Immunoassay method based on holographic imaging modal conversion and deep learning

Abstract

The application discloses an immune detection method based on holographic imaging modal conversion and deep learning, and belongs to the field of immune analysis and detection. The deep learning holographic modal conversion system is constructed based on an unpaired data set, so that the influence of diffraction interference on target quantification is minimized, thereby realizing holographic super-resolution imaging and overcoming the challenge in paired data set generation. In addition, a functionalized glass slide is designed, and a click reaction signal amplification strategy is combined, so that rapid and sensitive holographic imaging detection is realized. The functionalized glass slide is combined with a portable holographic imaging device, so that signal amplification is accurately performed, the detection efficiency of the system is greatly improved, the application scene is expanded, and antibiotic residues can be rapidly and sensitively detected. The application opens up a new research direction for high-sensitivity on-site convenient detection of trace target substances in portable devices and complex samples.

Description

TECHNICAL FIELD

[0001] The present application relates to an immune detection method based on holographic imaging modal conversion and deep learning, and belongs to the field of immune analysis and detection. BACKGROUND

[0002] Detection strategies based on immunological methods, such as enzyme-linked immunosorbent assay (ELISA), lateral flow immunoassay, and microsphere-mediated immunobiosensor, are widely used in clinical diagnosis, food safety, and environmental monitoring. Although these basic immunological methods have the advantages of high specificity and ease of use, their sensitivity and accuracy are limited when detecting trace amounts of target objects in complex real samples. Currently, microfield optical immunobiosensors based on immunomicrospheres have attracted widespread interest due to their high sensitivity, which can achieve high sensitivity and accuracy detection in a short time by capturing microsphere signals through target concentration-mediated microsphere number changes and microimaging. However, traditional microsphere counting detection analysis using ordinary optical microscopes is limited by small imaging field of view and poor portability. The imaging field of view decreases with increasing magnification, hindering the collection of various sample signals and easily leading to signal loss. In addition, sampling errors are easily magnified under a smaller imaging field of view, so it is crucial to achieve a larger imaging field of view at the same magnification for the field of immunosensing based on microsphere imaging.

[0003] Lensless holographic imaging technology has attracted widespread attention due to its low cost, comparable magnification to traditional microscopes, large imaging field of view, and no loss. Lensless holographic technology surpasses the optical limitations imposed by lenses, as it can directly place the specimen on the imaging photosensitive sensor to capture images without the use of lenses, and clear object image reconstruction can be achieved using corresponding image reconstruction algorithms.

[0004] The use of a coherent light source in lensless holographic imaging technology produces interference and diffraction patterns, appearing as concentric circular fringes, which greatly reduce the imaging quality and hinder analyte recognition. In order to restore the focusing information of the holographic image, a holographic reconstruction algorithm is needed for modal conversion (from holographic state to microscopic state), thus achieving super-resolution imaging. Current reconstruction algorithms are divided into physically and optically driven reconstruction methods and data-driven reconstruction methods based on deep learning. Physically and optically driven reconstruction methods rely on analyzing or iteratively calculating hardware parameters such as axial distance and wavelength, so they are usually suitable for static objects. However, these methods require a large number of measurements and adjustments, which are time-consuming and complex to operate.

[0005] In contrast, conventional deep learning data-driven algorithms rely on training on aligned and labeled image datasets, and heavily depend on the accuracy and stability of the aligned dataset during the learning process. However, in the field of real-time diagnosis and biomedical imaging, the background features of most imaging environments are complex, and the calibration process is also very complex, making it difficult to construct a large paired dataset with the same magnification and the same position under different modal conditions. Therefore, the practical application of deep learning modal conversion based on paired datasets still has considerable challenges. The unpaired dataset algorithm does not require special calibration, is convenient for data collection, and has no fixed quantity requirement. The unpaired dataset algorithm can effectively reduce the paired bias and enhance the resistance of the model to external factors. Therefore, realizing unpaired dataset modal conversion is crucial for the practical application of holographic imaging. SUMMARY

[0006] In order to solve the above problems, the present application provides an immune detection method based on holographic imaging modal conversion and deep learning, and the technical solution is as follows:

[0007] Step 1: mix the target object to be measured, the pre-prepared CuO2@SiO2 composite and the nano magnetic particle conjugate, and perform a competitive immune reaction; the surface of the CuO2@SiO2 composite is modified with detection antibodies that recognize the target object to be measured, and the surface of the nano magnetic particle conjugate is modified with complete antigens of the target object to be measured;

[0008] Step 2: magnetic separation, dissolving the remaining CuO2@SiO2 composite after reaction in hydrochloric acid and sodium ascorbate to release copper ions;

[0009] Step 3: copper ions as catalyst, at room temperature, rapidly catalyze the azide-alkyne cycloaddition reaction, so that the alkyne-modified polystyrene microspheres and the azide ultra-thin functionalized glass slide occur azide-alkyne cycloaddition reaction, i.e. click reaction;

[0010] Step 4: holographic imaging of the polystyrene microspheres fixed on the functionalized glass slide by click reaction obtained in step 3;

[0011] Step 5: constructing an unpaired modal conversion dataset by equigradient dilution of the polystyrene microsphere holographic imaging image and the microimaging image of the equigradient dilution of the polystyrene microspheres, and training a deep learning modal conversion model; and then using the trained deep learning modal conversion model to perform modal conversion on the holographic imaging image obtained in step 4, to obtain a microscopic modal image;

[0012] Step 6: input the micro modal image obtained after the modal conversion and the micro imaging image of the equigradient dilution polystyrene microspheres into a deep learning target detection network for training, and then use the trained deep learning target detection network to identify and count the micro modal image obtained after the modal conversion of the step 4 experiment;

[0013] Step 7: obtain the concentration information of the target object to be detected by using the linear relationship between the number of polystyrene microspheres and the concentration of the target object to be detected.

[0014] Optionally, the modal conversion model adopts a CycleGAN as a model backbone network.

[0015] Optionally, the deep learning target detection network adopts YOLOV7 as a backbone network.

[0016] Optionally, the copper ion is a monovalent copper ion generated through a reduction reaction of sodium ascorbate.

[0017] Optionally, the click reaction is a covalent bond formed by the reaction of azide and alkyne under the catalysis of monovalent copper ion.

[0018] Optionally, the polystyrene microspheres are 6μm polystyrene microspheres modified with alkyne on the surface.

[0019] Optionally, the method further comprises processing the holographic image by using a non-local mean filtering algorithm.

[0020] Optionally, the nano magnetic particle conjugate is a 150nm nano magnetic particle modified with a complete antigen of the target object to be detected on the surface.

[0021] Optionally, the addition concentration of the nano magnetic particle conjugate in the reaction system is 20μg / mL.

[0022] Optionally, the addition concentration of the CuO2@SiO2 composite in the reaction system is 20μg / mL.

[0023] Optionally, the surface of the functionalized glass slide is functionalized and modified with azide on the surface.

[0024] Optionally, the copper ion is a monovalent copper ion generated through a reduction reaction of sodium ascorbate.

[0025] Optionally, the target object to be detected is a small molecule antibiotic, including at least one of chloramphenicol, neomycin, and clarithromycin.

[0026] Optionally, the maximum addition amount of hydrochloric acid in the reaction system is 8mM.

[0027] Optionally, the polystyrene microsphere-azide complex is added to the reaction system at a concentration of 50 μg / mL.

[0028] Optionally, the complete antigen-alkyne complex is added to the reaction system at a concentration of 4 mg / mL.

[0029] Optionally, the functionalized glass slide is pre-made and stored at 4℃.

[0030] Optionally, the nano-magnetic particle conjugate is incubated in the reaction system at 20-40℃ for at least 10 min. Optionally, at 37℃ for 30 min.

[0031] The present application has the following advantages:

[0032] The present application combines a holographic non-paired data set super-resolution modal transformation system and a signal amplification strategy based on click chemistry, and can quickly and sensitively detect the target object.

[0033] The present application uses a portable holographic imaging technology instead of a traditional bulky microscope as a signal readout device. The holographic imaging technology has the advantages of wide field of view, non-destructive amplification, low cost, and good portability. Through a non-paired data set modal conversion algorithm, holographic super-resolution reconstruction is achieved, the diffraction interference caused by holographic imaging is solved, the background interference is greatly overcome, and the sensitivity of detection is ensured.

[0034] The present application combines a competitive immune reaction with a signal amplification system based on a click reaction, relies on the high specificity of the competitive immune reaction and the precipitation of a large number of copper ions in the CuO2@SiO2 nanomaterial, and based on the high-efficiency catalytic properties of the click reaction, realizes efficient and accurate signal transmission.

[0035] The present application designs a functionalized glass slide and can be combined with a portable holographic imaging device, which greatly improves the detection efficiency and application scenarios of the system while accurately amplifying the signal.

[0036] The detection system in the present application has high sensitivity, strong specificity, simple operation, high detection efficiency, and wide application prospect. BRIEF DESCRIPTION OF DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0038] Figure 1Flow chart of the holographic imaging based non-paired dataset modal transformation immunodetection method of the present application. Wherein (a) is the working flow chart of the holographic imaging based non-paired dataset modal transformation immunosensor for antibiotic detection; (b) is the architecture diagram of the lens-free holographic imaging hardware and software system.

[0039] Figure 2 Preparation principle and flow chart of the ultra-thin functionalized glass slide of the present application.

[0040] Figure 3 Flow chart of the non-paired dataset modal transformation algorithm system of the present application.

[0041] Figure 4 Training results of the non-paired dataset modal transformation algorithm of the present application, wherein (a) is the cycle consistency loss curve diagram; (b) is the object detection loss curve diagram; (c) is the mAP@0.5, (d) is the bounding box loss curve diagram; (e) is the Precision-Recall curve diagram.

[0042] Figure 5 Comparison chart of the original hologram image, filtered image and modal transformation image in different concentration gradients in the embodiment of the present application.

[0043] Figure 6 Consistency comparison chart between manual counting and deep learning object detection counting algorithm in the embodiment of the present application.

[0044] Figure 7 Effect of the modal transformation model in the embodiment of the present application in dealing with the superimposed diffraction interference phenomenon, the left side shows the effect of the modal transformation of the modal transformation model and the deep learning object detection in dealing with the superimposed diffraction of two and three polystyrene microspheres, and the right side shows the transformation effect of the modal transformation model in dealing with the superimposed diffraction interference of five polystyrene microspheres.

[0045] Figure 8 Gradient signal response standard curve and linear range of the holographic imaging based non-paired dataset modal transformation immunosensor for detecting chloramphenicol (CAP) in the embodiment of the present application.

[0046] Figure 9 Recovery rate of the standard addition in the determination of different concentrations of CAP in the embodiment of the present application.

[0047] Figure 10 Experimental result chart of the specific response of the holographic imaging based non-paired dataset modal transformation immunosensor to different targets in the embodiment of the present application.

[0048] Figure 11This is a comparison chart in this embodiment of the invention, comparing the modality transformation immunoassay method based on holographic imaging and the ELISA method for detecting CAP in real samples (urine, aquaculture wastewater, and fish) using heatmap distribution. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0050] Example 1

[0051] This embodiment provides an immune detection method based on holographic imaging modality conversion and deep learning. See [link to relevant documentation]. Figure 1 This embodiment is used to detect the concentration of chloramphenicol in the sample to be tested.

[0052] First, let's introduce the sources of the main reagents used in this embodiment:

[0053] Carboxyl-functionalized polystyrene microspheres (PSM-COOH, 6 μm) were purchased from Bangs Laboratories, Inc. (USA).

[0054] Carboxylated magnetic nanoparticles (MNP) 150 -COOH) was purchased from Ocean Nano-Tech (USA).

[0055] The azide (Azide-PEG4-NH2) and alkyne (Alkyze-PEG4-NH2) were purchased from Ruixi Biotechnology Co., Ltd. (Xi'an).

[0056] 1-Ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride (EDC), sodium N-hydroxysulfonated succinimide (Sulfo-NHS), phosphate-buffered saline (PBS), and 2-(N-morpholino)ethanesulfonate hydrate (MES) were purchased from Aladdin (Shanghai).

[0057] Bovine serum albumin (BSA) was purchased from Amresco (USA).

[0058] CuCl2·2H2O and polyvinylpyrrolidone (PVP) were purchased from Sinopharm Chemical Reagent Company (Shanghai).

[0059] Tetraethyl orthosilicate (TEOS), 3-aminopropyltriethoxysilane (APTES), and sodium ascorbate were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. (Shanghai).

[0060] Chloramphenicol (CAP), neomycin, and clarithromycin were purchased from Sigma Aldrich (USA).

[0061] Chloramphenicol antibody (CAP-Ab), BSA-chloramphenicol antigen (BSA-CAP-Ag) were purchased from Shengong Biotechnology Co., Ltd. (Shanghai).

[0062] The experimental water was deionized by a water purification system (Millipore, USA). All chemicals were of analytical purity and did not require further purification before use.

[0063] The preparation method of the related reagents used in this example:

[0064] PBS buffer (10 mM, pH = 7.4): Take 8.00 g of NaCl, 0.20 g of KCl, 0.20 g of KH2PO4 and 2.90 g of Na2HPO4·12H2O in a 1000 mL volumetric flask, and dilute to volume, shake well.

[0065] MES buffer (0.1 M, pH = 6.0): Take 21.325 g of MES and dilute to 1000 mL with deionized water as A solution; take 4 g of NaOH and dilute to 1000 mL with deionized water as B solution; mix 1000 mL of A solution and 400 mL of B solution, and shake well.

[0066] PBST, MEST solution: Add 0.5 mL of Tween-20 to the prepared 1000 mL PBS or MES buffer, and shake well.

[0067] The immunoassay method of this example mainly includes the following contents:

[0068] (1) Preparation of CuO2@SiO2 nanomaterials at room temperature

[0069] CuCl2·2H2O (17 mg) and polyvinylpyrrolidone PVP (1 g) were dissolved in 10 mL of ultrapure water, and NH3·H2O (10 mL, 1 mol / L) and H2O2 (200 μL, 10 mol / L) were added dropwise in turn, and stirred for 30 minutes. Add tetraethyl orthosilicate (TEOS, 100 μL), stir the mixture solution for 2 hours, centrifuge (10000 r / min, 5 minutes) to collect the crude product, rinse with ultrapure water three times, and freeze-dry overnight. The obtained CuO2@SiO2 product was stored at 4℃ in the dark for standby.

[0070] (2) Preparation of functionalized glass slides at room temperature

[0071] First, BSA-Azide was coupled in advance. BSA (2 mg / mL, 5 mL) was coupled with Azide-PEG4-NH2(10 mg / mL, 16 μL) in PBS (10 mM, pH = 7.4) and incubated for 2 h, and then the reaction was terminated with Tris-HCl (50 mM, pH = 8, 200 μL).

[0072] Second, functionalized glass slides were prepared. The ultrathin glass slides were soaked in 0.1 M NaOH for 1 h, and then washed with ultrapure water three times to activate the silicon hydroxyl groups on the surface of the ultrathin glass slides. Then the activated glass slides were immersed in an anhydrous ethanol solution containing APTES (1%, v / v) for 12 h to introduce amino groups into the activated glass slides. Subsequently, the glass slides were washed with ultrapure water three times and dried in an oven at 37 °C. After drying, the glass slides were incubated with a GA aqueous solution (2.5%, v / v) for 2 h, and then washed with ultrapure water three times. Finally, the pre-coupled BSA-Azide solution (4 mg / mL) was added dropwise to the surface of the glass slides and incubated for 2 h. After completion, the glass slides were washed with ultrapure water three times, and the functionalized glass slides were stored in a PBS solution at 4 °C.

[0073] (3) Preparation of complexes required for competitive immune reaction

[0074] Synthesis of CuO2@SiO2-CAP_Ab. CuO2@SiO2(2 mg) was soaked in an anhydrous ethanol solution containing APTES (1%, 500 uL, v / v) for 12 h to introduce amino groups onto the surface of the material. Subsequently, it was washed with ultrapure water three times and centrifuged (10000 rpm, 5 min). The dried CuO2@SiO2 was added to a GA solution (500 uL, 2.5%, v / v) and incubated for 2 h, and then washed with ultrapure water three times by centrifugation. Finally, 20 μg of CAP-Ab was added and incubated for 2 h, and then washed with ultrapure water three times by centrifugation. The unbound sites were blocked with BSA (1%, 500 uL) for 30 min, and then washed with PBST three times by centrifugation, and stored at 4 °C.

[0075] Synthesis of nano-magnetic particle-chloramphenicol complete antigen (MNP-CAP_BSA). MNP 150-COOH (1 mg) was washed twice with MEST, and the supernatant was removed after magnetic separation. EDC (5 mg / mL, 50 μL) and NHS (5 mg / mL, 25 μL) were added, respectively, and activated for 30 min at room temperature. After the carboxyl group was activated, it was washed twice with MEST. CAP_BSA (100 μg) was added to the carboxyl-activated MNP, and coupled for 3 h at 37°C. After magnetic separation, the supernatant was removed. Finally, BSA solution (1%, 1000 μL) was added to the above system to block the unbound sites, and reacted for 30 min at 37°C. After washing with PBST for 4 times, MNP-CAP_BSA was resuspended with PBST (1000 μL pH = 7.4, 0.5% BSA) and stored at 4°C.

[0076] (4) Chloramphenicol detection

[0077] MNP-CAP_BSA (20 μg / mL, 100 μL), CAP (100 μL), CuO2@SiO2-CAP_Ab (20 μg / mL, 100 μL) were mixed, and then a competitive immune reaction was carried out at 37°C for 25 min. The magnetic particles and CuO2@SiO2 immune-bound complex were separated by magnetic separation, and then the supernatant was taken. In the competitive immune reaction, the concentration of CAP was quantitatively related to the content of the remaining CuO2@SiO2 complex. Hydrochloric acid (8 mM, 100 μL) was added to the supernatant to release copper ions from the CuO2@SiO2-CAP_Ab after reaction by magnetic separation. The content of copper ions was directly quantitatively related to the content of CuO2@SiO2 complex, and an indirect quantitative relationship between the content of copper ions and the concentration of CAP was established. Then, sodium ascorbate (SA) (5 mM, 10 μL) was added to reduce Cu 2+ to Cu + , and finally the coupled Alkyze-PS 6μm (50 μg / mL, 100 μL) and the converted Cu + were added dropwise to the functionalized glass slide. The content of copper ions was quantitatively related to the polystyrene microspheres involved in the click reaction, thereby establishing a CAP concentration and polystyrene microsphere signal method reaction system. After blowing for 10 min, the glass slide was washed with a slow water flow, and the cleaned ultrathin glass slide was placed in a self-developed holographic imaging system for imaging.

[0078] (5) Modal conversion of unpaired data sets and construction of deep learning target detection data sets

[0079] The unpaired modal transformation dataset is composed of a microscope dataset and a holographic imaging dataset, wherein the microscope dataset is constructed by: gradient dilution of 6 μm polystyrene microspheres, imaging under 40 times magnification by a general optical microscope, random sampling method, then cutting the microscope imaging image into a 512*512 pixel image, and data enhancement for model training, and the microscope dataset has a total of 1200 images.

[0080] The holographic imaging dataset: the polystyrene microspheres with the same concentration gradient as described above are directly imaged on the self-developed lensless holographic imaging device, and the holographic image is subjected to non-local mean filtering and cutting processing, and a random sampling method is used to obtain a holographic image dataset of 400 images, thereby establishing an unpaired dataset for modal transformation.

[0081] In order to ensure the stability and accuracy of microsphere counting, the target detection dataset is composed of modal transformed images and microscope imaging dataset. The microscope modal images generated after modal transformation and the microscope imaging images with equal gradient dilution are subjected to target labeling, combined with a data enhancement algorithm to obtain 2824 target detection datasets, wherein the ratio of training set to validation set is 8:2, and a polystyrene microsphere recognition and counting model is trained using the target detection dataset.

[0082] (6) Unpaired dataset modal transformation algorithm and deep learning target detection algorithm construction

[0083] The modal transformation model uses CycleGAN as the backbone model network for transformation from the holographic domain to the microscope image domain. Unlike the traditional supervised modal generative adversarial network (GAN) which has only one generator and one discriminator and needs one-to-one mapping, the modal transformation model learns the mapping relationship between two different domains (holographic image domain and microscope image domain) to realize image transformation between the two domains, and maintains the consistency of transformation through inverse mapping. The model does not rely on paired datasets, but uses cycle consistency loss to realize image transformation of unpaired datasets.

[0084] The modal transformation model is composed of two mapping functions: generator G: X (holographic image domain) → Y (microscope image domain), generator F: Y (microscope image domain) → X (holographic image domain), discriminator D X , D Y , adversarial loss L1, L2 is used to control the difference of generated images in the domain, and cycle consistency loss Loss CycleThe model discriminator is a 70x70 PatchGAN network, and the generator is a Resnet_9blocks network, which adopts an Unaligned dataset mode. Due to hardware device limitations, the Batch_size is set to 1, and the training round number Epoch is set to 200, of which the number of rounds of learning rate linear decay is 100. During the training process, the initial learning rate of Adam is set to 0.0002, and then the learning rate is linearly decayed, the momentum term of Adam is set to 0.5, the training set image is set to 800x800, and grayscale images are used for training.

[0085] The target quantity quantification model adopts YOLOV7 as the backbone model network, and performs polystyrene microsphere quantity quantification after modal transformation of the image. The model has a multi-scale detection strategy and good generalization ability. The model first extracts specific feature information through a forward propagation network (Conv), then fuses feature information of different scales through an extended efficient layer aggregation network (E-ELAN), and then further extracts features and optimizes the network structure through a spatial pyramid pooling-cross stage partial network (SPPCSP) to maintain the network performance. Through the detection head, a prediction box is generated, and the internal parameters are adjusted through the back propagation algorithm to improve the model performance. Finally, the non-maximum suppression algorithm (NMS) is used to process overlapping predictions to retain the best prediction results. The model pre-training weight is yolov7-e6, the training image size is 640x640, the initial learning rate is 0.01, the Optimizer is SGD, the Optimizer weight decay is 0.0005, the SGD is 0.937, the Epoch is 300, the Batch_size is 32, and a multi-scale training method is used. Among them, MP and Upsample are used for downsampling and upsampling, respectively, and CBS is used for image convolution processing. In order to adapt to the statistics of polystyrene microspheres under the large field of view of holographic imaging, the maximum number of detection frames retained by max-det, i.e. the maximum number of object detections, is set to 10000.

[0086] The present embodiment is based on Python 3.8, Pytorch 3.0.1+cu11.8, Numpy, Opencv, Pandas, Visdom, piq, and other algorithm packages. The hardware used for model training is: GPU NVIDIA GeForce RTX 3090Ti, CPU i9-12900K.

[0087] Example 2: Detection of the content of chloramphenicol (CAP) in aquatic product samples.

[0088] The aquatic product samples were randomly purchased from a supermarket.

[0089] Pre-treatment: The aquatic product sample was ground into minced meat by a grinder. An unknown concentration of CAP solution was randomly added into the minced meat, which was mixed uniformly and then placed at 4℃ overnight. The CAP in the minced meat was extracted by ultrasonic extraction with a mixture of ethyl acetate / acetonitrile (1:1, v / v) twice, each for 30 minutes. The supernatant was dried by nitrogen and then re-suspended in PBS solution for subsequent detection.

[0090] The pre-treated aquatic product sample was detected for the content of chloramphenicol (CAP) by the immunoassay method described in Example 1, and the detection results are shown in Table 1. Figure 11 As shown in Table 1, among the 10 aquatic product samples, three samples were detected to contain CAP, and the same results were detected in the parallel samples, which was the same as the sample amount detected by the gold standard method ELISA.

[0091] Example 3: Detection of the content of chloramphenicol (CAP) in aquaculture wastewater samples

[0092] The aquaculture wastewater samples were randomly collected from an aquaculture area.

[0093] Pre-treatment: The impurities were filtered by using a 0.22 μm organic filter membrane, and an unknown concentration of CAP solution was randomly added for subsequent detection.

[0094] The pre-treated aquaculture wastewater sample was detected for the content of chloramphenicol (CAP) by the immunoassay method described in Example 1, and the detection results are shown in Table 2. Figure 11 As shown in Table 2, among the 10 aquaculture wastewater samples, three samples were detected to contain CAP, and the same results were detected in the parallel samples, which was the same as the sample amount detected by the gold standard method ELISA.

[0095] Example 4: Detection of the content of chloramphenicol (CAP) in urine

[0096] The urine samples were collected from 10 healthy people.

[0097] Pre-treatment: An unknown concentration of CAP solution was randomly added into the urine sample for subsequent detection.

[0098] The pre-treated urine sample was detected for the content of chloramphenicol (CAP) by the immunoassay method described in Example 1, and the detection results are shown in Table 3. Figure 11 As shown in Table 3, among the 10 urine samples, three samples were detected to contain CAP, and the same results were detected in the parallel samples, while one low concentration of CAP was not detected by the gold standard method ELISA, which showed that the present method had higher detection accuracy.

[0099] Some steps in the embodiments of the present application can be realized by software, and the corresponding software program can be stored in a readable storage medium, such as an optical disc or a hard disk.

[0100] The above description is only the preferred embodiment of the present application, and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. An immunoassay method, characterized in that, The method includes: Step 1: Mix the target analyte, the pre-prepared CuO2@SiO2 complex, and the nanomagnetic particle conjugate, and perform a competitive immunoassay; the surface of the CuO2@SiO2 complex is modified with a detection antibody that recognizes the target analyte, and the surface of the nanomagnetic particle conjugate is modified with the complete antigen of the target analyte; Step 2: Magnetic separation, the remaining post-reaction CuO2@SiO2 complex is dissolved in hydrochloric acid and sodium ascorbate to release copper ions; Step 3: Copper ions act as a catalyst to rapidly catalyze the azide-alkyne cycloaddition reaction at room temperature, causing the azide-alkyne cycloaddition reaction between the alkyne-modified polystyrene microspheres and the azide ultrathin functionalized glass slide, i.e., the click reaction; Step 4: Perform holographic imaging on the polystyrene microspheres obtained in Step 3 that are fixed on the functionalized glass slide by click reaction; Step 5: Construct an unpaired mode conversion dataset by combining the holographic imaging image of isogradiently diluted polystyrene microspheres and the microscopic imaging image of isogradiently diluted polystyrene microspheres, and train it using a deep learning mode conversion model. Then, use the trained deep learning mode conversion model to perform mode conversion on the holographic imaging image obtained in Step 4 to obtain the microscopic mode image. Step 6: Input the microscopic modal images obtained after modal conversion and the microscopic imaging images of polystyrene microspheres with equal gradient dilution into a deep learning target detection network for training. Then, use the trained deep learning target detection network to identify and count the microscopic modal images obtained in Step 4 after modal conversion. Step 7: Utilize the linear relationship between the number of polystyrene microspheres and the concentration of the target analyte to obtain the concentration information of the target analyte.

2. The immunoassay method according to claim 1, characterized in that, The deep learning modality conversion model uses CycleGAN as its backbone network.

3. The immunoassay method according to claim 1, characterized in that, The deep learning object detection network uses YOLOv7 as its backbone.

4. The immunoassay method according to claim 1, characterized in that, The copper ions are monovalent copper ions generated by the reduction reaction of sodium ascorbate.

5. The immunoassay method according to claim 1, characterized in that, The click reaction is a process in which azide and alkyne react to form a covalent bond under the catalysis of monovalent copper ions.

6. The immunoassay method according to claim 1, characterized in that, The polystyrene microspheres are 6μm polystyrene microspheres with alkynes modified on their surface.

7. The immunoassay method according to claim 1, characterized in that, The method also includes processing the holographic image using a nonlocal mean filtering algorithm.

8. The immunoassay method according to claim 1, characterized in that, The nanomagnetic particle conjugate is a 150nm nanomagnetic particle whose surface is modified with the complete antigen of the target analyte.

9. The immunoassay method according to claim 1, characterized in that, The concentration of the nanomagnetic particle coupling compound added to the reaction system was 20 μg / mL.

10. The immunoassay method according to claim 1, characterized in that, The concentration of the CuO2@SiO2 composite added to the reaction system was 20 μg / mL.

Images