Oxidized low-density lipoprotein detection method and equipment based on magnetic microsphere luminescence

Abstract

The invention discloses an oxidized low-density lipoprotein detection method and oxidized low-density lipoprotein detection equipment based on magnetic microsphere luminescence, and the method comprises the following steps: carrying out alkaline phosphatase labeling on an oxidized low-density lipoprotein antibody by using the oxidized low-density lipoprotein antibody, carrying out mixed reaction on the other antibody and modified carboxyl / toluenesulfonyl microspheres, and carrying out separation and purification to obtain a target product, a chemiluminescence immunoassay system is constructed, and oxidized low-density lipoprotein antigens in a sample, enzyme-labeled antibodies and labeled magnetic beads form a'magnetic bead-antibody-antigen-enzyme-labeled antibody 'compound. Then, through a two-step cleaning process, an optical signal value RLU is generated under the action of a light-emitting substrate APS-5, and finally, the optical signal value is amplified by a photomultiplier tube PMT in light-emitting equipment. According to the scheme, the specificity can be improved from the sample capturing end, and the sensitivity can be enhanced at the signal processing end.

Description

Technical Field

[0001] The present application relates to the field of bioinformatics, and in particular to a method and device for detecting oxidized low-density lipoprotein based on the luminescence of magnetic microspheres. Background Art

[0002] Oxidized low-density lipoprotein (ox-LDL) is one of the important markers of atherosclerosis, and its level is closely related to the occurrence and development of cardiovascular diseases. Accurate and sensitive detection of ox-LDL is of great significance for the early diagnosis and prevention of cardiovascular diseases. At present, the detection methods of ox-LDL mainly include enzyme-linked immunosorbent assay (ELISA), chemiluminescent immunoassay (CLIA), etc. However, these methods generally have problems such as insufficient sensitivity, poor specificity, and complex operation, which are difficult to meet clinical needs.

[0003] In recent years, immunoassays based on magnetic microspheres have been widely used in the field of biomarker detection due to their advantages such as large specific surface area, rapid separation, and low background interference. However, traditional magnetic microsphere surface modification methods often use simple protein or antibody direct adsorption, which can easily lead to problems such as changes in antibody conformation, decreased activity, and increased nonspecific adsorption. In addition, conventional chemiluminescence signal processing methods usually rely on simple standard curve methods, which are difficult to effectively handle background interference and nonlinear signal responses in complex samples, limiting the sensitivity and accuracy of detection.

[0004] Therefore, there is an urgent need for a technical solution that can improve specificity at the sample capture end and enhance sensitivity at the signal processing end. Summary of the invention

[0005] In order to solve the deficiencies of the prior art, the present invention provides a method and device for detecting oxidized low-density lipoprotein based on the luminescence of magnetic microspheres. The present invention solves the technical problems of high false positive rate and insufficient detection limit in the prior art marker detection.

[0006] The embodiment of the present application provides a method for detecting oxidized low-density lipoprotein based on the luminescence of magnetic microspheres, comprising: mixing superparamagnetic iron oxide microspheres with amphiphilic block copolymers, performing surface modification by a solvent volatilization method, and characterizing product characteristics by light scattering and potentiometric determination; using an oxidized low-density lipoprotein antibody to label the antibody with alkaline phosphatase, mixing another antibody with the modified carboxyl / toluenesulfonyl microspheres for reaction, and separating and purifying to obtain a target product; constructing a chemiluminescent immunoassay system, collecting kinetic curves of the chemiluminescence intensity of the magnetic microspheres modified with the antibody and oxidized low-density lipoprotein, and using pulse excitation to improve the collection sensitivity; and constructing a convolutional neural network to predict the concentration of oxidized low-density lipoprotein through the luminescence kinetic curve.

[0007] In a possible implementation, superparamagnetic iron oxide microspheres are mixed with amphiphilic block copolymers, surface modification is performed by solvent volatilization method, and light scattering and potentiometric method are used to characterize product properties, including: dispersing superparamagnetic iron oxide microspheres in deionized water, forming a suspension by ultrasonic treatment; preparing an amphiphilic block copolymer by ring-opening polymerization, and dissolving the copolymer in an organic solvent; dropping a copolymer solution containing carboxyl or toluenesulfonyl groups into the suspension, mixing under high-speed stirring conditions, and removing the organic solvent by rotary evaporation; using dynamic light scattering method and potentiometric method to characterize the particle size and surface charge of the obtained product, and calculating the modification density to verify the surface modification effect.

[0008] In a possible implementation, an oxidized low-density lipoprotein antibody is used to perform alkaline phosphatase labeling of the antibody, and another antibody is mixed with the modified carboxyl / toluenesulfonyl microspheres for reaction, and the target product is separated and purified, including: dissolving the antibody in a buffer, adding alkaline phosphatase, reacting at room temperature for a preset time, and obtaining an alkaline phosphatase-modified antibody; suspending the surface-modified microspheres in a buffer, adding an excess of DBCO-PEG4-NHS ester, reacting at room temperature for a preset time, and introducing carboxyl / toluenesulfonyl groups; mixing the antibody with the carboxyl / toluenesulfonyl-modified microspheres, stirring the reaction at room temperature for a specific time, covalently linking the antibody through the reaction, and removing unreacted antibodies through magnetic separation and washing.

[0009] In a possible implementation, a chemiluminescent immunoassay system is constructed, a kinetic curve of the chemiluminescent intensity of the magnetic microspheres modified with antibodies and oxidized low-density lipoprotein is collected, and pulse excitation is used to improve the collection sensitivity, including: selecting an enhanced chemiluminescent substrate, incubating the magnetic microspheres modified with antibodies with an oxidized low-density lipoprotein standard, adding an enzyme-labeled secondary antibody, and constructing a chemiluminescent immunoassay system; using an automated microplate reader to collect chemiluminescent signals at a preset wavelength, and obtaining a luminescent kinetic curve including characteristics such as maximum luminescent intensity, rising rate, and peak time; using a modified Gompertz function to describe the change in luminescent intensity over time, obtaining kinetic parameters related to the analyte concentration through nonlinear fitting, and establishing a quantitative analysis model; applying periodic strong light pulse stimulation during the continuous collection process, introducing additional oscillating luminescent components, and improving the detection sensitivity and signal-to-noise ratio by analyzing the composite luminescent signal.

[0010] In a possible implementation, a convolutional neural network is constructed to predict the concentration of oxidized low-density lipoprotein through a luminescence kinetic curve, including: constructing a training data set containing multiple concentration gradient samples, and repeatedly measuring each concentration multiple times to obtain luminescence kinetic curve data; designing a multi-layer convolutional neural network model, with time series curve data as input, including a convolution layer, a pooling layer, a fully connected layer, and a dropout layer; using an optimization algorithm to train the neural network model, and evaluating the model performance through cross-validation; randomly transforming the original curve data to expand the data set to improve the model's adaptability to data under different conditions; and evaluating the prediction accuracy of the model based on the logarithmic difference between the predicted concentration and the actual concentration.

[0011] In one possible implementation, wherein an automated microplate reader is used to collect chemiluminescent signals at a preset wavelength to obtain a luminescence kinetic curve including characteristics such as maximum luminescence intensity, rise rate, and peak time, including: ,in, represents the chemiluminescence intensity at time t, Indicates the maximum luminous intensity, is the rising rate constant, Indicates time, Indicates the turning point time, represents the background signal intensity, represents the decay rate constant.

[0012] In a possible implementation, in which periodic strong light pulse stimulation is applied during the continuous acquisition process, an additional oscillating luminescence component is introduced, and the detection sensitivity and signal-to-noise ratio are improved by analyzing the composite luminescence signal, including: ,in, represents the luminous intensity at time t after pulse excitation, represents the original chemiluminescence intensity, represents the additional luminescence intensity caused by pulse excitation, Indicates time, represents the decay time constant, Indicates the oscillation frequency.

[0013] An embodiment of the present application also provides an oxidized low-density lipoprotein detection device, comprising: a processor, a memory, and a system bus; wherein the processor and the memory are connected via the system bus; the memory is used to store one or more programs, and the one or more programs include instructions, which, when executed by the processor, enable the processor to execute the method described in any one of the above embodiments.

[0014] In the method and device for detecting oxidized low-density lipoprotein based on magnetic microsphere luminescence as provided above, the embodiment of the present application improves the traditional magnetic microsphere surface modification method and introduces a machine learning algorithm to optimize signal processing, which can improve specificity from the sample capture end and enhance sensitivity at the signal processing end. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0016] Figure 1 A schematic diagram of a method for detecting oxidized low-density lipoprotein based on magnetic microsphere luminescence provided in an embodiment of the present application; Figure 2 A schematic diagram of a chemiluminescence intensity curve provided in an embodiment of the present application; Figure 3 A schematic block diagram of a 7-layer convolutional neural network provided in an embodiment of the present application. DETAILED DESCRIPTION

[0017] Various exemplary embodiments of the present application will now be described in detail with reference to the accompanying drawings. It should be noted that unless otherwise specifically stated, the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present application.

[0018] Those skilled in the art will appreciate that the terms "first", "second" and the like in the embodiments of the present application are only used to distinguish different steps, devices or modules, etc., and do not represent any specific technical meaning, nor do they represent the necessary logical order between them. It should also be understood that in the embodiments of the present application, "multiple" may refer to two or more, and "at least one" may refer to one, two or more. It should also be understood that for any component, data or structure mentioned in the embodiments of the present application, in the absence of explicit limitation or contrary revelation given in the context, it can generally be understood as one or more. In addition, the term "and / or" in the present application is only a kind of association relationship describing the associated objects, indicating that there can be three kinds of relationships, for example, A and / or B, which can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in the present application generally indicates that the associated objects before and after are an "or" relationship. It should also be understood that the description of each embodiment in the present application emphasizes the differences between the embodiments, and the same or similar parts can refer to each other. For the sake of brevity, they will not be repeated one by one.

[0019] At the same time, it should be understood that, for ease of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship. The following description of at least one exemplary embodiment is actually only illustrative and is by no means intended to limit the present application and its application or use. The techniques, methods and devices known to ordinary technicians in the relevant fields may not be discussed in detail, but where appropriate, the techniques, methods and devices should be considered as part of the specification. It should be noted that similar numbers and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further discussed in subsequent drawings.

[0020] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0021] Figure 1 A schematic flow chart of a method for detecting oxidized low-density lipoprotein based on the luminescence of magnetic microspheres provided in the embodiments of the present application. It should be understood that the surface modification of submicron magnetic microspheres is an important biomedical technology for improving the performance and function of microspheres. This modification generally involves attaching specific molecules or polymers to the surface of the microspheres to give them specific properties or functions. In the embodiments of the present application, an amphiphilic copolymer composed of hydrophilic polyethylene glycol (PEG) and hydrophobic polylactic acid (PLA) is used. This copolymer can simultaneously meet the hydrophilic and hydrophobic requirements of the microsphere surface.

[0022] The PEG segments are exposed on the outer surface of the microspheres during the modification process, and this design has multiple advantages. First, PEG is biocompatible and can significantly reduce nonspecific adsorption, which is crucial for improving the specificity of detection and reducing background interference. Second, the PEG segments can form a hydration layer, increasing the dispersibility and stability of the microspheres in aqueous solution. At the same time, the PLA segments mainly interact hydrophobically with the surface of the magnetic microspheres. This interaction is achieved through van der Waals forces and hydrophobic effects, which can form a stable bond. The hydrophobic nature of PLA enables it to adhere tightly to the surface of the microspheres to form a stable coating. This design not only ensures the stability of the modified layer, but also allows the thickness and properties of the modified layer to be controlled by adjusting the molecular weight and structure of PLA.

[0023] On the basis of copolymer modification, the present application further introduces a click chemistry reaction to covalently link the anti-ox-LDL antibody to the end of the PEG segment. Click chemistry is a type of efficient and specific chemical reaction with the advantages of mild reaction conditions, high yield, and few by-products. In the present application, the copper-catalyzed azide-carboxyl or tosyl cycloaddition reaction (CuAAC) can be used, which is one of the most commonly used click chemistry reactions. This reaction can be carried out under physiological conditions without destroying the activity of the antibody.

[0024] This layered modification method has multiple advantages. First, it improves the efficiency of antibody attachment. Compared with directly attaching antibodies to the microsphere surface, using PEG segments as intermediates can reduce steric hindrance, making it easier for antibodies to approach the reaction site. Second, this method enhances the stability of antibody attachment. The formation of covalent bonds ensures that the antibody will not easily fall off the microsphere surface, which is crucial to ensure the reliability and repeatability of detection.

[0025] More importantly, this method allows the optimization of the spatial arrangement of antibodies by adjusting the PEG chain length. The adjustment of PEG chain length can affect the distribution density and accessibility of antibodies on the surface of microspheres. Longer PEG chains can keep antibodies away from the surface of microspheres, reduce steric hindrance, and increase the freedom to bind to target molecules. Shorter PEG chains can increase the antibody density per unit area. By precisely controlling the PEG chain length, the optimal balance can be found between antibody density and binding freedom, thereby maximizing the binding ability to ox-LDL.

[0026] In terms of signal processing, the present application embodiment introduces a machine learning algorithm, in particular a convolutional neural network (CNN) model based on deep learning, to process chemiluminescent signal data. This is a major innovation in traditional chemiluminescent signal processing methods. Traditional methods usually rely on a simple standard curve method, which is intuitive and easy to understand, but is easily affected by background noise and nonspecific binding, especially in the detection of low-concentration samples, which may cause significant errors.

[0027] The introduction of the CNN model provides a new approach to solving these problems. CNN is a deep learning model specifically designed to process data with a grid-like topological structure. In order to train the CNN model, it is first necessary to establish a large-scale training dataset containing ox-LDL samples of various concentrations. This dataset should contain complete chemiluminescence kinetic curves, not just a single peak intensity. These curves contain rich information, such as the rate of rise, peak time, decay rate, etc., which may be related to the concentration of ox-LDL. By using this comprehensive data to train the model, CNN can learn more features and patterns, thereby improving the accuracy of predictions.

[0028] The advantage of the CNN model is that it can automatically extract features from complex luminescence curves. Compared with traditional methods, CNN not only considers peak intensity, but also analyzes multi-dimensional features such as curve shape and rise rate. This multi-dimensional analysis capability allows the model to capture subtle differences that traditional methods may ignore, thereby greatly improving the accuracy and sensitivity of detection. This ability is particularly important in the detection of low-concentration samples, because in this case, the signal is often weak and easily masked by background noise.

[0029] In addition, the CNN model has the ability to continuously learn and optimize. By continuously inputting new data, the model can continuously update and improve its predictive ability. This feature enables the model to adapt to slight differences in different batches of reagents and instruments. In practical applications, even if the same reagents and instruments are used, there may be slight changes in the results due to factors such as environmental conditions and operator differences. Traditional methods may require frequent calibration to cope with these changes, while CNN models can automatically adjust by learning these changes, thereby improving the robustness of the method.

[0030] In the specific implementation process, the surface of the magnetic microspheres needs to be modified first. Figure 1 As shown, at step S101, superparamagnetic iron oxide microspheres are mixed with amphiphilic block copolymers, surface modified by solvent evaporation, and light scattering and potentiometry are used to characterize product properties. Solvent evaporation is a surface modification method that dissolves microspheres and copolymers in a volatile solvent, then slowly evaporates the solvent so that the copolymers are uniformly deposited on the surface of the microspheres. This method can achieve precise control of the surface of the microspheres and ensure the uniformity and stability of the modified layer. After the modification is completed, the product needs to be characterized to confirm the effect of the modification. Light scattering and potentiometry are used here. Light scattering can provide information about the size and dispersibility of the microspheres, while potentiometry can reflect the charge state of the microsphere surface. The combination of these two methods can comprehensively evaluate the effect of the modification, including changes in the size of the microspheres, changes in the surface charge, and improvements in dispersibility.

[0031] Specifically, the method includes: dispersing superparamagnetic iron oxide microspheres in deionized water and forming a suspension by ultrasonic treatment; preparing an amphiphilic block copolymer by ring-opening polymerization, and dissolving the copolymer in an organic solvent; dropping the copolymer solution into the suspension, mixing under high-speed shear conditions, and removing the organic solvent by rotary evaporation; using dynamic light scattering and potentiometric measurement to characterize the particle size and surface charge of the obtained product, and calculating the modification density to verify the surface modification effect.

[0032] In one embodiment, first, superparamagnetic Fe3O4 microspheres with a particle size of 500 nm are selected as the base material. 1 g of the microspheres are dispersed in 100 mL of deionized water and ultrasonically treated for 30 minutes to fully disperse.

[0033] Next, the amphiphilic block copolymer PEG-b-PLA was prepared. Polyethylene glycol monomethyl ether with a molecular weight of 5000 (mPEG5000, Sigma-Aldrich, 202371) and poly (L-lactic acid) with a molecular weight of 3000 (PLLA3000, Sigma-Aldrich, 765112) were selected as raw materials. The mPEG5000-b-PLLA3000 copolymer was obtained by ring-opening polymerization with stannate (Sn(Oct)2) as catalyst and reacted at 80°C in anhydrous toluene for 24 hours.

[0034] 0.5 g of the prepared copolymer was dissolved in 50 mL of dichloromethane and slowly added dropwise to the microsphere suspension, while stirring at 10000 rpm for 30 minutes using a high-speed disperser (e.g., IKA T25 digital ULTRA-TURRAX). The mixture was then transferred to a rotary evaporator and dichloromethane was slowly evaporated at 40° C. to obtain surface-modified magnetic microspheres.

[0035] The surface modification effect can be calculated by the following formula: modification density: σ = (m_polymer * N_A) / (M_polymer * S_microsphere), where σ is the modification density (molecules / nm²), m_polymer is the mass of the copolymer, N_A is the Avogadro constant, M_polymer is the molecular weight of the copolymer, and S_microsphere is the total surface area of ​​the microsphere.

[0036] Finally, the modification effect can be verified by measuring the change in particle size before and after modification and the change in zeta potential through dynamic light scattering (DLS). The typical modification density is about 0.5-1 molecules / nm², the particle size increases by about 10-20nm after modification, and the zeta potential changes from the original -5mV to about -15mV.

[0037] At step S102, an oxidized low-density lipoprotein antibody is used to perform alkaline phosphatase labeling of the antibody, and another antibody is mixed with the modified carboxyl / tosyl microspheres for reaction, and the target product is separated and purified. The azidation modification of the antibody usually introduces an azide group at a specific site of the antibody, while the carboxyl or tosyl modification of the microspheres introduces a carboxyl or tosyl group at the end of the PEG segment. Both modifications require precise control to ensure that the activity of the antibody and the performance of the microspheres are not affected. After the modification is completed, the modified antibody is mixed with the microspheres for reaction, and the covalent connection of the antibody is achieved through a click chemistry reaction. This reaction is usually carried out under mild conditions to protect the activity of the antibody. After the reaction is completed, separation and purification are required to remove unreacted antibodies and other impurities to obtain the final target product.

[0038] Specifically, the method includes: dissolving the antibody in a buffer solution, adding an excess of azide-PEG4-NHS ester, reacting at room temperature for a preset time to obtain an azidated antibody; suspending surface-modified microspheres in a buffer solution, adding an excess of DBCO-PEG4-NHS ester, reacting at room temperature for a preset time to introduce alkynyl groups; mixing the azidated antibody with carboxyl- or tosyl-modified microspheres, stirring the reaction at room temperature for a specific time, covalently linking the antibody through the reaction, and removing unreacted antibodies through magnetic separation and washing.

[0039] In one embodiment, a commercial Ox-LDL monoclonal antibody is selected for connection. First, the antibody is diluted to 1 mg / mL with PBS buffer (pH 7.4). Then, a 20-fold molar excess of azide-PEG4-NHS ester is added and reacted at room temperature for 2 hours to obtain an azidated antibody. At the same time, the surface-modified microspheres obtained in the previous step are suspended in PBS buffer, and a 20-fold molar excess of diphenylcyclooctyne is added and reacted at room temperature for 2 hours to introduce carboxyl or toluenesulfonyl groups. The azidated antibody and carboxyl- or toluenesulfonyl-modified microspheres are mixed, and the reaction is gently stirred at room temperature for 4 hours to achieve covalent connection of the antibody by click chemistry reaction. The unreacted antibody is removed by magnetic separation and PBS washing 3 times.

[0040] The antibody attachment efficiency can be determined by BCA protein quantification, and the typical attachment density is about 2000-3000 antibodies / μm². The attachment efficiency η can be calculated by the following formula: η = (m_initial - m_unbound) / m_initial *100%, where m_initial is the initial mass of the added antibody and m_unbound is the mass of the unattached antibody.

[0041] To optimize the spatial arrangement of antibodies, the PEG chain length can be adjusted to control it. For example, two copolymers, mPEG2000-b-PLLA3000 and mPEG10000-b-PLLA3000, can be prepared and mixed in different proportions. The antibody arrangement can be optimized by adjusting the mixing ratio. The optimal ratio can be determined by orthogonal experimental design. Usually, the mass ratio of 2000:10000 is better between 1:2 and 1:4.

[0042] In step S103, a chemiluminescent immunoassay system is constructed to collect the kinetic curve of the chemiluminescent intensity of the magnetic microspheres modified with antibodies and oxidized low-density lipoprotein, and pulse excitation is used to improve the collection sensitivity (such as Figure 2 This step involves mixing the antibody-modified magnetic microspheres with the ox-LDL sample and then measuring the kinetic curve of the chemiluminescence intensity. In order to improve the acquisition sensitivity, pulse excitation technology is used. Pulse excitation can reduce background noise and improve the signal-to-noise ratio, thereby increasing the sensitivity of the detection.

[0043] These include: selecting an enhanced chemiluminescent substrate, incubating the antibody-modified magnetic microspheres with an oxidized low-density lipoprotein standard, adding an enzyme-labeled secondary antibody, and constructing a chemiluminescent immunoassay system; using an automated microplate reader to collect chemiluminescent signals at a preset wavelength, and obtaining a luminescence kinetic curve containing characteristics such as maximum luminescence intensity, rise rate, and peak time; using a modified Gompertz function to describe the change in luminescence intensity over time, obtaining kinetic parameters related to the analyte concentration through nonlinear fitting, and establishing a quantitative analysis model; applying periodic strong light pulse stimulation during the continuous acquisition process, introducing additional oscillating luminescence components, and improving detection sensitivity and signal-to-noise ratio by analyzing the composite luminescence signal.

[0044] In one implementation scenario, when establishing a chemiluminescent detection system, an enhanced chemiluminescent substrate can be selected as a luminescent reagent. The magnetic microspheres modified with antibodies are incubated with different concentrations of Ox-LDL standards, and then a secondary antibody labeled with horseradish peroxidase (HRP) is added.

[0045] An automated microplate reader was used to collect chemiluminescent signals. The excitation wavelength was set to 450nm, the emission wavelength was set to 480nm, the acquisition time was set to 300 seconds, and the interval was set to 0.1 seconds. The obtained luminescence kinetic curve contains rich information, including not only the maximum luminescence intensity, but also the rise rate, peak time and other characteristics.

[0046] The change of chemiluminescence intensity I(t) over time can be described by the following modified Gompertz function: Wherein, an automated microplate reader is used to collect chemiluminescence signals at a preset wavelength to obtain a luminescence kinetic curve including characteristics such as maximum luminescence intensity, rise rate, and peak time, including: ,in, represents the chemiluminescence intensity at time t, Indicates the maximum luminous intensity, is the rising rate constant, Indicates time, Indicates the turning point time, represents the background signal intensity, represents the decay rate constant. These parameters can be obtained by nonlinear fitting and are all related to the ox-LDL concentration.

[0047] In order to improve the detection sensitivity, pulse excitation technology can be used. That is, during the continuous acquisition process, a 50ms strong light pulse (1000 lux) is given every 10 seconds to significantly improve the signal-to-noise ratio. Among them, applying periodic strong light pulse stimulation during the continuous acquisition process introduces additional oscillating luminescence components, and improving the detection sensitivity and signal-to-noise ratio by analyzing the composite luminescence signal, including: ,in, represents the luminous intensity at time t after pulse excitation, represents the original chemiluminescence intensity, represents the additional luminescence intensity caused by pulse excitation, Indicates time, represents the decay time constant, This method can increase the detection limit from the original 1ng / mL to 0.1ng / mL.

[0048] In step S104, a convolutional neural network is constructed to predict the concentration of oxidized low-density lipoprotein through the luminescence kinetic curve (e.g. Figure 3 The design of the network structure needs to consider the characteristics of the input data (such as the length of the curve and the sampling frequency) as well as the expected output (ox-LDL concentration). The training of the model requires a large amount of labeled data, including the complete luminescence kinetic curves of ox-LDL samples with different concentrations. During the training process, the network parameters need to be continuously adjusted to minimize the prediction error. Finally, an independent validation set is needed to evaluate the performance of the model to ensure that it has good generalization ability.

[0049] These include: constructing a training data set containing multiple concentration gradient samples, repeatedly measuring each concentration multiple times to obtain luminescence kinetic curve data; designing a multi-layer convolutional neural network model with time series curve data as input, including convolutional layer, pooling layer, fully connected layer and dropout layer; using an optimization algorithm to train the neural network model and evaluating the model performance through cross-validation; randomly transforming the original curve data to expand the data set to improve the model's adaptability to data under different conditions; and evaluating the model's prediction accuracy based on the logarithmic difference between the predicted concentration and the actual concentration.

[0050] In summary, the embodiment of the present application combines the surface modification technology and the deep learning algorithm method to provide a new solution for the high-sensitivity detection of ox-LDL. It not only improves the specificity from the sample capture end, but also enhances the sensitivity from the signal processing end, effectively solving the technical problems of high false positive rate and insufficient detection limit in ox-LDL detection.

[0051] Figure 2 A schematic diagram of a chemiluminescence intensity curve provided in an embodiment of the present application. As can be seen from the figure, the luminescence intensity in the unexcited state is relatively stable, while the luminescence intensity increases significantly after pulse excitation and a more complex fluctuation pattern appears. This reflects the significant impact of pulse excitation on the system dynamics. The chemiluminescence signal under unexcited conditions presents a typical Gompertz function characteristic, that is, there is a rapid rise process in the initial stage, and then gradually stabilizes. This signal pattern is consistent with the kinetic characteristics of most chemiluminescent reactions and reflects the gradual consumption process of the luminescent substrate. However, it is often difficult to detect extremely low concentration analytes relying solely on this spontaneous luminescence because the signal intensity is relatively weak and easily interfered by background noise.

[0052] In contrast, the signal after pulse excitation technology showed significantly different characteristics. First, the basic form of the Gompertz function is still maintained in the overall trend, which shows that pulse excitation does not change the essential mechanism of the chemiluminescent reaction. But on this basis, additional luminescence components caused by pulse excitation are superimposed. These additional luminescence appear as a series of rapidly decaying spikes, and their envelopes show an exponential decay trend. Each spike corresponds to a pulse excitation event, and the peak intensity is much higher than the signal intensity in the unexcited state. This signal enhancement effect originates from the instantaneous energy input caused by light excitation, which causes more luminescent molecules to be excited to a high-energy state at the same time, thereby generating stronger photon release.

[0053] It is worth noting that the signal after pulse excitation also shows obvious oscillation characteristics. This oscillation may originate from the nonlinear response of the system and reflects the energy exchange process between excited state molecules and ground state molecules. The frequency and decay rate of the oscillation contain rich kinetic information, which can be used to further improve the specificity and sensitivity of the detection.

[0054] From a quantitative perspective, the pulse excitation technology significantly improves the peak intensity of the signal. In the time range shown in the figure, the maximum signal intensity in the unexcited state is about 100 arbitrary units, while the peak signal after pulse excitation can reach about 200 units, which is about doubled.

[0055] Figure 3A schematic block diagram of a 7-layer convolutional neural network provided in an embodiment of the present application. It should be understood that the system shown in the figure is exemplary and non-restrictive. This means that the system architecture involved is not limited to a specific form or design, but is presented as an example. In other words, the architecture shown in the figure can be regarded as a way of expression to clearly describe related concepts and relationships, and does not exclude other forms of architecture. Therefore, when interpreting the architecture in the picture, it should be understood that the model is flexible and diverse, and its purpose is to provide an exemplary description rather than a restrictive provision on a specific form.

[0056] In one embodiment, a training data set containing 1000 samples is constructed, and the Ox-LDL concentration ranges from 0.1 ng / mL to 1000 ng / mL, which is logarithmically uniformly distributed. Each concentration is measured 10 times, and 10,000 luminescence kinetic curves are obtained.

[0057] A 7-layer convolutional neural network (CNN) model is designed, with the input being 300×1 time series data (i.e., 300-second luminescence curve). The network structure is as follows: - 1D convolutional layer (32 filters, kernel size 3) - Max pooling layer (pooling size is 2) - 1D convolutional layer (64 filters, kernel size 3) - Max pooling layer (pooling size is 2) - Fully connected layer (128 neurons) - Dropout layer (dropout rate 0.5) - Fully connected layer (1 neuron, outputs the predicted ox-LDL concentration) The Adam optimizer was used, the learning rate was set to 0.001, the batch size was 32, and the training was performed for 500 rounds. The model performance was evaluated by 5-fold cross validation, and the typical root mean square error (RMSE) could reach 0.05log(ng / mL).

[0058] To further improve the generalization ability of the model, data augmentation technology was introduced. The original curve was subjected to random time shift (±5 seconds), amplitude scaling (0.9-1.1 times) and Gaussian noise addition (σ=0.01), and the data set was expanded 10 times. This method can make the model better adapt to the slight differences between different batches of reagents and instruments.

[0059] The prediction accuracy P of the model can be evaluated by the following formula: P = 1 - Σ|log(C_pred) - log(C_true)| / (N * log(C_max / C_min)), where C_pred is the predicted concentration, C_true is the actual concentration, N is the number of samples, and C_max and C_min are the upper and lower limits of the concentration range. Typical P values ​​can reach above 0.95.

[0060] Furthermore, an embodiment of the present application also provides an ox-LDL detection device, comprising: a processor, a memory, and a system bus; the processor and the memory are connected via the system bus; the memory is used to store one or more programs, the one or more programs comprising instructions, and when the instructions are executed by the processor, the processor executes any of the above methods.

[0061] Furthermore, an embodiment of the present application also provides a computer program product, which, when executed on a terminal device, enables the terminal device to execute any of the above methods.

[0062] It can be known from the description of the above implementation mode that those skilled in the art can clearly understand that all or part of the steps in the above-mentioned embodiment method can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product can be stored in a storage medium such as ROM / RAM, a disk, an optical disk, etc., including several instructions for enabling a computer device (which can be a personal computer, a server, or a network communication device such as a media gateway, etc.) to execute the methods described in the various embodiments of the present application or certain parts of the embodiments.

[0063] It should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same or similar parts between the various embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0064] It should also be noted that, in the embodiments of the present application, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements includes not only those elements, but also includes other elements that are not clearly listed, or also includes elements inherent to such process, method, article or equipment. In the absence of more restrictions, the elements limited by the statement "comprise one..." do not exclude the presence of other identical elements in the process, method, article or equipment including the elements.

[0065] The above description of the disclosed embodiments enables professionals and technicians in the field to implement or use the present application. Various modifications to these embodiments will be apparent to professionals and technicians in the field, and the general principles defined in the embodiments of the present application can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to these embodiments shown in the embodiments of the present application, but will conform to the widest range consistent with the principles and novel features disclosed in the embodiments of the present application.

Claims

1. A method for detecting oxidized low-density lipoprotein based on magnetic microsphere luminescence, characterized in that: include: Superparamagnetic iron oxide microspheres were mixed with amphiphilic block copolymers, surface modified by solvent evaporation method, and characterized by light scattering and potentiometry. An oxidized low-density lipoprotein antibody is used to label the antibody with alkaline phosphatase, and another antibody is mixed with the modified carboxyl / tosyl microspheres for reaction, and the target product is separated and purified; A chemiluminescent immunoassay system was constructed to collect the kinetic curve of the chemiluminescent intensity of the antibody-modified magnetic microspheres and oxidized low-density lipoprotein, and pulse excitation was used to improve the collection sensitivity; A convolutional neural network was constructed to predict the concentration of oxidized low-density lipoprotein based on the luminescence kinetic curve.

2. The method for detecting oxidized low-density lipoprotein according to claim 1, characterized in that: in, Superparamagnetic iron oxide microspheres were mixed with amphiphilic block copolymers, surface modified by solvent evaporation, and characterized by light scattering and potentiometry, including: Superparamagnetic iron oxide microspheres are dispersed in deionized water and formed into a suspension by ultrasonic treatment; preparing an amphiphilic block copolymer by ring-opening polymerization, and dissolving the copolymer in an organic solvent; Adding the copolymer solution containing carboxyl or toluenesulfonyl groups dropwise into the suspension, mixing under high-speed stirring conditions, and removing the organic solvent by rotary evaporation; Dynamic light scattering and potentiometric methods were used to characterize the particle size and surface charge of the obtained product, and the modification density was calculated to verify the surface modification effect.

3. The method for detecting oxidized low-density lipoprotein according to claim 1, characterized in that: in, An oxidized low-density lipoprotein antibody is used to label the antibody with alkaline phosphatase, and another antibody is mixed with the modified carboxyl / tosyl microspheres for reaction, and the target product is separated and purified, including: The antibody is dissolved in a buffer, alkaline phosphatase is added, and the reaction is carried out at room temperature for a preset time to obtain an alkaline phosphatase-modified antibody; The surface-modified microspheres were suspended in a buffer, an excess of DBCO-PEG4-NHS ester was added, and the reaction was carried out at room temperature for a preset time to introduce carboxyl / tosyl groups; The antibody is mixed with the carboxyl / tosyl modified microspheres, stirred for reaction at room temperature for a specific time, the antibody is covalently linked by reaction, and unreacted antibody is removed by magnetic separation and washing.

4. The method for detecting oxidized low-density lipoprotein according to claim 1, characterized in that: in, Construct a chemiluminescent immunoassay system to collect the kinetic curve of the chemiluminescent intensity of the antibody-modified magnetic microspheres and oxidized low-density lipoprotein, and use pulse excitation to improve the collection sensitivity, including: An enhanced chemiluminescent substrate was selected, and the antibody-modified magnetic microspheres were incubated with oxidized low-density lipoprotein standards, and an enzyme-labeled secondary antibody was added to construct a chemiluminescent immunoassay system; An automated microplate reader is used to collect chemiluminescent signals at a preset wavelength to obtain a luminescence kinetic curve including characteristics such as maximum luminescence intensity, rise rate, and peak time; The modified Gompertz function was used to describe the change of luminescence intensity over time, and the kinetic parameters related to the analyte concentration were obtained through nonlinear fitting to establish a quantitative analysis model. During the continuous acquisition process, periodic strong light pulse stimulation is applied to introduce additional oscillating luminescence components, and the detection sensitivity and signal-to-noise ratio are improved by analyzing the composite luminescence signals.

5. The method for detecting oxidized low-density lipoprotein according to claim 1, characterized in that: in, Construct a convolutional neural network to predict the concentration of oxidized low-density lipoprotein through the luminescence kinetic curve, including: Construct a training data set containing multiple concentration gradient samples, and repeat the measurement multiple times for each concentration to obtain luminescence kinetic curve data; Design a multi-layer convolutional neural network model with time series curve data as input, including convolution layer, pooling layer, fully connected layer and dropout layer; Use optimization algorithms to train neural network models and evaluate model performance through cross-validation; The original curve data is randomly transformed to expand the data set to improve the model's adaptability to data under different conditions; The prediction accuracy of the model was evaluated based on the logarithmic difference between the predicted concentration and the actual concentration.

6. The method for detecting oxidized low-density lipoprotein according to claim 4, characterized in that: in, The chemiluminescent signal is collected at a preset wavelength using an automated microplate reader to obtain a luminescence kinetic curve containing characteristics such as maximum luminescence intensity, rise rate, and peak time, including: ， in, represents the chemiluminescence intensity at time t, Indicates the maximum luminous intensity, is the rising rate constant, Indicates time, Indicates the turning point time, represents the background signal intensity, represents the decay rate constant.

7. The method for detecting oxidized low-density lipoprotein according to claim 4, characterized in that: in, During the continuous acquisition process, periodic strong light pulse stimulation is applied to introduce additional oscillating luminescence components, and the detection sensitivity and signal-to-noise ratio are improved by analyzing the composite luminescence signal, including: ， in, represents the luminous intensity at time t after pulse excitation, represents the original chemiluminescence intensity, represents the additional luminescence intensity caused by pulse excitation, Indicates time, represents the decay time constant, Indicates the oscillation frequency.

8. An oxidized low-density lipoprotein detection device, characterized in that: include: A processor, a memory, and a system bus; wherein the processor and the memory are connected via the system bus; The memory is used to store one or more programs, wherein the one or more programs include instructions, and when the instructions are executed by the processor, the processor executes the method according to any one of claims 1 to 7.

Images