ratio are combined to obtain the multi-frequency microwave S-parameter data. An effective separation particle size range is set.

Images