Method, system, equipment and medium for extracting characteristics of microbeads in high-density gene chips

Abstract

The present invention provides a method, system, device, and medium for extracting features from microbeads in a high-density gene chip, comprising: obtaining a chip scan image of the microbeads whose features are to be extracted; inputting the chip scan image into a constructed microbead brightness classification model to perform brightness classification; establishing a mapping relationship between the output weights of the trained microbead brightness classification model and the brightness classification results of the microbeads, generating a feature extraction kernel for the microbeads whose features are to be extracted, and performing feature extraction on the microbeads whose features are to be extracted based on the feature extraction kernel. The present invention constructs a feature extraction kernel by constructing a mapping relationship between different weights output by the microbead brightness classification model and the brightness of the microbeads, thereby making the microbeads more adaptable to actual data during feature extraction and avoiding the influence of subjective factors when manually setting weighted mean weights.

Description

Technical Field

[0001] The present invention relates to the technical field of biochips, and in particular to a method, system, equipment and medium for extracting characteristics of microbeads in a high-density gene chip. Background Art

[0002] The chip scan image (grayscale image) is the most original image of chip quality. The grayscale value of the point on the image reflects the fluorescence intensity information of the probe at the corresponding chip position. The grayscale image can be used to perform preliminary quality control of the chip. For each chip scan image, after successfully locating the center pixel coordinates of each microsphere in the image, if the grayscale representation of these microspheres is required, the most direct method is to take the center pixel coordinates of each microbead as the center, take a rectangular area or a circular area with an appropriate side length or diameter, and calculate the mean grayscale value of all pixels in this area to achieve feature extraction.

[0003] However, in the prior art, the selection of regions is usually done manually, and when calculating the mean grayscale value of each pixel in the pre-fetch, each pixel value is not weighted, resulting in the pixel values ​​of pixels farther / closer to the center contributing equally to the final mean value, which is obviously unreasonable. Artificially assigning thresholds from large to small based on the distance of pixels from the center point, from near to far, introduces a strong subjective factor, resulting in inaccurate and non-robust final results. Therefore, it is necessary to design a more reasonable, accurate, and robust feature extraction method that adapts to actual data. Summary of the Invention

[0004] In order to solve the above technical problems, the present invention provides a method, system, device and medium for extracting features of microbeads in high-density gene chips. By learning from a large amount of historical data, appropriate weights are obtained to weight each pixel value in the microbead area, so that the feature extraction of microbeads is more adapted to the actual data and avoids the influence of subjective factors when manually setting the weighted mean weight.

[0005] In a first aspect, the present invention provides a method for extracting features of microbeads in a high-density gene chip, comprising:

[0006] Acquire a chip scanning image of the microbeads whose features are to be extracted;

[0007] Inputting the chip scan image into the constructed microbead brightness and darkness classification model to classify the brightness and darkness;

[0008] Establishing a mapping relationship between the different weights output by the trained microbead brightness and darkness classification model and the brightness and darkness classification results, generating a feature extraction kernel for the microbeads to be extracted, and performing feature extraction on the microbeads to be extracted based on the feature extraction kernel;

[0009] The weights are parameters of the microbead light and dark classification model after training.

[0010] According to a specific embodiment of the present invention, the process of constructing the microbead brightness and darkness classification model includes the following steps:

[0011] Selecting successfully decoded microbeads from different high-density gene chips, obtaining a microbead region of the same size as the desired feature extraction kernel and a corresponding label value on each successfully decoded microbead, and using these as a data set;

[0012] Build a classification model architecture based on weight extraction layers and activation functions;

[0013] Inputting the data set into the classification model framework to perform classification training on the brightness and darkness of microbeads to complete the construction of the microbead brightness and darkness classification model;

[0014] Among them, the weight extraction layer outputs a feature vector or feature map as the output result after processing the data set based on a single-layer fully connected layer or a single-layer convolutional layer, and the output result is subjected to the activation function to classify the brightness and darkness of the beads, and the classification result is used as the optimization target to train the bead brightness and darkness classification model. After the training is completed, the parameters of the weight extraction layer are taken out as weights to generate feature extraction kernels with different weights and mapping relationships with the brightness and darkness of the beads.

[0015] According to a specific embodiment of the present invention, the step of the weight extraction layer generating output results with different weights after processing the data set based on a single-layer fully connected layer includes: during training, converting the result of the data set after processing by the single-layer fully connected layer from a one-dimensional form to a two-dimensional form, and then generating output results based on different weights, wherein the number of nodes in the single-layer fully connected layer is set to 1.

[0016] According to a specific embodiment of the present invention, the step of the weight extraction layer generating output results with different weights after processing the data set based on the single-layer convolution layer includes: during training, the result of the data set processed by the single-layer convolution layer is directly used as the output result based on different weights.

[0017] According to a specific embodiment of the present invention, the output brightness of the microbeads includes the ratio r between the bright beads and the dark beads of the microbeads, wherein the bright bead error is 1 / r of the dark bead error, and the loss function is set to one of the BCE loss function, the Focal loss function or the Contrastive loss function.

[0018] According to a specific embodiment of the present invention, the size of the feature extraction kernel generated during training is a preset size, and all weights in the feature extraction kernel are multiplied by a fixed constant so that the sum of all weights is 1.

[0019] According to a specific embodiment of the present invention, the activation function is one of a sigmoid function, a softsign function or a swish function.

[0020] In a second aspect, the present invention further provides a feature extraction system for microbeads in a high-density gene chip, the feature extraction system being used to implement the feature extraction method for microbeads in a high-density gene chip as described in the first aspect, the system comprising:

[0021] An acquisition module, used to acquire a chip scanning image of the microbeads whose features are to be extracted;

[0022] A processing module is used to pre-build a microbead brightness and darkness classification model, and input the chip scanning image into the microbead brightness and darkness classification model to perform brightness and darkness classification;

[0023] an extraction module, configured to establish a mapping relationship between the different weights output by the trained microbead brightness classification model and the brightness classification results, generate a feature extraction kernel for the microbeads whose features are to be extracted, and perform feature extraction on the microbeads whose features are to be extracted based on the feature extraction kernel;

[0024] The weights are parameters of the microbead light and dark classification model after training.

[0025] In a third aspect, the present invention further provides an electronic device, comprising:

[0026] at least one processor; and,

[0027] a memory communicatively connected to the at least one processor; wherein,

[0028] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the feature extraction method for microbeads in a high-density gene chip described in the first aspect.

[0029] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the method for extracting features of microbeads in a high-density gene chip according to the first aspect.

[0030] In summary, the present invention provides a method, system, device and medium for extracting features of microbeads in high-density gene chips, including: obtaining a chip scanning image of the microbeads whose features are to be extracted; inputting the chip scanning image into a constructed microbead brightness classification model to perform brightness classification; establishing a mapping relationship between the different weights output by the trained microbead brightness classification model and the brightness classification results, generating a feature extraction core for the microbeads whose features are to be extracted, and performing feature extraction on the microbeads whose features are to be extracted based on the feature extraction core. The present invention constructs a feature extraction core by constructing a mapping relationship between different weights output by the microbead brightness classification model and the brightness of the microbeads, so that the microbeads are more adaptable to the actual data during feature extraction, and also avoids the influence of subjective factors when manually setting weighted mean weights. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0032] Figure 1 A schematic diagram of the process of extracting features of microbeads in a high-density gene chip provided by the present invention;

[0033] Figure 2 This is a schematic diagram of the application of the feature extraction core in the feature extraction method of microbeads in a high-density gene chip provided by the present invention;

[0034] Figure 3 This is a schematic diagram of the framework of the feature extraction system for microbeads in a high-density gene chip provided by the present invention. DETAILED DESCRIPTION

[0035] The following describes the embodiments of the present disclosure through specific concrete examples, and those skilled in the art can easily understand other advantages and effects of the present disclosure from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all of the embodiments. The present disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present disclosure. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present disclosure.

[0036] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein may be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on this disclosure, it should be understood by those skilled in the art that an aspect described herein may be implemented independently of any other aspect, and two or more of these aspects may be combined in various ways. For example, any of the aspects described herein may be used to implement an apparatus and / or practice a method. In addition, other structures and / or functionalities other than one or more of the aspects described herein may be used to implement this apparatus and / or practice this method.

[0037] Additionally, in the following description, specific details are provided to provide a thorough understanding of the examples. However, one skilled in the art will appreciate that the aspects described can be practiced without these specific details.

[0038] refer to Figure 1 The present invention provides a method for extracting features of microbeads in a high-density gene chip, comprising:

[0039] S101 obtains a scanning image of the chip of the microbeads whose features are to be extracted;

[0040] S102. Inputting the chip scan image into the constructed microbead brightness classification model for brightness and darkness classification;

[0041] S103. Establishing a mapping relationship between the different weights output by the trained microbead brightness classification model and the brightness classification results, generating a feature extraction kernel for the microbeads to be extracted, and performing feature extraction on the microbeads to be extracted based on the feature extraction kernel;

[0042] The weights are parameters of the microbead light and dark classification model after training.

[0043] In an embodiment of the present application, the brightness and darkness classification of microbeads in a high-density gene chip is achieved by constructing a microbead brightness and darkness classification model. The implementation of the above steps is an embodiment of a specific application based on the completion of the construction of the microbead brightness and darkness classification model. After the model is constructed, the brightness and darkness of the acquired chip scan image can be classified. As the technical content for implementing this embodiment, the specific construction process of the microbead brightness and darkness classification model will be explained below.

[0044] It should be noted that the construction process of the microbead brightness and darkness classification model in this embodiment includes the preparation of a data set, the construction of a classification model architecture, and the model training process, which will be described in detail below.

[0045] In this embodiment, beads that have been successfully decoded are selected from different high-density gene chips, and the bead area with the same size as the desired feature extraction kernel and the corresponding label value are obtained on each successfully decoded bead, and used as a data set. Generally, the beads in high-density gene chips are usually 3-5 mm in diameter. Magnetic or fluorescent microspheres, and specific probes such as DNA, antibodies, etc. are fixed on their surface. Each microbead type corresponds to a unique probe and a unique code, wherein the unique probe is targeted at a specific gene sequence / protein marker, and the unique code is identified by a combination of fluorescent dyes or physical markers. Common decoding modes include fluorescent coding decoding, spatial positioning decoding, and sequence barcode decoding. In this embodiment, these methods can be used to decode the microbeads of a high-density gene chip. After obtaining the successfully decoded microbeads, a microbead area of ​​the same size as the desired feature extraction kernel is taken for training the model. At the same time, the label value of the successfully decoded microbead can also be obtained.

[0046] The dataset constructed through the above processing is of good quality, and the labels obtained and used for training are also relatively reliable, so that different weights can be set according to the number ratio of different categories in the loss function below.

[0047] After the above-mentioned data set is constructed, the process of constructing the classification model architecture begins. In this embodiment, the classification model architecture is implemented based on the weight extraction layer and the activation function. In order to achieve the purpose of this embodiment, the weight extraction layer can be set to a single-layer fully connected layer or a single-layer convolutional layer. The parameters after training the single-layer fully connected layer or the large-layer convolutional layer are used as weights to generate output results with different weights.

[0048] In this embodiment, if a single fully connected layer is used, the number of nodes in the single fully connected layer is set to 1, and the two-dimensional image data is flattened in the input layer so that the number of nodes in the input layer is the number of pixels in the bead area. When training, the output of the single fully connected layer is The one-dimensional form is converted to The two-dimensional form outputs the output results based on different weights.

[0049] In this embodiment, for a single-layer convolution layer, the spatial structure is usually maintained and does not need to be adjusted. Therefore, the size of the single-layer convolution layer can be set to the same as the size of the microbead area, and the two-dimensional image data can be directly input into the input layer. When training, the result of processing the data set through the single-layer convolution layer is directly used as the output result based on different weights. It should be noted that the weight extraction layer in this embodiment can be a single-layer fully connected layer, a single-layer convolution layer, or a logistic regression model as described above. The purpose of the present invention is to use the obtained model parameters as weights to establish a mapping relationship between different weights and the brightness and darkness of the microbeads. Therefore, in practical applications, a single-layer fully connected layer, a single-layer convolution layer, or a logistic regression model can be used as an option for the weight extraction layer, thereby achieving the purpose of the present invention.

[0050] After the above process, the weight extraction layer processes the data set based on a single-layer fully connected layer or a single-layer convolutional layer, outputting a feature vector or feature map, and using the output feature vector or feature map as the output result. On the one hand, the output result is subjected to binary classification through an activation function to generate a classification result of the brightness and darkness of the microbeads, and the entire model is trained with the classification result as the optimization target. Since the selected training data comes from microbeads that are successfully decoded in accordance with experimental expectations, the output brightness and darkness of the microbeads includes the ratio r between the bright and dark beads of the microbeads, where the bright bead error is 1 / r of the dark bead error, and the loss function is set to the BCE loss function. It should be noted that the classification of the brightness and darkness of the microbeads is based on the grayscale value of each successfully decoded microbead in the data set as the input image and result label, and training extraction is performed to achieve the brightness and darkness classification of the microbeads. This process can achieve the classification of the brightness and darkness of the microbeads in any way, and this embodiment does not specifically limit it.

[0051] On the other hand, after training is complete, the parameters of a single fully connected layer or a single convolutional layer are extracted as weights and converted into the desired feature extraction kernel size. Simultaneously, all weights are multiplied by a fixed constant so that the sum of all weights in the feature extraction kernel is 1. In this embodiment, a mapping relationship is established between the weights and the brightness of the beads, i.e., each weight corresponds to a brightness value of a bead. Large weight values ​​can be set to correspond to high brightness values ​​of the beads. This will be illustrated with an example below. At this point, the feature extraction kernel generation process completes the training phase. In step S103, after the beads to be extracted are obtained, they are subjected to the bead brightness classification model to generate brightness classification results. Based on the parameters output by the trained model and the brightness values ​​corresponding to the weights, a mapping relationship is established between the weights and the brightness classification results of the beads to be extracted, and the corresponding feature extraction kernel is generated. Feature extraction is then performed based on this feature extraction kernel.

[0052] It should be noted that the activation function in this embodiment can be a sigmoid function, the loss function can be a BCE loss function, the activation function can also be a softsign function or a swish function, the loss function can also be a Focal loss function or a Contrastive loss function, and in practice it can also be other activation functions or other loss functions to achieve the corresponding functions. This embodiment does not make specific limitations.

[0053] In order to more clearly describe the content of the embodiments of this application, Figure 2 shows a schematic diagram of the application of feature extraction kernel, such as Figure 2 As shown in the figure, the mapping relationship between the weight distribution in the feature extraction kernel and the brightness of the beads is such that the pixels closer to the center have higher weights. Because the central area of ​​the feature extraction kernel also corresponds to the center of each bead, the pixel values ​​in this area can better reflect the brightness of the bead itself; the weights of the areas slightly farther from the center are lower because these areas correspond to the edge of each bead, where the grayscale values ​​of bright beads are attenuated compared to the center of the bead, and therefore are not as representative of the grayscale values ​​of the beads as the center area of ​​the bead; the weights of the areas farthest from the center are even negative because these areas correspond to the edge of the bead or even the background area, and cannot reasonably reflect the brightness of the beads. The negative weights here also highlight the weights of the areas close to the center.

[0054] Meanwhile, Table 1 shows the test results of comparative tests conducted between the existing common mean method and the method of this embodiment by setting multiple groups of test data.

[0055] Table 1

[0056]

[0057] In Table 1, "decoding rate" is defined as the ratio of the number of beads whose decoding results meet experimental expectations to the total number of beads on the chip; "decoded bead type ratio" is defined as the ratio of the number of successfully decoded bead types (a certain number of beads of each bead type are considered decoded successfully) to the total number of bead types actually decoded in the experiment. After processing using the feature extraction method described in this example, both methods showed significant improvements compared to existing common averaging methods.

[0058] Based on the same inventive concept, the present application also provides a feature extraction system for implementing the aforementioned method for extracting features from microbeads in a high-density gene chip. The solution provided by this system is similar to the solution described in the aforementioned method. Therefore, the specific limitations of one or more embodiments of the feature extraction system for microbeads in a high-density gene chip provided below can be found in the aforementioned limitations on the feature extraction method and will not be further elaborated here.

[0059] In one embodiment, reference Figure 3 This embodiment further provides a feature extraction system 300 for microbeads in a high-density gene chip, comprising:

[0060] An acquisition module 301 is used to acquire a chip scanning image of a microbead whose features are to be extracted;

[0061] Processing module 302, for pre-building a microbead brightness and darkness classification model, and inputting the chip scan image into the microbead brightness and darkness classification model for brightness and darkness classification;

[0062] An extraction module 303 is configured to establish a mapping relationship between the different weights output by the trained microbead brightness classification model and the brightness classification results, generate a feature extraction kernel for the microbeads whose features are to be extracted, and perform feature extraction on the microbeads whose features are to be extracted based on the feature extraction kernel;

[0063] The weights are parameters of the microbead light and dark classification model after training.

[0064] In one embodiment, the processing module 302 is specifically used to pre-construct a microbead brightness and darkness classification model, including obtaining a data set for model training, constructing a classification model architecture based on a weight extraction layer and an activation function, and inputting the data set into the classification model architecture to perform microbead brightness and darkness classification training to complete the construction of the microbead brightness and darkness classification model.

[0065] Figure 3 The specific implementation of the system shown can refer to the specific implementation of the aforementioned method embodiment, which will not be repeated here.

[0066] In addition, an embodiment of the present application further provides an electronic device, including:

[0067] at least one processor; and,

[0068] a memory communicatively connected to the at least one processor; wherein,

[0069] The memory stores instructions that can be executed by the at least one processor. The instructions are executed by the at least one processor to enable the at least one processor to perform the feature extraction method for microbeads in a high-density gene chip in the above method embodiment.

[0070] An embodiment of the present application further provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the method for extracting features of microbeads in a high-density gene chip in the aforementioned method embodiment.

[0071] It should be noted that the computer-readable medium described herein may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. Computer-readable storage media may be, but are not limited to, electrical, magnetic, optical, electromagnetic, infrared semiconductor systems, devices, or components, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0072] In this disclosure, a computer-readable storage medium may be any tangible medium that contains or stores a program for use by or in conjunction with an instruction execution system, apparatus, or device. Furthermore, in this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device. Program code embodied on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wire, fiber optic cable, RF (radio frequency), or any suitable combination thereof.

[0073] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0074] The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device: obtains a chip scan image of the microbeads whose features are to be extracted; inputs the chip scan image into the constructed microbead brightness and darkness classification model for brightness and darkness classification; establishes a mapping relationship between the different weights output by the trained microbead brightness and darkness classification model and the brightness and darkness classification results, generates a feature extraction core for the microbeads whose features are to be extracted, and extracts features of the microbeads whose features are to be extracted based on the feature extraction core; wherein the weights are parameters of the microbead brightness and darkness classification model after training.

[0075] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer, partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0076] The flow charts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flow chart or block diagram can represent a part of a module or program segment code, and the part of this module or program segment code contains one or more executable instructions for realizing the prescribed logical functions. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a sequence different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of the boxes in the block diagram and / or flow chart, can be implemented by a dedicated hardware-based system that performs the prescribed function or operation, or can be implemented by a combination of dedicated hardware and computer instructions.

[0077] It should be understood that various parts of the present disclosure can be implemented in hardware, software, firmware, or a combination thereof.

[0078] The above description is merely a specific embodiment of the present disclosure, but the scope of protection of the present disclosure is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this disclosure should be included in the scope of protection of the present disclosure. Therefore, the scope of protection of the present disclosure should be based on the scope of protection of the claims.

Claims

1. A method for extracting features of microbeads in a high-density gene chip, characterized in that: The steps include: Acquire a chip scanning image of the microbeads whose features are to be extracted; Inputting the chip scan image into the constructed microbead brightness and darkness classification model to classify the brightness and darkness; Establishing a mapping relationship between the different weights output by the trained microbead brightness and darkness classification model and the brightness and darkness classification results, generating a feature extraction kernel for the microbeads to be extracted, and performing feature extraction on the microbeads to be extracted based on the feature extraction kernel; The construction process of the microbead brightness and darkness classification model includes the following steps: Selecting successfully decoded microbeads from different high-density gene chips, obtaining a microbead region of the same size as the desired feature extraction kernel and a corresponding label value on each successfully decoded microbead, and using these as a data set; Build a classification model architecture based on weight extraction layers and activation functions; Inputting the data set into the classification model framework to perform classification training on the brightness and darkness of microbeads to complete the construction of the microbead brightness and darkness classification model; Among them, the weight extraction layer outputs a feature vector or feature map as the output result after processing the data set based on a single-layer fully connected layer or a single-layer convolutional layer, and the output result is subjected to the activation function to classify the brightness and darkness of the beads, and the classification result is used as the optimization target to train the bead brightness and darkness classification model. After the training is completed, the parameters of the weight extraction layer are taken out as weights to generate feature extraction kernels with different weights and mapping relationships with the brightness and darkness of the beads.

2. The method for extracting characteristics of microbeads in a high-density gene chip according to claim 1, characterized in that: The step of the weight extraction layer generating output results with different weights after processing the data set based on the single-layer fully connected layer includes: during training, converting the result of the data set after being processed by the single-layer fully connected layer from a one-dimensional form to a two-dimensional form and then generating output results based on different weights, wherein the number of nodes in the single-layer fully connected layer is set to 1.

3. The method for extracting characteristics of microbeads in a high-density gene chip according to claim 1, characterized in that: The step of generating output results with different weights after the weight extraction layer processes the data set based on the single-layer convolution layer includes: during training, directly using the result of the data set processed by the single-layer convolution layer as the output result based on different weights.

4. The method for extracting characteristics of microbeads in a high-density gene chip according to claim 2 or 3, characterized in that: The output brightness of the microbeads includes the ratio r between the bright and dark microbeads, where the bright bead error is 1 / r of the dark bead error, and the loss function is set to one of the BCE loss function, the Focal loss function, or the Contrastive loss function.

5. The method for extracting characteristics of microbeads in a high-density gene chip according to claim 2 or 3, characterized in that: The size of the feature extraction kernel generated during training is a preset size, and all weights in the feature extraction kernel are multiplied by a fixed constant so that the sum of all weights is 1.

6. The method for extracting characteristics of microbeads in a high-density gene chip according to claim 1, characterized in that: The activation function is one of a sigmoid function, a softsign function or a swish function.

7. A feature extraction system for microbeads in a high-density gene chip, the feature extraction system being used to implement the feature extraction method for microbeads in a high-density gene chip according to any one of claims 1 to 6, characterized in that: The system comprises: An acquisition module, used to acquire a chip scanning image of the microbeads whose features are to be extracted; A processing module is used to pre-build a microbead brightness and darkness classification model, and input the chip scanning image into the microbead brightness and darkness classification model to perform brightness and darkness classification; The extraction module is used to establish a mapping relationship between the different weights output by the trained microbead brightness and darkness classification model and the brightness and darkness classification results, generate a feature extraction kernel for the microbeads whose features are to be extracted, and perform feature extraction on the microbeads whose features are to be extracted based on the feature extraction kernel.

8. An electronic device, characterized in that: include: at least one processor; as well as, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method for extracting features of microbeads in a high-density gene chip according to any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium, characterized in that The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the method for extracting characteristics of microbeads in a high-density gene chip according to any one of claims 1 to 6.

Images