WO2023138162A1 - 样本检测方法、装置、系统、电子设备及计算机可读介质 - Google Patents

样本检测方法、装置、系统、电子设备及计算机可读介质 Download PDF

Info

Publication number
WO2023138162A1
WO2023138162A1 PCT/CN2022/129869 CN2022129869W WO2023138162A1 WO 2023138162 A1 WO2023138162 A1 WO 2023138162A1 CN 2022129869 W CN2022129869 W CN 2022129869W WO 2023138162 A1 WO2023138162 A1 WO 2023138162A1
Authority
WO
WIPO (PCT)
Prior art keywords
sample
detection
generate
tested
target
Prior art date
Application number
PCT/CN2022/129869
Other languages
English (en)
French (fr)
Inventor
井超
Original Assignee
北京流荧生物科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京流荧生物科技有限公司 filed Critical 北京流荧生物科技有限公司
Publication of WO2023138162A1 publication Critical patent/WO2023138162A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0012—Biomedical image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
    • G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
    • G16B25/00—ICT specially adapted for hybridisation; ICT specially adapted for gene or protein expression
    • G16B25/20—Polymerase chain reaction [PCR]; Primer or probe design; Probe optimisation

Definitions

  • the present application relates to the field of in vitro detection, in particular, to a sample detection method, device, system, electronic equipment and computer-readable medium based on a single optical detection channel.
  • IVD In Vitro Diagnosis
  • chemiluminescence immunoassay CLIA including direct chemiluminescence assay, chemiluminescence enzyme immunoassay (CLEIA), electrochemiluminescence immunoassay ECLIA/ECL
  • chemiluminescence immunoassay CLIA including direct chemiluminescence assay, chemiluminescence enzyme immunoassay (CLEIA), electrochemiluminescence immunoassay ECLIA/ECL
  • SIMOA single molecule immunoassay
  • SIMOA enzyme-linked immunosorbent assay ELISA
  • MSD multifactor ultrasensitive electrochemiluminescence immunoassay Meso Scale Discovery
  • PCR loop-mediated isothermal amplification LAMP; other isothermal nucleic acid amplification assays; other molecular diagnostic techniques (gene sequencing), etc.
  • the modes of qualitative or quantitative detection by optical means are roughly divided into the following types: 1. Observation of color or fluorescence intensity with the naked eye based on a microscope; 2. Optical sensors (such as photon counters, etc.) measure color absorption, luminous intensity or fluorescence intensity; 3. Cameras detect and identify color changes or fluorescence directly or through a microscope, generally using image processing technology.
  • a multi-channel optical system is required, which increases the complexity of the optical path design of the optical equipment and the cost of the instrument. Each channel has a different combination of filter lenses and beam splitters, and some technologies require special consumables. Complex detection means limit the development of in vitro diagnostic technology. Therefore, there is a need for a new sample detection method, device, system, electronic device and computer-readable medium based on a single optical detection channel.
  • the present application provides a sample detection method, device, system, electronic device, and computer-readable medium based on a single optical detection channel, which can realize multiplexed detection through a single-channel optical system, and simultaneously perform qualitative or quantitative accurate identification of multiple target analytes without the need for special consumables.
  • a sample detection method based on a single optical detection channel comprising: obtaining a sample to be tested, the sample to be tested contains multiple target analytes; the sample to be tested is subjected to biochemical reaction processing based on magnetic particles to generate a reaction solution, the reaction solution contains multiple multiple complexes, wherein the multiple multiple complexes correspond to multiple target analytes; the reaction solution is excited so that the multiple multiple complexes emit multiple lights;
  • the optical detection channel acquires the plurality of lights to generate a sample detection picture; generates a detection result of each target analyte in the multiple target analytes based on the sample detection picture; generates a detection result of the sample to be tested based on the detection result of each target analyte.
  • performing a biochemical reaction treatment based on magnetic particles on the sample to be tested to generate a reaction solution includes: performing a non-enzymatic biochemical reaction treatment based on magnetic particles on the sample to be tested based on biocarrier technology to generate the reaction solution.
  • the non-enzymatic biochemical reaction treatment based on magnetic particles is performed on the sample to be tested based on biological carrier technology to generate the reaction solution, including: performing non-enzymatic biochemical reaction treatment on the sample to be tested so that multiple target analytes in the sample to be tested are respectively combined with multiple biological carriers to generate the reaction solution; using magnetic particles, the target analyte is separated under the action of an external magnetic field.
  • performing non-enzymatic biochemical reaction treatment on the sample to be tested so that multiple target analytes in the sample to be tested are respectively combined with multiple biological carriers to generate the reaction solution includes: performing non-enzymatic biochemical reaction treatment on the sample to be tested so that multiple target analytes in the sample to be tested are respectively combined with corresponding magnetic particles, antibodies, analytes, and fluorescent nanobead signal probes to generate the reaction solution.
  • exciting the reaction solution so that the multiple multi-component complexes emit multiple lights includes: exciting the reaction solution through a light source device or an electrical device or a chemical reagent to make the multiple multi-component complexes emit multiple lights, and the multiple lights include lights with different intensities and/or frequencies.
  • exciting the reaction solution to make the multiple multi-component complexes emit multiple lights includes: exciting the reaction solution with a light source to make the multiple multi-component complexes emit multiple lights.
  • the reaction solution is excited by a light source so that the multiple multi-component complexes emit multiple lights, comprising: placing the reaction solution on a transparent plane; the magnetic particles in the solution to be reacted are deposited on the surface of the plane, and the reaction solution placed on the transparent plane is excited by a blue light source so that the multiple multi-component complexes emit multiple lights.
  • acquiring the multiple lights based on a single optical detection channel to generate a sample detection picture includes: acquiring the multiple lights based on a filter lens in a single optical detection channel; an optical sensor generating the sample detection picture based on the multiple lights.
  • the optical sensor generating the sample detection picture based on the multiple lights includes: a CMOS optical sensor generating the sample detection picture based on the multiple lights; or a CCD optical sensor generating the sample detection picture based on the multiple lights.
  • generating the detection result of each of the multiple target analytes based on the sample detection picture includes: preprocessing the sample detection picture; generating a feature matrix of the sample detection picture; inputting the feature matrix into a multi-target recognition model to generate the detection result of each of the multiple target analytes.
  • performing preprocessing on the sample detection picture includes: performing filtering processing on the sample detection picture; and/or performing speckle detection processing on the sample detection picture.
  • generating the feature matrix of the sample detection picture includes: extracting feature data of each pixel in the sample detection picture; and generating the feature matrix based on feature data of all pixels.
  • extracting feature data of each pixel in the sample detection picture includes: extracting the color and position of each pixel in the sample detection picture.
  • generating the feature matrix based on feature data of all pixels further includes: performing feature transformation processing on the feature matrix.
  • inputting the feature matrix into a multi-target recognition model to generate a detection result for each target analyte in the multiple target analytes includes: inputting the feature matrix into the multi-target recognition model; the multi-target recognition model classifies and predicts pixels in the sample detection picture one by one based on the feature matrix; generates a detection result for each target analyte in the multiple target analytes based on the classification prediction results.
  • generating the detection result of each target analyte in the plurality of target analytes includes: generating the total number of pixels and the total brightness corresponding to each target analyte in the plurality of target analytes; generating the detection result of each target analyte in the plurality of target analytes based on the total number of pixels and the total brightness.
  • generating the detection result of each target analyte among the plurality of target analytes based on the total number of pixels and the total brightness includes: generating the corresponding quantity and concentration of each target analyte based on the calibration formula and the total number of pixels and total brightness corresponding to each target analyte; generating the detection result of the sample to be tested based on the corresponding quantity and concentration of each target analyte.
  • it also includes: generating a plurality of sample training pictures through a plurality of detection samples; respectively labeling a plurality of training labels for the plurality of sample training pictures, wherein the plurality of training labels correspond to the various target analytes; training a machine learning model based on a plurality of sample training pictures with a plurality of training labels to generate the multi-target recognition model.
  • labeling the multiple sample training pictures with multiple training labels respectively includes: labeling each pixel in the multiple sample training pictures with a training label respectively.
  • training the machine learning model based on a plurality of sample training pictures with multiple training labels to generate the multi-target recognition model includes: training a multi-layer perceptron based on a plurality of sample training pictures with multiple training labels to generate the multi-target recognition model; or training a random forest classifier based on a plurality of sample training pictures with multiple training labels to generate the multi-target recognition model.
  • a sample detection device based on a single optical detection channel includes: a sample module, used to obtain a sample to be tested, and the sample to be tested contains multiple target analytes; a reaction module, used to perform biochemical reaction processing on the sample to be tested based on magnetic particles to generate a reaction solution, the reaction solution contains multiple multicomponent complexes, wherein the multiple multiplex complexes correspond to multiple target analytes; an excitation separation module, used to excite the reaction solution so that the multiple multiplex complexes emit multiple lights;
  • the image module is used to acquire the multiple lights based on a single optical detection channel to generate a sample detection image; the detection module is used to generate the detection result of each target analyte in the multiple target analytes based on the sample detection image; the result module is used to generate the detection result of the sample to be tested based on the detection result of each target analyte.
  • the detection module includes: a processing unit for preprocessing the sample detection picture; a feature unit for generating a feature matrix of the sample test picture; a recognition unit for inputting the feature matrix into a multi-target recognition model to generate a detection result for each target analyte among the multiple target analytes.
  • the detection module further includes: a model training unit, configured to generate a plurality of sample training pictures through a plurality of detection samples; a plurality of training labels are respectively marked for a plurality of sample training pictures, wherein the plurality of training labels correspond to the various target analytes; a machine learning model is trained based on a plurality of sample training pictures with a plurality of training labels to generate the multi-target recognition model.
  • a model training unit configured to generate a plurality of sample training pictures through a plurality of detection samples
  • a plurality of training labels are respectively marked for a plurality of sample training pictures, wherein the plurality of training labels correspond to the various target analytes
  • a machine learning model is trained based on a plurality of sample training pictures with a plurality of training labels to generate the multi-target recognition model.
  • a sample detection system based on a single optical detection channel includes: a reaction device for obtaining a sample to be tested, the sample to be tested contains multiple target analytes; the sample to be tested is subjected to a biochemical reaction treatment based on magnetic particles to generate a reaction solution, the reaction solution contains multiple multiplex complexes, wherein the multiple multiplex complexes correspond to multiple target analytes; an excitation separation device is used to excite the reaction solution so that the multiple multiplex complexes emit multiple lights; collected on the surface of the magnetic particle; an image device, used to acquire the plurality of lights based on a single optical detection channel to generate a sample detection picture; a server, used to generate the detection result of each target analyte in the multiple target analytes based on the sample detection picture; generate the detection result of the sample to be tested based on the detection result of each target analyte.
  • an electronic device which includes: one or more processors; a storage device for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the method as above.
  • a computer-readable medium on which a computer program is stored, and when the program is executed by a processor, the above method is implemented.
  • the sample to be tested contains multiple target analytes; the sample to be tested is subjected to biochemical reaction processing based on magnetic particles to generate a reaction solution, the reaction solution contains multiple multiplex compounds, wherein the multiple multiplex complexes correspond to multiple target analytes; the reaction solution is excited so that the multiple multiplex complexes emit multiple lights; the multiple lights are obtained based on a single optical detection channel to generate a sample detection image;
  • the detection result of each target analyte among the multiple target analytes; the method of generating the detection result of the sample to be tested based on the detection result of each target analyte can realize multiplex detection through a single-channel optical system, and simultaneously perform qualitative or quantitative accurate identification of multiple target analytes without the need for special consumables.
  • Fig. 1 is a schematic diagram of a sample detection system based on a single optical detection channel according to an exemplary embodiment.
  • Fig. 2 is a flow chart of a sample detection method based on a single optical detection channel according to an exemplary embodiment.
  • Fig. 3 is a schematic diagram of a sample detection method based on a single optical detection channel according to another exemplary embodiment.
  • Fig. 4 is a schematic diagram of a sample detection method based on a single optical detection channel according to another exemplary embodiment.
  • Fig. 5 is a flow chart of a sample detection method based on a single optical detection channel according to an exemplary embodiment.
  • Fig. 6 is a flow chart of a sample detection method based on a single optical detection channel according to an exemplary embodiment.
  • Fig. 7 is a flow chart of a sample detection method based on a single optical detection channel according to an exemplary embodiment.
  • Fig. 8 is a block diagram of a sample detection device based on a single optical detection channel according to another exemplary embodiment.
  • Fig. 9 is a block diagram of an electronic device according to an exemplary embodiment.
  • Fig. 10 is a block diagram showing a computer readable medium according to an exemplary embodiment.
  • Example embodiments will now be described more fully with reference to the accompanying drawings.
  • Example embodiments may, however, be embodied in many forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided so that this application will be thorough and complete, and will fully convey the concept of example embodiments to those skilled in the art.
  • the same reference numerals denote the same or similar parts in the drawings, and thus their repeated descriptions will be omitted.
  • Fig. 1 is a schematic diagram of a sample detection system based on a single optical detection channel according to an exemplary embodiment.
  • a sample detection system 10 based on a single optical detection channel may include a reaction device 101 , an excitation separation device 102 , an image device 103 , a network 104 and a server 105 .
  • the network 104 is used as a medium for providing a communication link between the image device 103 and the server 105 .
  • Network 104 may include various connection types, such as wires, wireless communication links, or fiber optic cables, among others.
  • the reaction device 101 is used to obtain a sample to be tested, the sample to be tested contains multiple target analytes; the sample to be tested is subjected to a biochemical reaction treatment based on magnetic particles to generate a reaction solution, and the reaction solution contains multiple multiple complexes, wherein the multiple multiple complexes correspond to multiple target analytes;
  • the excitation separation device 102 is used to excite the reaction solution so that the multiple multi-component complexes emit various lights; under the action of an external magnetic field, the target analyte is separated from the sample to be tested and collected on the surface of the magnetic particle;
  • the image device 103 is used to acquire the multiple lights based on a single optical detection channel to generate a sample detection picture
  • the server 105 is configured to generate a detection result of each target analyte among the plurality of target analytes based on the sample detection picture; and generate a detection result of the sample to be tested based on the detection result of each target analyte.
  • the server 105 may be an entity server, and may also be composed of multiple servers, for example, a part of the server 105 may be used as a sample detection device in this application, for example, to generate a detection result of each target analyte in the multiple target analytes based on the sample detection picture; generate a detection result of the sample to be tested based on the detection result of each target analyte; and a part of the server 105 may also be used, for example, as a model training device in the present application, for generating multiple sample training pictures through multiple detection samples; labeling multiple training labels for the multiple sample training pictures , wherein the multiple training labels correspond to the multiple target analytes; the machine learning model is trained based on multiple sample training pictures with multiple training labels to generate the multi-target recognition model.
  • the excitation method can be excitation by a fluorescent light source.
  • the functions of the excitation separation device 102 and the imaging device 103 can be realized by a single-channel optical detection device in the prior art.
  • the sample detection system based on a single optical detection channel in this application may include: a reaction device, an optical detection device, a network and a server.
  • Fig. 2 is a flow chart of a sample detection method based on a single optical detection channel according to an exemplary embodiment.
  • the sample detection method 20 based on a single optical detection channel at least includes steps S202 to S212.
  • a sample to be tested is obtained, and the sample to be tested contains multiple target analytes.
  • the sample to be tested can be a serum sample of a patient, and there may be multiple target analytes in the sample, and the target analytes can include: p24 protein produced by HIV-1 virus, N protein produced by SARS-COV-2 virus, etc.
  • the target analytes can include: p24 protein produced by HIV-1 virus, N protein produced by SARS-COV-2 virus, etc.
  • the biochemical reaction treatment based on magnetic particles is performed on the sample to be tested to generate a reaction solution, the reaction solution contains multiple multi-component complexes, wherein the multiple multi-component complexes correspond to multiple target analytes.
  • the sample to be tested can be subjected to biochemical reaction treatment based on biological carrier technology to generate the reaction solution.
  • the biochemical reaction may be a non-enzymatic biochemical reaction.
  • the magnetic particles can be magnetic beads in the prior art. Based on magnetic particles, the target can be quickly and accurately captured and separated. In the traditional method of the prior art, the biochemical reaction takes several hours, and at the same time, the capture efficiency for low-concentration target substances is also very low. In this application, the biochemical reaction based on magnetic particles improves the reaction speed and shortens the detection time in practical applications.
  • the sample to be tested may be subjected to non-enzymatic biochemical reaction treatment based on magnetic particles based on biological carrier technology to generate the reaction solution. More specifically, for example, non-enzymatic biochemical reaction treatment is performed on the sample to be tested so that multiple target analytes in the sample to be tested are combined with multiple biological carriers to generate the reaction solution; magnetic particles are used to separate the target analytes under the action of an external magnetic field.
  • the sample to be tested can be treated with non-enzymatic biochemical reactions so that multiple target analytes in the sample to be tested can be combined with corresponding magnetic particles, antibodies, analytes, and fluorescent nanobead signal probes to generate the reaction solution.
  • the serum sample can be subjected to immunobiochemical reaction processing, and the specific processing process can refer to the description in FIG. 3 .
  • Magnetic particle-p24 capture antibody-p24 protein-p24 detection antibody-green fluorescent nanobead signal probe
  • the surface of the magnetic particles may capture multiple corresponding proteins and nano-scale fluorescent nano-bead signal probes, or may not capture them at all.
  • the protein and nano-scale fluorescent nano-bead signal probes are not captured, the magnetic particles remain in their original state. Therefore, after the immunobiochemical reaction, the reaction solution may also contain magnetic particles.
  • the target analyte is separated from the sample to be tested under the action of an external magnetic field and collected on the surface of the magnetic particle.
  • the reaction solution is excited to make the multiple multi-component complexes emit multiple lights.
  • the reaction solution can be excited by light source equipment or electrical equipment or chemical reagents to make the multiple multi-component complexes emit multiple lights. More specifically, the reaction solution can be excited by light sources such as halogen lamps and LEDs. It can also be excited by light sources such as mercury vapor lamps, LEDs, and lasers.
  • the reaction solution can be excited by a fluorescent light source so that the multiple multi-component complexes emit multiple fluorescences.
  • reaction solution can be placed on a transparent plane; the magnetic particles in the solution to be reacted are deposited on the surface of the plane, and the reaction solution placed on the transparent plane is excited by a blue light source so that the multiple multi-component complexes emit various lights.
  • the reacted solution can be placed on a transparent plane with low fluorescence noise.
  • the fluorescence noise parameter of the transparent plane can be the fluorescence emission intensity of the autofluorescence of the material.
  • the unit is: Arbitrary Unit (depending on the testing instrument). Considerable materials: glass, polymethylmethacrylate (PMMA/polymethylmethacrylate), cycloolefin polymer (COP/Cyclo Olefin Polymer), etc.
  • the fluorescence intensity of a substance is related to the wavelength of the excitation light and the wavelength of the measured emission light.
  • the data of the fluorescence intensity of the substance is expressed in the form of a matrix.
  • the rows and columns correspond to different excitation light wavelengths and emission light wavelengths.
  • Each matrix element is the fluorescence intensity of the excitation light and emission light wavelength, which is called the Excitation Emission Matrix (EEM).
  • EEM Excitation Emission Matrix
  • the light emitted by the target analyte is not a single wavelength of light, but light across a band, and the intensity distribution in the entire band is completely determined by the excited substance itself.
  • the use of an emission filter matched with blue fluorescence does not filter out 100% of other color light in the wavelength band away from green light, so after excitation by blue fluorescence, different non-green fluorescence corresponding to various target analytes can be displayed in the image.
  • the multiple lights are acquired based on a single optical detection channel to generate a sample detection picture.
  • the plurality of lights may be acquired based on a filter lens in a single optical detection channel; an optical sensor generates the sample detection picture based on the plurality of lights.
  • the sample detection picture can be generated based on the multiple lights by a CMOS optical sensor; the sample detection picture can also be generated based on the multiple lights by a CCD optical sensor.
  • Excitation can be done in different ways. It can be excited by a white light source. At this time, all magnetic particles are captured by the CCD camera sensor under bright field conditions.
  • various particles in the reaction solution will be excited to emit light of different colors through the excitation of the blue light source and the filtering of the corresponding filter lens, and the particles of different colors can be captured by the CCD sensor, in which the magnetic particles show yellow, the green fluorescent nanobead signal probe shows green, the red fluorescent nanobead signal probe shows orange, and the background shows black.
  • the blue light excitation method can obtain different lights of various target analytes, and each substance has a different light band and intensity combination, which is beneficial to identify different target analytes.
  • the detection result of each target analyte in the plurality of target analytes is generated based on the sample detection picture, and the sample detection picture can be preprocessed; the feature matrix of the sample detection picture is generated; the feature matrix is input into the multi-target recognition model, and the detection result of each target analyte in the multiple target analytes is generated.
  • the detection result of the sample to be tested is generated based on the detection result of each target analyte.
  • the quantity and concentration of p24 protein and N protein can be obtained indirectly through the conversion of the calibration formula, and the qualitative or quantitative judgment of the patient's health status can be obtained.
  • the calibration formula can be obtained from the fitting curve in the experiment. For example, before the actual application, multiple preclinical experimental data are obtained, and based on the clinical test data, the curve of the number of fluorescent particles VS the protein concentration of the controlled quantity is calculated, and then a calibration formula is generated.
  • the sample to be tested contains multiple target analytes; the sample to be tested is subjected to a biochemical reaction treatment based on magnetic particles to generate a reaction solution, and the reaction solution contains multiple multiplexes.
  • a sample detection picture By generating a sample detection picture; generating a detection result of each of the multiple target analytes based on the sample detection picture; and generating a detection result of the sample to be tested based on the detection result of each target analyte, multiplex detection can be realized through a single-channel optical system, and multiple target analytes can be accurately identified qualitatively or quantitatively at the same time without the need for special consumables.
  • the sample detection method based on a single optical detection channel of the present application realizes the multiplexed detection of a single optical detection channel, and one imaging picture can cover information of various luminescent substances, and then indirectly calculate the quantity of corresponding various target analytes and their concentrations in the sample species.
  • the sample detection method based on a single optical detection channel of the present application, after image preprocessing, the remaining effective pixels after each preprocessing are scanned, and the picture information of the effective pixels is converted into data information.
  • Machine learning processes data information, which speeds up the processing speed and accuracy of the machine learning model.
  • the sample detection method based on a single optical detection channel of the present application realizes the identification and counting of luminescent substances one by one without the need for special consumables (such as microfluidic kits) and corresponding fluid control.
  • Fig. 5 is a flow chart of a sample detection method based on a single optical detection channel according to another exemplary embodiment.
  • the process 50 shown in FIG. 5 is a detailed description of S210 "generating the detection result of each of the multiple target analytes based on the sample detection picture" in the process shown in FIG. 2 .
  • preprocessing is performed on the sample detection picture.
  • Filter processing may be performed on the sample detection picture; speckle detection processing may also be performed on the sample detection picture.
  • a feature matrix of the sample detection picture is generated.
  • the feature data of each pixel in the sample detection picture may be extracted; the feature matrix is generated based on the feature data of all pixels.
  • the color and position of each pixel in the sample detection picture may be extracted as feature data.
  • feature transformation processing may also be performed on the feature matrix, and after feature transformation processing, the feature matrix is generated based on the colors and positions of all pixels. More specifically, feature transformation processing such as data scaling and label encoding can be performed on the feature matrix. Pixel compression processing can also be performed on the feature matrix, and pixels whose brightness is lower than a certain threshold can be classified as background colors, so that subsequent image analysis is not performed on the pixels to save computing time and computing space.
  • the feature matrix is input into a multi-target recognition model to generate a detection result for each target analyte in the multiple target analytes.
  • the feature matrix can be input into a multi-target recognition model; the multi-target recognition model classifies and predicts the pixels in the sample detection picture one by one based on the feature matrix; generates the detection result of each target analyte in the multiple target analytes based on the classification prediction results.
  • the total number of pixels and the total brightness corresponding to each target analyte among the plurality of target analytes can be generated; and the detection result of each target analyte among the plurality of target analytes can be generated based on the total number of pixels and the total brightness.
  • the quantity and concentration corresponding to each target analyte can be generated based on the calibration formula and the total number of pixels and total brightness corresponding to each target analyte; the detection result of the sample to be tested can be generated based on the quantity and concentration corresponding to each target analyte.
  • Fig. 6 is a flow chart of a sample detection method based on a single optical detection channel according to another exemplary embodiment.
  • the process 60 shown in FIG. 6 is a supplementary description to the process shown in FIG. 5 .
  • multiple sample training pictures are generated from multiple detection samples.
  • different machine learning models need to be trained for different samples. Serum samples from multiple patients can be obtained as training samples, and then biochemical reactions and excitations are performed according to the method in this application to generate multiple sample training pictures.
  • Sample training pictures can also be generated for the patient's body fluids, urine, etc. for subsequent training of the machine learning model.
  • a plurality of training labels are respectively labeled for the plurality of sample training pictures, wherein the plurality of training labels correspond to the plurality of target analytes. More specifically, a training label may be marked for each pixel in the plurality of sample training pictures.
  • Each pixel in the sample training image is marked with a training label, and the training labels can be target analyte 1, target analyte 2, target analyte 3, and so on.
  • the types of marked target analytes may be more than the detected target analytes in the actual analysis, such as the specific examples above in this application, the detected target analytes are: p24 protein produced by HIV-1 virus, N protein produced by SARS-COV-2 virus, but corresponding to this sample detection, the X protein corresponding to other viruses was also detected in the machine learning model used.
  • the type of target analyte that the machine learning model can detect depends on the type of target analyte in the sample.
  • the machine learning model is trained based on multiple sample training pictures with multiple training labels to generate the multi-target recognition model.
  • the multi-layer perceptron can be trained based on multiple sample training pictures with multiple training labels to generate the multi-target recognition model; the random forest classifier can also be trained based on multiple sample training pictures with multiple training labels to generate the multi-target recognition model.
  • the data corresponding to 60% of the pixels in the sample training image can be used as the training data set to train the machine learning model.
  • Model hyperparameters can also be optimized through ten-fold cross-validation. For multi-layer perceptrons, hyperparameters include the number of neurons in the hidden layer, activation function types, etc.; for random forest classifiers, hyperparameters include the number of decision trees, maximum depth type, etc.
  • the optimal parameters and the optimal feature transformation methods of multi-layer perceptron and random forest classifier are also selected.
  • an initial model is constructed respectively, and the pixel feature matrix in the sample training image is input into the initial model to calculate the predicted label, and the predicted label is compared with the corresponding real label to determine whether the predicted label is consistent with the real label, count the number of predicted labels consistent with the real label, and calculate the proportion of the number of predicted labels consistent with the real label in the number of all predicted labels.
  • the method for adjusting the parameters in the initial model may be performed by using a stochastic gradient descent algorithm, a gradient descent algorithm or a normal equation. If the number of times to adjust the parameters of the initial model exceeds the preset number of times, the model used to construct the adjusted model can be replaced to improve model training efficiency.
  • the data corresponding to 40% of the pixels can also be taken as the verification data set, and the final evaluation of the model is performed using the machine-related measurement parameters of the confusion matrix. If the final evaluation fails, more high-quality patient sample pictures can be collected to retrain the model, and different feature engineering schemes can be adopted to adjust the model.
  • the model after the model training is completed, the model can be deployed to make predictions on patient samples. After the model is deployed and implemented, the calculation effect of the model can also be tested to confirm that the model continues to be effective.
  • the detection method can be: compare the recognition of the image with the preset parameter threshold, and compare the diagnosis result with other reliable reference results in reality.
  • the optics used in practice and in training will be calibrated. If there is a problem with the optics used in training, the model will be retrained after the fix. If the data is abnormal, it means that the training samples do not cover all patient conditions. Retrain the model on data that adds anomalies to the training data.
  • Fig. 7 is a flow chart of a sample detection method based on a single optical detection channel according to another exemplary embodiment.
  • the process 70 shown in FIG. 7 is an application description of the method in this application in an actual application scenario.
  • a sample to be tested is obtained, which may be, for example, a patient's blood sample, urine sample, saliva or other body fluids.
  • a reaction solution is generated after the biochemical reaction.
  • Serum samples were processed for immunobiochemical reactions using magnetic particles, capture antibodies, detection antibodies and fluorescent nanobead signal probes.
  • the senor acquires digitized images or videos. Place the reacted solution on a transparent surface with low optical noise. Excitation by the blue light source and filtering by the corresponding filter lens. Various particles will be excited to emit different light colors, and different particles can be captured by the CCD sensor. Among them, the magnetic particles are yellow, the green light particles are green, the red light particles are orange, and the background is black.
  • a diagnosis result is generated.
  • the quantity and concentration of p24 protein and N protein can be obtained indirectly through the conversion of the calibration formula, and the qualitative or quantitative judgment of the patient's health status can be obtained.
  • a treatment plan is recommended according to the diagnosis result.
  • Preliminary treatment methods corresponding to different diagnosis results can be set in the system in advance. After the diagnosis results of the patients in this test are obtained, the treatment plan will be automatically retrieved to assist the doctor in on-site processing and speed up.
  • the sample detection method based on a single optical detection channel of the present application realizes the multiplexed detection of a single optical detection channel, and one imaging picture can cover information of various luminescent substances, and then indirectly calculate the quantity of corresponding various target analytes and their concentrations in the sample species.
  • the preprocessing of traditional image processing technology assists the final recognition of machine learning, but in the sample detection method based on a single optical detection channel in this application, after preprocessing, the remaining effective pixels after each preprocessing are scanned, and the image information of the effective pixels is converted into data information.
  • Machine learning processes data information, which speeds up the processing speed and accuracy of the machine learning model.
  • the sample detection method based on a single optical detection channel of the present application realizes the identification and counting of luminescent substances one by one without the need for special consumables (such as microfluidic kits) and corresponding fluid control.
  • Fig. 8 is a block diagram of a sample detection device based on a single optical detection channel according to an exemplary embodiment.
  • a sample detection device 80 based on a single optical detection channel includes: a sample module 802 , a reaction module 804 , an excitation separation module 806 , an image module 808 , a detection module 810 , and a result module 812 .
  • the detection module 810 may include: a processing unit 8102 , a feature unit 8104 , an identification unit 8106 , and a model training unit 8108 .
  • the sample module 802 is used to obtain a sample to be tested, which contains multiple target analytes
  • the reaction module 804 is used to perform biochemical reaction processing on the sample to be tested based on magnetic particles to generate a reaction solution, the reaction solution contains a variety of multiple complexes, wherein the multiple multiple complexes correspond to multiple target analytes;
  • the excitation and separation module 806 is used to excite the reaction solution so that the various multi-component complexes emit various lights; under the action of an external magnetic field, the target analyte is separated from the sample to be tested and collected on the surface of the magnetic particle;
  • the picture module 808 is used to acquire the multiple lights based on a single optical detection channel to generate a sample detection picture
  • the detection module 810 is used to generate a detection result of each target analyte in the plurality of target analytes based on the sample detection picture;
  • the result module 812 is used for generating the detection result of the sample to be tested based on the detection result of each target analyte.
  • the detection module 810 includes: a processing unit 8102 for preprocessing the sample detection picture; a feature unit 8104 for generating a feature matrix of the sample test picture; a recognition unit 8106 for inputting the feature matrix into a multi-target recognition model to generate a detection result for each target analyte in the multiple target analytes.
  • the detection module 810 also includes: a model training unit 8108 for generating a plurality of sample training pictures through a plurality of detection samples; labeling a plurality of training labels for the plurality of sample training pictures, wherein the plurality of training labels correspond to the various target analytes; training a machine learning model based on a plurality of sample training pictures with a plurality of training labels to generate the multi-target recognition model.
  • a model training unit 8108 for generating a plurality of sample training pictures through a plurality of detection samples; labeling a plurality of training labels for the plurality of sample training pictures, wherein the plurality of training labels correspond to the various target analytes; training a machine learning model based on a plurality of sample training pictures with a plurality of training labels to generate the multi-target recognition model.
  • the sample to be tested contains multiple target analytes; the sample to be tested is subjected to biochemical reaction processing based on magnetic particles to generate a reaction solution, and the reaction solution contains multiple multiple complexes, wherein the multiple multiple complexes correspond to multiple target analytes; the reaction solution is excited so that the multiple multiple complexes emit multiple lights; target analytes are separated from the sample to be tested and collected on the surface of the magnetic particle under the action of an external magnetic field; the multiple multiple complexes are obtained based on a single optical detection channel.
  • Fig. 9 is a block diagram of an electronic device according to an exemplary embodiment.
  • FIG. 9 An electronic device 900 according to this embodiment of the present application is described below with reference to FIG. 9 .
  • the electronic device 900 shown in FIG. 9 is only an example, and should not limit the functions and scope of use of this embodiment of the present application.
  • electronic device 900 takes the form of a general-purpose computing device.
  • Components of the electronic device 900 may include, but are not limited to: at least one processing unit 910, at least one storage unit 920, a bus 930 connecting different system components (including the storage unit 920 and the processing unit 910), a display unit 940, and the like.
  • the storage unit stores program codes, and the program codes can be executed by the processing unit 910, so that the processing unit 910 executes the steps described in this specification according to various exemplary implementations of the present application.
  • the processing unit 910 may execute the steps shown in FIG. 2 , FIG. 5 , FIG. 6 , and FIG. 7 .
  • the storage unit 920 may include a readable medium in the form of a volatile storage unit, such as a random access storage unit (RAM) 9201 and/or a cache storage unit 9202 , and may further include a read-only storage unit (ROM) 9203 .
  • RAM random access storage unit
  • ROM read-only storage unit
  • the storage unit 920 may also include a program/utility 9204 having a set (at least one) of program modules 9205.
  • program modules 9205 include, but are not limited to: an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include the implementation of a network environment.
  • Bus 930 may represent one or more of several types of bus structures, including a memory cell bus or memory cell controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus structures.
  • the electronic device 900 may also communicate with one or more external devices 900′ (e.g., keyboards, pointing devices, Bluetooth devices, etc.), enabling a user to communicate with any device that the electronic device 900 interacts with, and/or any device that enables the electronic device 900 to communicate with one or more other computing devices (e.g., a router, modem, etc.). Such communication may occur through input/output (I/O) interface 950 .
  • the electronic device 900 can also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN) and/or a public network such as the Internet) through the network adapter 960 .
  • networks such as a local area network (LAN), a wide area network (WAN) and/or a public network such as the Internet
  • the network adapter 960 can communicate with other modules of the electronic device 900 through the bus 930 . It should be understood that although not shown in the figure, other hardware and/or software modules may be used in conjunction with the electronic device 900, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, among others.
  • the technical solution according to the embodiment of the present application can be embodied in the form of a software product, and the software product can be stored in a non-volatile storage medium (which can be a CD-ROM, a U disk, a mobile hard disk, etc.) or on the network, and include several instructions so that a computing device (which can be a personal computer, a server, or a network device, etc.) executes the above-mentioned method according to the embodiment of the application.
  • a computing device which can be a personal computer, a server, or a network device, etc.
  • the software product may utilize any combination of one or more readable media.
  • the readable medium may be a readable signal medium or a readable storage medium.
  • the readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any combination thereof. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, random access memory (RAM), read only memory (ROM), erasable programmable read only memory (EPROM or flash memory), optical fiber, portable compact disk read only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the foregoing.
  • the computer readable storage medium may include a data signal carrying readable program code in baseband or as part of a carrier wave traveling as a data signal. Such propagated data signals may take many forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing.
  • a readable storage medium may also be any readable medium other than a readable storage medium that can send, propagate or transport a program for use by or in conjunction with an instruction execution system, apparatus or device.
  • the program code contained on the readable storage medium may be transmitted by any suitable medium, including but not limited to wireless, cable, optical cable, RF, etc., or any suitable combination of the above.
  • Program code for carrying out the operations of the present application may be written in any combination of one or more programming languages, including object-oriented programming languages—such as Java, C++, etc., and conventional procedural programming languages—such as “C” or similar programming languages.
  • the program code may execute entirely on the user computing device, partly on the user device, as a stand-alone software package, partly on the user computing device and partly on a remote computing device, or entirely on the remote computing device or server.
  • the remote computing device may be connected to the user computing device through any kind of network, including a local area network (LAN) or a wide area network (WAN), or alternatively, may be connected to an external computing device (e.g., via the Internet using an Internet service provider).
  • LAN local area network
  • WAN wide area network
  • Internet service provider e.g., via the Internet using an Internet service provider
  • the above-mentioned computer-readable medium carries one or more programs, and when the above-mentioned one or more programs are executed by one of the devices, the computer-readable medium realizes the following functions: obtain a sample to be tested, and the sample to be tested contains multiple target analytes; perform a biochemical reaction treatment based on magnetic particles on the sample to be tested to generate a reaction solution, and the reaction solution contains multiple multiple complexes, wherein the multiple multiple complexes correspond to multiple target analytes; Excite the reaction solution so that the multiple multiple complexes emit multiple lights; The sample is separated from the sample and collected on the surface of the magnetic particle; the plurality of lights are acquired based on a single optical detection channel to generate a sample detection image; the detection result of each target analyte in the plurality of target analytes is generated based on the sample detection image; the detection result of the sample to be tested is generated based on the detection result of each target analyte.
  • the computer-readable medium can also realize the following functions: generating a plurality of sample training pictures through a plurality of detection samples; labeling a plurality of training labels for the plurality of sample training pictures, wherein the plurality of training labels correspond to the various target analytes; training a machine learning model based on a plurality of sample training pictures with a plurality of training labels to generate the multi-target recognition model.
  • modules in the above embodiments can be distributed in the device according to the description of the embodiment, and corresponding changes can also be made in one or more devices that are only different from the embodiment.
  • the modules in the above embodiments can be combined into one module, and can also be further split into multiple sub-modules.
  • the technical solution according to the embodiment of the present application can be embodied in the form of a software product, and the software product can be stored in a non-volatile storage medium (which can be a CD-ROM, a U disk, a mobile hard disk, etc.) or on a network, and include several instructions so that a computing device (which can be a personal computer, a server, a mobile terminal, or a network device, etc.) executes the method according to the embodiment of the application.
  • a non-volatile storage medium which can be a CD-ROM, a U disk, a mobile hard disk, etc.
  • a computing device which can be a personal computer, a server, a mobile terminal, or a network device, etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • General Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Biophysics (AREA)
  • Evolutionary Biology (AREA)
  • Chemical & Material Sciences (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • Multimedia (AREA)
  • Software Systems (AREA)
  • Genetics & Genomics (AREA)
  • Molecular Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Biotechnology (AREA)
  • Evolutionary Computation (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Databases & Information Systems (AREA)
  • Quality & Reliability (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Radiology & Medical Imaging (AREA)
  • Investigating Or Analyzing Materials By The Use Of Magnetic Means (AREA)
  • Investigating, Analyzing Materials By Fluorescence Or Luminescence (AREA)

Abstract

本申请涉及一种样本检测方法、装置、系统、电子设备及计算机可读介质。本申请的样本检测方法包括：获取待测样本；对待测样本进行基于磁性粒子的生化反应处理以生成反应溶液；对反应溶液进行激发以使多种多元复合物发出多种光；在外界磁场作用下将目标分析物从待测样本中分离出来以及收集在磁性粒子的表面；基于单一光学检测通道获取多种光以生成样本检测图片；基于样本检测图片生成多种目标分析物中每一种目标分析物的检测结果；基于每一种目标分析物的检测结果生成待测样本的检测结果。本申请通过单通道的光学系统实现多路复用检测，而且在不需要特殊耗材的前提下，同时对多目标分析物进行定性或定量的精准识别。

Description

样本检测方法、装置、系统、电子设备及计算机可读介质 技术领域
本申请涉及体外检测领域，具体而言，涉及一种基于单一光学检测通道的样本检测方法、装置、系统、电子设备及计算机可读介质。
背景技术
体外诊断，即IVD(In Vitro Diagnosis)，是指在人体之外，通过对人体样本(血液、体液、组织等)进行检测而获取临床诊断信息，进而判断疾病或机体功能的产品和服务。
目前的体外诊断里的免疫诊断技术和分子诊断技术，可分为：化学发光免疫检测CLIA(包括直接化学发光分析，化学发光酶免疫分析(CLEIA)，电化学发光免疫分析ECLIA/ECL)；单分子免疫测定SIMOA；酶联免疫吸附测定ELISA；多重因子超灵敏电化学发光免疫分析Meso Scale Discovery(MSD)；荧光免疫测定FIA；放射免疫测定RIA；胶体金Colloidal gold；聚合酶链式反应PCR；环状介导等温扩增LAMP；其他等温核酸扩增测定；其他分子诊断技术(基因测序)等。
其中，通过光学手段进行定性或定量的检测的模式大概分如下几种：1.基于显微镜的肉眼观察颜色或荧光的强度；2.光学传感器(比如光子计数器等)测量颜色吸收、发光强度或荧光强度；3.相机直接或通过显微镜对于颜色变化或荧光进行检测和识别，一般会辅助使用图像处理技术。其中需要多通道的光学系统，增加了光学设备的光路设计复杂性以及仪器造价成本，每个通道有不同的滤光透镜和分束镜组合，有的技术还需要特殊耗材。复杂的检测手段限制了体外诊断技术的发展。因此，需要一种新的基于单一光学检测通道的样本检测方法、装置、系统、电子设备及计算机可读介质。
在所述背景技术部分公开的上述信息仅用于加强对本申请的背景的 理解，因此它可以包括不构成对本领域普通技术人员已知的现有技术的信息。
发明内容
有鉴于此，本申请提供一种基于单一光学检测通道的样本检测方法、装置、系统、电子设备及计算机可读介质，能够通过单通道的光学系统实现多路复用检测，而且在不需要特殊耗材的前提下，同时对多目标分析物进行定性或定量的精准识别。
本申请的其他特性和优点将通过下面的详细描述变得显然，或部分地通过本申请的实践而习得。
根据本申请的一方面，提出一种基于单一光学检测通道的样本检测方法，该方法包括：获取待测样本，所述待测样本中包含多种目标分析物；对所述待测样本进行基于磁性粒子的生化反应处理以生成反应溶液，所述反应溶液中包含多种多元复合物，其中，多种多元复合物与多种目标分析物对应；对所述反应溶液进行激发以使所述多种多元复合物发出多种光；在外界磁场作用下将目标分析物从待测样本中分离出来以及收集在磁性粒子的表面；基于单一光学检测通道获取所述多种光以生成样本检测图片；基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果；基于每一种目标分析物的检测结果生成所述待测样本的检测结果。
在本申请的一种示例性实施例中，对所述待测样本进行基于磁性粒子的生化反应处理以生成反应溶液，包括：基于生物载体技术对所述待测样本进行基于磁性粒子的非酶生化反应处理以生成所述反应溶液。
在本申请的一种示例性实施例中，基于生物载体技术对所述待测样本进行基于磁性粒子的非酶生化反应处理以生成所述反应溶液，包括：对所述待测样本进行非酶生化反应处理以使得待测样本中的多种目标分析物分别和多种生物载体结合以生成所述反应溶液；利用磁性粒子，在外界磁场作用下对目标分析物进行分离。
在本申请的一种示例性实施例中，对所述待测样本进行非酶生化反应处理以使得待测样本中的多种目标分析物分别和多种生物载体结合以生成所述反应溶液，包括：对所述待测样本进行非酶生化反应处理以使得待测样本中的多种目标分析物分别和与其对应的磁性粒子、抗体、待测物、荧光纳米珠信号探针结合以生成所述反应溶液。
在本申请的一种示例性实施例中，对所述反应溶液进行激发以使所述多种多元复合物发出多种光，包括：通过光源设备或电学设备或化学试剂对所述反应溶液进行激发以使所述多种多元复合物发出多种光，所述多种光包括具有不同的强度和/或频率的光。
在本申请的一种示例性实施例中，对所述反应溶液进行激发以使所述多种多元复合物发出多种光，包括：通过光源对所述反应溶液进行激发以使所述多种多元复合物发出多种光。
在本申请的一种示例性实施例中，通过光源对所述反应溶液进行激发以使所述多种多元复合物发出多种光，包括：将所述反应溶液放置于透明平面；待反应溶液中的磁性粒子沉淀在平面的表面上，通过蓝色光源对置于透明平面的所述反应溶液进行激发以使所述多种多元复合物发出多种光。
在本申请的一种示例性实施例中，基于单一光学检测通道获取所述多种光以生成样本检测图片，包括：基于单一光学检测通道中的滤光透镜获取所述多种光；光学传感器基于所述多种光生成所述样本检测图片。
在本申请的一种示例性实施例中，光学传感器基于所述多种光生成所述样本检测图片，包括：CMOS光学传感器基于所述多种光生成所述样本检测图片；或CCD光学传感器基于所述多种光生成所述样本检测图片。
在本申请的一种示例性实施例中，基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果，包括：对所述样本检测图片进行预处理；生成所述样本检测图片的特征矩阵；将所述特征矩阵输入多目标识别模型中，生成所述多种目标分析物中每一种目标分析物的检测 结果。
在本申请的一种示例性实施例中，对所述样本检测图片进行预处理，包括：对所述样本检测图片进行滤波处理；和/或对所述样本检测图片进行斑点检测处理。
在本申请的一种示例性实施例中，生成所述样本检测图片的特征矩阵，包括：提取所述样本检测图片中每一个像素的特征数据；基于所有像素的特征数据生成所述特征矩阵。
在本申请的一种示例性实施例中，提取所述样本检测图片中每一个像素的特征数据，包括：提取所述样本检测图片中每一个像素的颜色和位置。
在本申请的一种示例性实施例中，基于所有像素的特征数据生成所述特征矩阵，还包括：对所述特征矩阵进行特征变换处理。
在本申请的一种示例性实施例中，将所述特征矩阵输入多目标识别模型中，生成所述多种目标分析物中每一种目标分析物的检测结果，包括：将所述特征矩阵输入多目标识别模型中；所述多目标识别模型基于所述特征矩阵对所述样本检测图片中的像素逐一进行分类预测；基于分类预测结果生成所述多种目标分析物中每一种目标分析物的检测结果。
在本申请的一种示例性实施例中，生成所述多种目标分析物中每一种目标分析物的检测结果，包括：生成所述多种目标分析物中每一种目标分析物对应的像素总数和总亮度；基于像素总数和总亮度生成所述多种目标分析物中每一种目标分析物的检测结果。
在本申请的一种示例性实施例中，基于像素总数和总亮度生成所述多种目标分析物中每一种目标分析物的检测结果，包括：基于校准公式和每一种目标分析物对应的像素总数和总亮度生成每一种目标分析物对应的数量和浓度；基于每一种目标分析物对应的数量和浓度生成所述待测样本的检测结果。
在本申请的一种示例性实施例中，还包括：通过多个检测样本生成多 个样本训练图片；为所述多个样本训练图片分别标注多个训练标签，其中，所述多个训练标签与所述多种目标分析物对应；基于带有多个训练标签的多个样本训练图片对机器学习模型进行训练以生成所述多目标识别模型。
在本申请的一种示例性实施例中，为所述多个样本训练图片分别标注多个训练标签，包括：为所述多个样本训练图片中的每个像素分别标注训练标签。
在本申请的一种示例性实施例中，基于带有多个训练标签的多个样本训练图片对机器学习模型进行训练以生成所述多目标识别模型，包括：基于带有多个训练标签的多个样本训练图片对多层感知器进行训练以生成所述多目标识别模型；或基于带有多个训练标签的多个样本训练图片对随机森林分类器进行训练以生成所述多目标识别模型。
根据本申请的一方面，提出一种基于单一光学检测通道的样本检测装置，该装置包括：样本模块，用于获取待测样本，所述待测样本中包含多种目标分析物；反应模块，用于对所述待测样本进行基于磁性粒子的生化反应处理以生成反应溶液，所述反应溶液中包含多种多元复合物，其中，多种多元复合物和多种目标分析物对应；激发分离模块，用于对所述反应溶液进行激发以使所述多种多元复合物发出多种光；在外界磁场作用下将目标分析物从待测样本中分离出来以及收集在磁性粒子的表面；图片模块，用于基于单一光学检测通道获取所述多种光以生成样本检测图片；检测模块，用于基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果；结果模块，用于基于每一种目标分析物的检测结果生成所述待测样本的检测结果。
在本申请的一种示例性实施例中，所述检测模块，包括：处理单元，用于对所述样本检测图片进行预处理；特征单元，用于生成所述样本检测图片的特征矩阵；识别单元，用于将所述特征矩阵输入多目标识别模型中，生成所述多种目标分析物中每一种目标分析物的检测结果。
在本申请的一种示例性实施例中，所述检测模块，还包括：模型训练单元，用于通过多个检测样本生成多个样本训练图片；为多个样本训练图 片分别标注多个训练标签，其中，所述多个训练标签与所述多种目标分析物对应；基于带有多个训练标签的多个样本训练图片对机器学习模型进行训练以生成所述多目标识别模型。
根据本申请的一方面，提出一种基于单一光学检测通道的样本检测系统，该系统包括：反应装置，用于获取待测样本，所述待测样本中包含多种目标分析物；对所述待测样本进行基于磁性粒子的生化反应处理以生成反应溶液，所述反应溶液中包含多种多元复合物，其中，多种多元复合物与多种目标分析物对应；激发分离装置，用于对所述反应溶液进行激发以使所述多种多元复合物发出多种光；在外界磁场作用下将目标分析物从待测样本中分离出来以及收集在磁性粒子的表面；图像装置，用于基于单一光学检测通道获取所述多种光以生成样本检测图片；服务器，用于基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果；基于每一种目标分析物的检测结果生成所述待测样本的检测结果。
根据本申请的一方面，提出一种电子设备，该电子设备包括：一个或多个处理器；存储装置，用于存储一个或多个程序；当一个或多个程序被一个或多个处理器执行，使得一个或多个处理器实现如上文的方法。
根据本申请的一方面，提出一种计算机可读介质，其上存储有计算机程序，该程序被处理器执行时实现如上文中的方法。
根据本申请的基于单一光学检测通道的样本检测方法、装置、系统、电子设备及计算机可读介质，通过获取待测样本，所述待测样本中包含多种目标分析物；对所述待测样本进行基于磁性粒子的生化反应处理以生成反应溶液，所述反应溶液中包含多种多元复合物，其中，多种多元复合物与多种目标分析物对应；对所述反应溶液进行激发以使所述多种多元复合物发出多种光；基于单一光学检测通道获取所述多种光以生成样本检测图片；基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果；基于每一种目标分析物的检测结果生成所述待测样本的检测结果的方式，能够通过单通道的光学系统实现多路复用检测，而且在不需要特殊耗材的前提下，同时对多目标分析物进行定性或定量的精准识别。
应当理解的是，以上的一般描述和后文的细节描述仅是示例性的，并不能限制本申请。
附图说明
通过参照附图详细描述其示例实施例，本申请的上述和其它目标、特征及优点将变得更加显而易见。下面描述的附图仅仅是本申请的一些实施例，对于本领域的普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1是根据一示例性实施例示出的一种基于单一光学检测通道的样本检测系统的示意图。
图2是根据一示例性实施例示出的一种基于单一光学检测通道的样本检测方法的流程图。
图3是根据另一示例性实施例示出的一种基于单一光学检测通道的样本检测方法的示意图。
图4是根据另一示例性实施例示出的一种基于单一光学检测通道的样本检测方法的示意图。
图5是根据一示例性实施例示出的一种基于单一光学检测通道的样本检测方法的流程图。
图6是根据一示例性实施例示出的一种基于单一光学检测通道的样本检测方法的流程图。
图7是根据一示例性实施例示出的一种基于单一光学检测通道的样本检测方法的流程图。
图8是根据另一示例性实施例示出的一种基于单一光学检测通道的样本检测装置的框图。
图9是根据一示例性实施例示出的一种电子设备的框图。
图10是根据一示例性实施例示出的一种计算机可读介质的框图。
具体实施方式
现在将参考附图更全面地描述示例实施例。然而，示例实施例能够以多种形式实施，且不应被理解为限于在此阐述的实施例；相反，提供这些 实施例使得本申请将全面和完整，并将示例实施例的构思全面地传达给本领域的技术人员。在图中相同的附图标记表示相同或类似的部分，因而将省略对它们的重复描述。
此外，所描述的特征、结构或特性可以以任何合适的方式结合在一个或更多实施例中。在下面的描述中，提供许多具体细节从而给出对本申请的实施例的充分理解。然而，本领域技术人员将意识到，可以实践本申请的技术方案而没有特定细节中的一个或更多，或者可以采用其它的方法、组元、装置、步骤等。在其它情况下，不详细示出或描述公知方法、装置、实现或者操作以避免模糊本申请的各方面。
附图中所示的方框图仅仅是功能实体，不一定必须与物理上独立的实体相对应。即，可以采用软件形式来实现这些功能实体，或在一个或多个硬件模块或集成电路中实现这些功能实体，或在不同网络和/或处理器装置和/或微控制器装置中实现这些功能实体。
附图中所示的流程图仅是示例性说明，不是必须包括所有的内容和操作/步骤，也不是必须按所描述的顺序执行。例如，有的操作/步骤还可以分解，而有的操作/步骤可以合并或部分合并，因此实际执行的顺序有可能根据实际情况改变。
应理解，虽然本文中可能使用术语第一、第二、第三等来描述各种组件，但这些组件不应受这些术语限制。这些术语乃用以区分一组件与另一组件。因此，下文论述的第一组件可称为第二组件而不偏离本申请概念的教示。如本文中所使用，术语“及/或”包括相关联的列出项目中的任一个及一或多者的所有组合。
本领域技术人员可以理解，附图只是示例实施例的示意图，附图中的模块或流程并不一定是实施本申请所必须的，因此不能用于限制本申请的保护范围。
图1是根据一示例性实施例示出的一种基于单一光学检测通道的样本检测系统的示意图。
如图1所示，基于单一光学检测通道的样本检测系统10可以包括反应装置101、激发分离装置102、图像装置103，网络104和服务器105。 网络104用以在图像装置103和服务器105之间提供通信链路的介质。网络104可以包括各种连接类型，例如有线、无线通信链路或者光纤电缆等等。
反应装置101用于获取待测样本，所述待测样本中包含多种目标分析物；对所述待测样本进行基于磁性粒子的生化反应处理以生成反应溶液，所述反应溶液中包含多种多元复合物，其中，多种多元复合物与多种目标分析物对应；
激发分离装置102用于对所述反应溶液进行激发以使所述多种多元复合物发出多种光；在外界磁场作用下将目标分析物从待测样本中分离出来以及收集在磁性粒子的表面；
图像装置103用于基于单一光学检测通道获取所述多种光以生成样本检测图片；
服务器105用于基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果；基于每一种目标分析物的检测结果生成所述待测样本的检测结果。
更进一步的，服务器105可以是一个实体的服务器，还可例如为多个服务器组成，服务器105中的一部分可例如作为本申请中的样本检测装置，用于基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果；基于每一种目标分析物的检测结果生成所述待测样本的检测结果；以及服务器105中的一部分还可例如作为本申请中的模型训练装置，用于通过多个检测样本生成多个样本训练图片；为所述多个样本训练图片分别标注多个训练标签，其中，所述多个训练标签与所述多种目标分析物对应；基于带有多个训练标签的多个样本训练图片对机器学习模型进行训练以生成所述多目标识别模型。
值得一提的是，在本申请中的方法的一个具体的实施例中，激发方式可为荧光光源激发，在这种情况下，激发分离装置102和图像装置103的功能可通过现有技术中的单一通道的光学检测设备实现，在这种情况下，本申请中的基于单一光学检测通道的样本检测系统可包括：反应装置、光 学检测设备，网络和服务器。
图2是根据一示例性实施例示出的一种基于单一光学检测通道的样本检测方法的流程图。基于单一光学检测通道的样本检测方法20至少包括步骤S202至S212。
如图2所示，在S202中，获取待测样本，所述待测样本中包含多种目标分析物。
在本申请的一个具体的实施例中，待测样本可为某病人的血清样本，样本中可能多种目标分析物，目标分析物可包括：HIV-1病毒产生的p24蛋白、SARS-COV-2病毒产生的N蛋白等。后文中将会以本实施例进行说明，可理解的是，本申请中的方法还可应用于其他的待测样本和目标分析物的检测中。
在S204中，对所述待测样本进行基于磁性粒子的生化反应处理以生成反应溶液，所述反应溶液中包含多种多元复合物，其中，多种多元复合物与多种目标分析物对应。可基于生物载体技术对所述待测样本进行生化反应处理以生成所述反应溶液。其中，生化反应可为非酶生化反应。其中，磁性粒子可为现有技术中的磁珠。基于磁性粒子可以对目标物进行快速和精准地捕捉和分离。在现有技术传统方式中，生化反应的时间大致需要多个小时，同时对低浓度的目标物的捕捉效率也会非常低。在本申请中基于磁性粒子的生化反应提高了反应速度，在实际应用中缩短了检测时间。
在一个实施例中，可基于生物载体技术对所述待测样本进行基于磁性粒子的非酶生化反应处理以生成所述反应溶液。更具体地，可例如，对所述待测样本进行非酶生化反应处理以使得待测样本中的多种目标分析物分别和多种生物载体结合以生成所述反应溶液；利用磁性粒子，在外界磁场作用下对目标分析物进行分离。
在一个实施例中，可对所述待测样本进行非酶生化反应处理以使得待测样本中的多种目标分析物分别和与其对应的磁性粒子、抗体、待测物、荧光纳米珠信号探针结合以生成所述反应溶液。
更具体的，可对血清样本进行免疫生化反应处理，具体的处理过程可参考图3中的描述。使用磁性粒子，捕获抗体，检测抗体和萤光粒子，对对待测样本中的多种目标物质进行捕获和标记。更具体的，可对两种抗原蛋白进行捕获和标记，形成多路复用的多元复合物结构。
在如上文所述的待测样本中，生成的多元复合物的结构分别是：
磁性粒子–p24捕获抗体–p24蛋白–p24检测抗体–绿色荧光纳米珠信号探针；
磁性粒子–N蛋白捕获抗体–N蛋白–N蛋白检测抗体–红色荧光纳米珠信号探针。
值得一提的是，磁性粒子的表面可能捕捉多个相应蛋白及纳米级荧光纳米珠信号探针，也可能完全捕捉不到，在未捕获到蛋白及纳米级荧光纳米珠信号探针时，磁性粒子保持原有的状态，所以，免疫生化反应之后，反应溶液中还可能包含磁性粒子。
在S206中，在外界磁场作用下将目标分析物从待测样本中分离出来以及收集在磁性粒子的表面。
在S208中，对所述反应溶液进行激发以使所述多种多元复合物发出多种光。可通过光源设备或电学设备或化学试剂对所述反应溶液进行激发以使所述多种多元复合物发出多种光。更具体的，可通过卤素灯，LED等光源对反应溶液进行激发。还可通过汞蒸气灯，LED，激光等光源进行激发。
在一个实施例中，可通过荧光光源对所述反应溶液进行激发以使所述多种多元复合物发出多种荧光。
更具体的，可将所述反应溶液放置于透明平面；待反应溶液中的磁性粒子沉淀在平面的表面上，通过蓝色光源对置于透明平面的所述反应溶液进行激发以使所述多种多元复合物发出多种光。
在一个具体的应用场景中，可将反应后的溶液置于荧光噪声很低的透 明平面，在一个实施例中，透明平面的荧光噪声参数可为材料自发荧光的荧光发射强度。单位是：Arbitrary Unit(取决于检测仪器)。可考虑的材料：玻璃，聚甲基丙烯酸甲酯(PMMA/polymethylmethacrylate)，环烯烃聚合物(COP/Cyclo Olefin Polymer)等。
物质的荧光强度与激发光的波长和所测量发射光的波长有关，将物质的荧光强度的数据用矩阵形式表示，行和列对应不同的激发光波长和发射光波长，每个矩阵元分别为该激发光、发射光波长的荧光强度，称之为激发—发射矩阵(Excitation Emission Matrix，EEM)。
在本申请的一个实施例中，如图4所述，当目标分析物被某一个波长的光激发之后，目标分析物发出的光不是单一波长的光，而是跨一个波段的光，而且在整个波段中的强度分布完全由被激发物质自身决定。配合使用与蓝色荧光匹配的滤波片(emission filter)在远离绿光的波段中的并没有100％的滤掉其他的颜色光，所以通过蓝色的荧光激发后，在图像中可以显示出多种目标分析物对应的不同的非绿色的荧光。
在S210中，基于单一光学检测通道获取所述多种光以生成样本检测图片。可基于单一光学检测通道中的滤光透镜获取所述多种光；光学传感器基于所述多种光生成所述样本检测图片。
在一个实施例中，可通过CMOS光学传感器基于所述多种光生成所述样本检测图片；还可通过CCD光学传感器基于所述多种光生成所述样本检测图片。
可通过不同的方式进行激发，可通过白光源的激发，此时，所有的磁性粒子都在明场条件下被CCD相机传感器捕捉。
在一个具体的应用场景中，可通过蓝色光源的激发及相应滤光透镜的过滤，反应溶液中的各种粒子会被激发出不同颜色的光，不同颜色粒子都可以被CCD传感器捕捉，其中磁性粒子显示黄色，绿色荧光纳米珠信号探针显示绿色，红色荧光纳米珠信号探针显示橙色，背景显示黑色。
本申请中的这种采用蓝光激发的方式，能够得到多种目标分析物的不 同的光，且每种物质的光波段和强度组合不同，有利于识别出不同的目标分析物。
在S212中，基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果，可对所述样本检测图片进行预处理；生成所述样本检测图片的特征矩阵；将所述特征矩阵输入多目标识别模型中，生成所述多种目标分析物中每一种目标分析物的检测结果。
“基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果”的详细内容将在图5对应的实施例中进行详细描述。
在S214中，基于每一种目标分析物的检测结果生成所述待测样本的检测结果。可根据固态载体-磁性粒子和免疫测定标记物-萤光粒子的计数，通过校准公式的转换，间接获得p24蛋白和N蛋白的数量和浓度，得到病人的健康情况的定性或者定量的判定。
更具体的，校准公式可由实验中的拟合曲线得到。可例如，在实际应用之前，获取多次临床前实验数据，基于临床试验数据推算萤光粒子个数VS控制好数量的蛋白浓度的曲线，进而生成校准公式。
根据本申请的基于单一光学检测通道的样本检测方法，通过获取待测样本，所述待测样本中包含多种目标分析物；对所述待测样本进行基于磁性粒子的生化反应处理以生成反应溶液，所述反应溶液中包含多种多元复合物，其中，多种多元复合物与多种目标分析物对应；对所述反应溶液进行激发以使所述多种多元复合物发出多种光；在外界磁场作用下将目标分析物从待测样本中分离出来以及收集在磁性粒子的表面；基于单一光学检测通道获取所述多种光以生成样本检测图片；基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果；基于每一种目标分析物的检测结果生成所述待测样本的检测结果的方式，能够通过单通道的光学系统实现多路复用检测，而且在不需要特殊耗材的前提下，同时对多目标分析物进行定性或定量的精准识别。
本申请的基于单一光学检测通道的样本检测方法，实现了单一的光学检测通道的多路复用检测，一张成像图片中可以涵盖多种发光物质的信息，继而间接计算出相应多种目标分析物的数量及其在样本种的浓度。
在本申请的基于单一光学检测通道的样本检测方法中，在图像预处理后，扫描每个预处理后剩下的有效像素，将有效像素的图片信息转化为数据信息，机器学习处理的是数据信息，加快了机器学习模型的处理速度和准确度。
本申请的基于单一光学检测通道的样本检测方法，无需特殊耗材(比如微流体试剂盒)及相应流体控制的前提下，实现发光物质的逐个识别和计数。
应清楚地理解，本申请描述了如何形成和使用特定示例，但本申请的原理不限于这些示例的任何细节。相反，基于本申请公开的内容的教导，这些原理能够应用于许多其它实施例。
图5是根据另一示例性实施例示出的一种基于单一光学检测通道的样本检测方法的流程图。图5所示的流程50是对图2所示的流程中S210“基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果”的详细描述。
如图5所示，在S502中，对所述样本检测图片进行预处理。可对所述样本检测图片进行滤波处理；还可对所述样本检测图片进行斑点检测处理。
在S504中，生成所述样本检测图片的特征矩阵。可提取所述样本检测图片中每一个像素的特征数据；基于所有像素的特征数据生成所述特征矩阵。
在一个实施例中，可提取所述样本检测图片中每一个像素的颜色和位置作为特征数据。
在一个实施例中，还可对所述特征矩阵进行特征变换处理，在特征变换处理之后，基于所有像素的颜色和位置生成所述特征矩阵。更具体的，可对特征矩阵进行数据缩放，标签编码等特征变换处理。还可对特征矩阵进行像素压缩处理，可将亮度低于某个阈值的像素点归类为背景颜色，从而不对该像素点进行后续的图像分析，以节约计算时间和计算空间。
在S506中，将所述特征矩阵输入多目标识别模型中，生成所述多种目标分析物中每一种目标分析物的检测结果。可将所述特征矩阵输入多目标识别模型中；所述多目标识别模型基于所述特征矩阵对所述样本检测图片中的像素逐一进行分类预测；基于分类预测结果生成所述多种目标分析物中每一种目标分析物的检测结果。
在一个实施例中，可生成所述多种目标分析物中每一种目标分析物对应的像素总数和总亮度；基于像素总数和总亮度生成所述多种目标分析物中每一种目标分析物的检测结果。
更具体的，可基于校准公式和每一种目标分析物对应的像素总数和总亮度生成每一种目标分析物对应的数量和浓度；基于每一种目标分析物对应的数量和浓度生成所述待测样本的检测结果。
图6是根据另一示例性实施例示出的一种基于单一光学检测通道的样本检测方法的流程图。图6所示的流程60是对图5所示的流程的补充描述。
如图6所示，在S602中，通过多个检测样本生成多个样本训练图片。对应于上文中的具体的实施例，针对于不同的样本需要训练不同的机器学习模型。可获取多个病人的血清样本作为训练样本，进而根据本申请中的方式进行生化反应和激发生成多个样本训练图片。
还可针对病人的体液、尿液等分别生成样本训练图片以进行后续的机器学习模型的训练。
在S604中，为所述多个样本训练图片分别标注多个训练标签，其中，所述多个训练标签与所述多种目标分析物对应。更具体的，可为所述多个样本训练图片中的每个像素分别标注训练标签。
为样本训练图片中的每个像素分别标注训练标签，训练标签可为目标分析物1，目标分析物2，目标分析物3等等。
可理解的是，标注的目标分析物的种类可多于实际分析中的检测出的目标分析物，如本申请上文中的具体实施例，检测出的目标分析物为：HIV- 1病毒产生的p24蛋白、SARS-COV-2病毒产生的N蛋白，但对应于本次的样本检测，所使用的机器学习模型中还检出其他病毒对应的X蛋白等。机器学习模型能够检出的目标分析物的种类取决于样本中目标分析物的种类。
在S606中，基于带有多个训练标签的多个样本训练图片对机器学习模型进行训练以生成所述多目标识别模型。可基于带有多个训练标签的多个样本训练图片对多层感知器进行训练以生成所述多目标识别模型；还可基于带有多个训练标签的多个样本训练图片对随机森林分类器进行训练以生成所述多目标识别模型。
可将样本训练图像中的60％的像素点对应的数据为训练数据组，对机器学习模型进行训练。还可通过十折交叉验证优化模型超参数，对于多层感知器，超参数包括隐藏层的神经元数量，激活函数类型等等；对于随机森林分类器，超参数包括决策树的数量，最大深度类型等等
随着超参数的优化，多层感知器和随机森林分类器的最优参数及最优的特征转化的方式也被选定。
具体的，针对每个样本训练图像，分别构建初始模型，将样本训练图像中的像素特征矩阵输入初始模型中，以计算得到预测标签，将预测标签与相应的真实的标签进行比对，判断预测标签与真实的标签是否一致，统计与真实的标签一致的预测标签的数量，并计算与真实的标签一致的预测标签的数量在所有预测标签的数量中的占比，若所述占比大于或等于预设占比值，则初始模型收敛，得到训练完成的多目标识别模型。
若所述占比小于所述预设占比值，则调整初始模型中的参数，通过调整后的初始模型重新预测各个对象的预测标签，直至所述占比大于或等于预设占比值。其中，调整初始模型中的参数的方法可以采用随机梯度下降算法、梯度下降算法或正规方程进行。若调整初始模型的参数的次数超过预设次数时，可以更换构建调整模型所使用的模型，以提高模型训练效率。
在模型训练结束后，还可取40％像素点对应的数据作为验证数据组，并使用混淆矩阵机器相关度量参数，进行模型的最终评估。如果最终评估失败，可以采集更多高质量的病人样本图片重新训练模型，也可以采取不 同的特征工程方案进行模型调整。
在一个实施例中，在模型训练完成后，即可部署模型对病人的样本进行预测。可在模型部署后实施后，还可对模型的计算效果进行检测以确认模型持续有效，检测方式可为：比较对于图像的识别和预先设置的参数阈值，将诊断结果与现实中其它可靠的参考结果进行比较。
如果数据不匹配，将校准实际中和训练中使用的光学系统。如果训练中使用的光学系统有问题，将在修复后重新训练模型。如果是数据异常，代表训练样本没有涵盖所有的病人状况。将在训练数据中增加异常的数据重新训练模型。
图7是根据另一示例性实施例示出的一种基于单一光学检测通道的样本检测方法的流程图。图7所示的流程70本申请中的方法在一个实际的应用场景中的应用描述。
如图7所示，在S702中，获取待测样本，可例如为病人的血样、尿样、唾液或其他体液。
在S704中，生化反应后生成反应溶液。对血清样本进行免疫生化反应处理，使用磁性粒子，捕获抗体，检测抗体和荧光纳米珠信号探针。
在S706中，传感器获取数字化图像或者视频。将反应后的溶液置于光噪声很低的透明平面。通过蓝色光源的激发及相应滤光透镜的过滤。各种微粒会被激发出不同的光颜色，不同粒子都可以被CCD传感器捕捉，其中磁性粒子显示黄色，绿色光粒子显示绿色，红色光粒子现实橙色，背景显示黑色。
在S708中，输入机器学习模型中进行计算。对相机拍摄的样本图片进行预处理，包括滤波和斑点检测处理等；对处理后的图片逐像素进行特征提取，得到相应的特征矩阵；使用训练好的机器学习模型对特征矩阵逐像素进行分类预测，得到每种目标粒子的像素总数和总亮度。
在S710中，生成诊断结果。根据固态载体-磁性粒子和免疫测定标记物-萤光粒子的计数，通过校准公式的转换，间接获得p24蛋白和N蛋白的数量和浓度，得到病人的健康情况的定性或者定量的判定。
在S712中，根据诊断结果推荐治疗方案。可事先在系统中设置不同的诊断结果对应的初步处理手段，在得到本次检测的病人的诊断结果后，自动调取治疗方案，辅助医生现场处理，提高速度。
本申请的基于单一光学检测通道的样本检测方法，实现了单一的光学检测通道的多路复用检测，一张成像图片中可以涵盖多种发光物质的信息，继而间接计算出相应多种目标分析物的数量及其在样本种的浓度。
传统图像处理技术的预处理辅助机器学习的最终识别，而在本申请的基于单一光学检测通道的样本检测方法中，在预处理后,扫描每个预处理后剩下的有效像素，将有效像素的图片信息转化为数据信息，机器学习处理的是数据信息，加快了机器学习模型的处理速度和准确度。
本申请的基于单一光学检测通道的样本检测方法，无需特殊耗材(比如微流体试剂盒)及相应流体控制的前提下，实现发光物质的逐个识别和计数。
本领域技术人员可以理解实现上述实施例的全部或部分步骤被实现为由CPU执行的计算机程序。在该计算机程序被CPU执行时，执行本申请提供的上述方法所限定的上述功能。所述的程序可以存储于一种计算机可读存储介质中，该存储介质可以是只读存储器，磁盘或光盘等。
此外，需要注意的是，上述附图仅是根据本申请示例性实施例的方法所包括的处理的示意性说明，而不是限制目的。易于理解，上述附图所示的处理并不表明或限制这些处理的时间顺序。另外，也易于理解，这些处理可以是例如在多个模块中同步或异步执行的。
下述为本申请装置实施例，可以用于执行本申请方法实施例。对于本申请装置实施例中未披露的细节，请参照本申请方法实施例。
图8是根据一示例性实施例示出的一种基于单一光学检测通道的样本检测装置的框图。如图8所示，基于单一光学检测通道的样本检测装置80包括：样本模块802，反应模块804，激发分离模块806，图片模块808，检测模块810，结果模块812。更具体的，检测模块810可包括：处理单元8102，特征单元8104，识别单元8106，模型训练单元8108。
样本模块802用于获取待测样本，所述待测样本中包含多种目标分析物；
反应模块804用于对所述待测样本进行基于磁性粒子的生化反应处理以生成反应溶液，所述反应溶液中包含多种多元复合物，其中，多种多元复合物和多种目标分析物对应；
激发分离模块806用于对所述反应溶液进行激发以使所述多种多元复合物发出多种光；在外界磁场作用下将目标分析物从待测样本中分离出来以及收集在磁性粒子的表面；
图片模块808用于基于单一光学检测通道获取所述多种光以生成样本检测图片；
检测模块810用于基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果；
结果模块812用于基于每一种目标分析物的检测结果生成所述待测样本的检测结果。
其中，所述检测模块810包括：处理单元8102用于对所述样本检测图片进行预处理；特征单元8104用于生成所述样本检测图片的特征矩阵；识别单元8106用于将所述特征矩阵输入多目标识别模型中，生成所述多种目标分析物中每一种目标分析物的检测结果。
所述检测模块810还包括：模型训练单元8108用于通过多个检测样本生成多个样本训练图片；为多个样本训练图片分别标注多个训练标签，其中，所述多个训练标签与所述多种目标分析物对应；基于带有多个训练标签的多个样本训练图片对机器学习模型进行训练以生成所述多目标识别模型。
根据本申请的基于单一光学检测通道的样本检测装置，通过获取待测样本，所述待测样本中包含多种目标分析物；对所述待测样本进行基于磁性粒子的生化反应处理以生成反应溶液，所述反应溶液中包含多种多元复合物，其中，多种多元复合物与多种目标分析物对应；对所述反应 溶液进行激发以使所述多种多元复合物发出多种光；在外界磁场作用下将目标分析物从待测样本中分离出来以及收集在磁性粒子的表面；基于单一光学检测通道获取所述多种光以生成样本检测图片；基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果；基于每一种目标分析物的检测结果生成所述待测样本的检测结果的方式，能够通过单通道的光学系统实现多路复用检测，而且在不需要特殊耗材的前提下，同时对多目标分析物进行定性或定量的精准识别。
图9是根据一示例性实施例示出的一种电子设备的框图。
下面参照图9来描述根据本申请的这种实施方式的电子设备900。图9显示的电子设备900仅仅是一个示例，不应对本申请实施例的功能和使用范围带来任何限制。
如图9所示，电子设备900以通用计算设备的形式表现。电子设备900的组件可以包括但不限于：至少一个处理单元910、至少一个存储单元920、连接不同系统组件(包括存储单元920和处理单元910)的总线930、显示单元940等。
其中，所述存储单元存储有程序代码，所述程序代码可以被所述处理单元910执行，使得所述处理单元910执行本说明书中描述的根据本申请各种示例性实施方式的步骤。例如，所述处理单元910可以执行如图2，图5，图6，图7中所示的步骤。
所述存储单元920可以包括易失性存储单元形式的可读介质，例如随机存取存储单元(RAM)9201和/或高速缓存存储单元9202，还可以进一步包括只读存储单元(ROM)9203。
所述存储单元920还可以包括具有一组(至少一个)程序模块9205的程序/实用工具9204，这样的程序模块9205包括但不限于：操作系统、一个或者多个应用程序、其它程序模块以及程序数据，这些示例中的每一个或某种组合中可能包括网络环境的实现。
总线930可以为表示几类总线结构中的一种或多种，包括存储单元总线或者存储单元控制器、外围总线、图形加速端口、处理单元或者使用多 种总线结构中的任意总线结构的局域总线。
电子设备900也可以与一个或多个外部设备900’(例如键盘、指向设备、蓝牙设备等)通信，使得用户能与该电子设备900交互的设备通信，和/或该电子设备900能与一个或多个其它计算设备进行通信的任何设备(例如路由器、调制解调器等等)通信。这种通信可以通过输入/输出(I/O)接口950进行。并且，电子设备900还可以通过网络适配器960与一个或者多个网络(例如局域网(LAN)，广域网(WAN)和/或公共网络，例如因特网)通信。网络适配器960可以通过总线930与电子设备900的其它模块通信。应当明白，尽管图中未示出，可以结合电子设备900使用其它硬件和/或软件模块，包括但不限于：微代码、设备驱动器、冗余处理单元、外部磁盘驱动阵列、RAID系统、磁带驱动器以及数据备份存储系统等。
通过以上的实施方式的描述，本领域的技术人员易于理解，这里描述的示例实施方式可以通过软件实现，也可以通过软件结合必要的硬件的方式来实现。因此，如图10所示，根据本申请实施方式的技术方案可以以软件产品的形式体现出来，该软件产品可以存储在一个非易失性存储介质(可以是CD-ROM，U盘，移动硬盘等)中或网络上，包括若干指令以使得一台计算设备(可以是个人计算机、服务器、或者网络设备等)执行根据本申请实施方式的上述方法。
所述软件产品可以采用一个或多个可读介质的任意组合。可读介质可以是可读信号介质或者可读存储介质。可读存储介质例如可以为但不限于电、磁、光、电磁、红外线、或半导体的系统、装置或器件，或者任意以上的组合。可读存储介质的更具体的例子(非穷举的列表)包括：具有一个或多个导线的电连接、便携式盘、硬盘、随机存取存储器(RAM)、只读存储器(ROM)、可擦式可编程只读存储器(EPROM或闪存)、光纤、便携式紧凑盘只读存储器(CD-ROM)、光存储器件、磁存储器件、或者上述的任意合适的组合。
所述计算机可读存储介质可以包括在基带中或者作为载波一部分传 播的数据信号，其中承载了可读程序代码。这种传播的数据信号可以采用多种形式，包括但不限于电磁信号、光信号或上述的任意合适的组合。可读存储介质还可以是可读存储介质以外的任何可读介质，该可读介质可以发送、传播或者传输用于由指令执行系统、装置或者器件使用或者与其结合使用的程序。可读存储介质上包含的程序代码可以用任何适当的介质传输，包括但不限于无线、有线、光缆、RF等等，或者上述的任意合适的组合。
可以以一种或多种程序设计语言的任意组合来编写用于执行本申请操作的程序代码，所述程序设计语言包括面向对象的程序设计语言—诸如Java、C++等，还包括常规的过程式程序设计语言—诸如“C”语言或类似的程序设计语言。程序代码可以完全地在用户计算设备上执行、部分地在用户设备上执行、作为一个独立的软件包执行、部分在用户计算设备上部分在远程计算设备上执行、或者完全在远程计算设备或服务器上执行。在涉及远程计算设备的情形中，远程计算设备可以通过任意种类的网络，包括局域网(LAN)或广域网(WAN)，连接到用户计算设备，或者，可以连接到外部计算设备(例如利用因特网服务提供商来通过因特网连接)。
上述计算机可读介质承载有一个或者多个程序，当上述一个或者多个程序被一个该设备执行时，使得该计算机可读介质实现如下功能：获取待测样本，所述待测样本中包含多种目标分析物；对所述待测样本进行基于磁性粒子的生化反应处理以生成反应溶液，所述反应溶液中包含多种多元复合物，其中，多种多元复合物与多种目标分析物对应；对所述反应溶液进行激发以使所述多种多元复合物发出多种光；在外界磁场作用下将目标分析物从待测样本中分离出来以及收集在磁性粒子的表面；基于单一光学检测通道获取所述多种光以生成样本检测图片；基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果；基于每一种目标分析物的检测结果生成所述待测样本的检测结果。该计算机可读介质还可实现如下功能：通过多个检测样本生成多个样本训练图片；为所述多个样本训练图片分别标注多个训练标签，其中，所述多个训练标签与所述多种目标分析物对应；基于带有多个训练标签的多个样本训练图片对机器学习模型进行训练以生成所述多目标识别模型。
本领域技术人员可以理解上述各模块可以按照实施例的描述分布于装置中，也可以进行相应变化唯一不同于本实施例的一个或多个装置中。上述实施例的模块可以合并为一个模块，也可以进一步拆分成多个子模块。
通过以上的实施例的描述，本领域的技术人员易于理解，这里描述的示例实施例可以通过软件实现，也可以通过软件结合必要的硬件的方式来实现。因此，根据本申请实施例的技术方案可以以软件产品的形式体现出来，该软件产品可以存储在一个非易失性存储介质(可以是CD-ROM，U盘，移动硬盘等)中或网络上，包括若干指令以使得一台计算设备(可以是个人计算机、服务器、移动终端、或者网络设备等)执行根据本申请实施例的方法。
以上具体地示出和描述了本申请的示例性实施例。应可理解的是，本申请不限于这里描述的详细结构、设置方式或实现方法；相反，本申请意图涵盖包含在所附权利要求的精神和范围内的各种修改和等效设置。

Claims (26)

  1. 一种基于单一光学检测通道的样本检测方法，其特征在于，包括：
    获取待测样本，所述待测样本中包含多种目标分析物；
    对所述待测样本进行基于磁性粒子的生化反应处理以生成反应溶液，所述反应溶液中包含多种多元复合物，其中，多种多元复合物与多种目标分析物对应；
    对所述反应溶液进行激发以使所述多种多元复合物发出多种光；
    在外界磁场作用下将目标分析物从待测样本中分离出来以及收集在磁性粒子的表面；
    基于单一光学检测通道获取所述多种光以生成样本检测图片；
    基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果；
    基于每一种目标分析物的检测结果生成所述待测样本的检测结果。
  2. 如权利要求1所述的样本检测方法，其特征在于，对所述待测样本进行基于磁性粒子的生化反应处理以生成反应溶液包括：
    基于生物载体技术对所述待测样本进行基于磁性粒子的非酶生化反应处理以生成所述反应溶液。
  3. 如权利要求2所述的样本检测方法，其特征在于，基于生物载体技术对所述待测样本进行基于磁性粒子的非酶生化反应处理以生成所述反应溶液包括：
    对所述待测样本进行非酶生化反应处理以使得待测样本中的多种目标分析物分别和多种生物载体结合以生成所述反应溶液；
    利用磁性粒子，在外界磁场作用下对目标分析物进行分离。
  4. 如权利要求3所述的样本检测方法，其特征在于，对所述待测样本进行非酶生化反应处理以使得待测样本中的多种目标分析物分别和多种生物载体结合以生成所述反应溶液包括：
    对所述待测样本进行非酶生化反应处理以使得待测样本中的多种目标分析物分别和与其对应的磁性粒子、抗体、待测物、荧光纳米珠信号探针结合以生成所述反应溶液。
  5. 如权利要求1所述的样本检测方法，其特征在于，对所述反应溶液进行激发以使所述多种多元复合物发出多种光包括：
    通过光源设备或电学设备或化学试剂对所述反应溶液进行激发以使所述多种多元复合物发出多种光，所述多种光包括具有不同的强度和/或频率的光。
  6. 如权利要求1所述的样本检测方法，其特征在于，对所述反应溶液进行激发以使所述多种多元复合物发出多种光包括：
    通过光源对所述反应溶液进行激发以使所述多种多元复合物发出多种光。
  7. 如权利要求6所述的样本检测方法，其特征在于，通过光源对所述反应溶液进行激发以使所述多种多元复合物发出多种光包括：
    将所述反应溶液放置于透明平面；
    待反应溶液中的磁性粒子沉淀在平面的表面上，通过蓝色光源对置于透明平面的所述反应溶液进行激发以使所述多种多元复合物发出多种光。
  8. 如权利要求1所述的样本检测方法，其特征在于，基于单一光学检测通道获取所述多种光以生成样本检测图片包括：
    基于单一光学检测通道中的滤光透镜获取所述多种光；
    光学传感器基于所述多种光生成所述样本检测图片。
  9. 如权利要求8所述的样本检测方法，其特征在于，光学传感器基于所述多种光生成所述样本检测图片包括：
    CMOS光学传感器基于所述多种光生成所述样本检测图片；或
    CCD光学传感器基于所述多种光生成所述样本检测图片。
  10. 如权利要求1所述的样本检测方法，其特征在于，基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果包括：
    对所述样本检测图片进行预处理；
    生成所述样本检测图片的特征矩阵；
    将所述特征矩阵输入多目标识别模型中，生成所述多种目标分析物中每一种目标分析物的检测结果。
  11. 如权利要求10所述的样本检测方法，其特征在于，对所述样本检测图片进行预处理包括：
    对所述样本检测图片进行滤波处理；和/或
    对所述样本检测图片进行斑点检测处理。
  12. 如权利要求10所述的样本检测方法，其特征在于，生成所述样本检测图片的特征矩阵包括：
    提取所述样本检测图片中每一个像素的特征数据；
    基于所有像素的特征数据生成所述特征矩阵。
  13. 如权利要求12所述的样本检测方法，其特征在于，提取所述样本检测图片中每一个像素的特征数据包括：
    提取所述样本检测图片中每一个像素的颜色和位置。
  14. 如权利要求12所述的样本检测方法，其特征在于，基于所有像 素的特征数据生成所述特征矩阵还包括：
    对所述特征矩阵进行特征变换处理。
  15. 如权利要求10所述的样本检测方法，其特征在于，将所述特征矩阵输入多目标识别模型中，生成所述多种目标分析物中每一种目标分析物的检测结果包括：
    将所述特征矩阵输入多目标识别模型中；
    所述多目标识别模型基于所述特征矩阵对所述样本检测图片中的像素逐一进行分类预测；
    基于分类预测结果生成所述多种目标分析物中每一种目标分析物的检测结果。
  16. 如权利要求10所述的样本检测方法，其特征在于，生成所述多种目标分析物中每一种目标分析物的检测结果包括：
    生成所述多种目标分析物中每一种目标分析物对应的像素总数和总亮度；
    基于像素总数和总亮度生成所述多种目标分析物中每一种目标分析物的检测结果。
  17. 如权利要求16所述的样本检测方法，其特征在于，基于像素总数和总亮度生成所述多种目标分析物中每一种目标分析物的检测结果包括：
    基于校准公式和每一种目标分析物对应的像素总数和总亮度生成每一种目标分析物对应的数量和浓度；
    基于每一种目标分析物对应的数量和浓度生成所述待测样本的检测结果。
  18. 如权利要求10所述的样本检测方法，其特征在于，还包括：
    通过多个检测样本生成多个样本训练图片；
    为所述多个样本训练图片分别标注多个训练标签，其中，所述多个训练标签与所述多种目标分析物对应；
    基于带有多个训练标签的多个样本训练图片对机器学习模型进行训练以生成所述多目标识别模型。
  19. 如权利要求18所述的样本检测方法，其特征在于，为所述多个样本训练图片分别标注多个训练标签包括：
    为所述多个样本训练图片中的每个像素分别标注训练标签。
  20. 如权利要求18所述的样本检测方法，其特征在于，基于带有多个训练标签的多个样本训练图片对机器学习模型进行训练以生成所述多目标识别模型包括：
    基于带有多个训练标签的多个样本训练图片对多层感知器进行训练以生成所述多目标识别模型；或
    基于带有多个训练标签的多个样本训练图片对随机森林分类器进行训练以生成所述多目标识别模型。
  21. 一种基于单一光学检测通道的样本检测装置，其特征在于，包括：
    样本模块，用于获取待测样本，所述待测样本中包含多种目标分析物；
    反应模块，用于对所述待测样本进行基于磁性粒子的生化反应处理以生成反应溶液，所述反应溶液中包含多种多元复合物，其中，多种多元复合物和多种目标分析物对应；
    激发分离模块，用于对所述反应溶液进行激发以使所述多种多元复合物发出多种光，在外界磁场作用下将目标分析物从待测样本中分离出来以及收集在磁性粒子的表面；
    图片模块，用于基于单一光学检测通道获取所述多种光以生成样本检 测图片；
    检测模块，用于基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果；
    结果模块，用于基于每一种目标分析物的检测结果生成所述待测样本的检测结果。
  22. 如权利要求21所述的样本检测装置，其特征在于，所述检测模块包括：
    处理单元，用于对所述样本检测图片进行预处理；
    特征单元，用于生成所述样本检测图片的特征矩阵；
    识别单元，用于将所述特征矩阵输入多目标识别模型中，生成所述多种目标分析物中每一种目标分析物的检测结果。
  23. 如权利要求22所述的样本检测装置，其特征在于，所述检测模块还包括：
    模型训练单元，用于通过多个检测样本生成多个样本训练图片；为多个样本训练图片分别标注多个训练标签，其中，所述多个训练标签与所述多种目标分析物对应；基于带有多个训练标签的多个样本训练图片对机器学习模型进行训练以生成所述多目标识别模型。
  24. 一种基于单一光学检测通道的样本检测系统，其特征在于，包括：
    反应装置，用于获取待测样本，所述待测样本中包含多种目标分析物，对所述待测样本进行基于磁性粒子的生化反应处理以生成反应溶液，所述反应溶液中包含多种多元复合物，其中，多种多元复合物与多种目标分析物对应；
    激发分离装置，用于对所述反应溶液进行激发以使所述多种多元复合物发出多种光，在外界磁场作用下将目标分析物从待测样本中分离出来以及收集在磁性粒子的表面；
    图像装置，用于基于单一光学检测通道获取所述多种光以生成样本检测图片；
    服务器，用于基于所述样本检测图片生成所述多种目标分析物中每一种目标分析物的检测结果，基于每一种目标分析物的检测结果生成所述待测样本的检测结果。
  25. 一种电子设备，其特征在于，包括：
    一个或多个处理器；
    存储装置，用于存储一个或多个程序；
    当所述一个或多个程序被所述一个或多个处理器执行，使得所述一个或多个处理器实现如权利要求1-20中任一所述的方法。
  26. 一种计算机可读介质，其上存储有计算机程序，其特征在于，所述程序被处理器执行时实现如权利要求1-20中任一所述的方法。
PCT/CN2022/129869 2022-01-18 2022-11-04 样本检测方法、装置、系统、电子设备及计算机可读介质 WO2023138162A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210056449.4A CN116503302A (zh) 2022-01-18 2022-01-18 样本检测方法、装置、系统、电子设备及计算机可读介质
CN202210056449.4 2022-01-18

Publications (1)

Publication Number Publication Date
WO2023138162A1 true WO2023138162A1 (zh) 2023-07-27

Family

ID=87327164

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/129869 WO2023138162A1 (zh) 2022-01-18 2022-11-04 样本检测方法、装置、系统、电子设备及计算机可读介质

Country Status (2)

Country Link
CN (1) CN116503302A (zh)
WO (1) WO2023138162A1 (zh)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107003311A (zh) * 2014-10-29 2017-08-01 蒙斯大学 疟疾检测
US20180361396A1 (en) * 2015-07-02 2018-12-20 University College Dublin National University Of Ireland, Dublin An Optical Detection Based on Non-Linear Magnetophoretic Transport of Magnetic Particle for Particle and Biological Sensing and Separation
CN110261608A (zh) * 2019-05-29 2019-09-20 江苏大学 基于磁性荧光探针的食品大肠杆菌菌落可视化检测及自动化计数方法
CN112462056A (zh) * 2020-11-19 2021-03-09 武汉大学 一种用于现场检测尿液中细菌的尿检平台及其使用方法
CN113138270A (zh) * 2020-01-20 2021-07-20 上海纳衍生物科技有限公司 一种血液样本的多目标物检测方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107003311A (zh) * 2014-10-29 2017-08-01 蒙斯大学 疟疾检测
US20180361396A1 (en) * 2015-07-02 2018-12-20 University College Dublin National University Of Ireland, Dublin An Optical Detection Based on Non-Linear Magnetophoretic Transport of Magnetic Particle for Particle and Biological Sensing and Separation
CN110261608A (zh) * 2019-05-29 2019-09-20 江苏大学 基于磁性荧光探针的食品大肠杆菌菌落可视化检测及自动化计数方法
CN113138270A (zh) * 2020-01-20 2021-07-20 上海纳衍生物科技有限公司 一种血液样本的多目标物检测方法
CN112462056A (zh) * 2020-11-19 2021-03-09 武汉大学 一种用于现场检测尿液中细菌的尿检平台及其使用方法

Also Published As

Publication number Publication date
CN116503302A (zh) 2023-07-28

Similar Documents

Publication Publication Date Title
US20220374264A1 (en) Methods, systems, and devices for real time execution and optimization of concurrent test protocols on a single device
US20220274109A1 (en) Single-use test device for imaging blood cells
KR102149318B1 (ko) 소량 응집 검정
JP2019168471A (ja) サンプルを代表する光を検出すること及び利用すること
KR101431843B1 (ko) 생체인식 분자에 접합된 마이크로비드를 사용하여 병원체를검출하는 방법
US20190056304A1 (en) Method of imaging blood cells
JP5606285B2 (ja) 分析方法および装置
US20160341752A1 (en) Methods, systems, and devices for real time execution and optimization of concurrent test protocols on a single device
US20210311038A1 (en) Indicator-based analysis of a sample
US11313857B2 (en) System and method for identifying and quantifying species with nanopores, using complexes of nanoparticles with carrier particles
WO2014127269A1 (en) Methods, systems, and devices for real time execution and optimization of concurrent test protocols on a single device
KR20220100854A (ko) 인공 지능 기반 세포 분석을 위한 시스템 및 방법
EP3882603A1 (en) Information processing device, information processing method, and computer program
Ahmadsaidulu et al. Microfluidic point-of-care diagnostics for multi-disease detection using optical techniques: a review
US20190056384A1 (en) Single-use test device for imaging assay beads
JP5543310B2 (ja) イムノクロマトグラフ検査方法および装置
KR20170087169A (ko) 생체분자의　계수분석　방법, 키트 및 장치, 그리고　그들의 이용
WO2023138162A1 (zh) 样本检测方法、装置、系统、电子设备及计算机可读介质
JP5543888B2 (ja) イムノクロマトグラフ検査方法および装置
US20190056385A1 (en) Method of imaging assay beads in a biological sample
US20210356460A1 (en) Method for Analyzing Immunoglobulins and Other Analytes in an Immunoassay
KR20170142157A (ko) 생체분자의　계수분석　방법, 키트 및 장치, 그리고　그들의 이용
JP7010293B2 (ja) 情報処理装置、情報処理方法及びプログラム
US20170108435A1 (en) Fluorometer
Wei Temporal Multiplexed Fluorescence Imaging Device for Bead-based Immunoassay Detection

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22921586

Country of ref document: EP

Kind code of ref document: A1