Sperm bank sperm supply management system and method and storage medium

Abstract

The invention discloses a sperm library sperm supply management system, a sperm library sperm supply management method and a storage medium, and solves the problem that the efficiency and accuracy of sperm quality detection are reduced because multi-source heterogeneous physical examination information cannot be identified and evaluated from multiple dimensions when a re-reception module of an existing system screens the physical examination information of a sperm donation volunteer. The system comprises a user side, an information processing module, an evaluation management module, a supply management module and a sperm database. The system is provided with a user side which customizes a volunteer sperm donation plan based on basic information of a sperm donation volunteer and an information processing module which can perform normalization processing on multi-source heterogeneous physical examination parameters by adopting edge calculation and pre-construct a sperm quality evaluation model. The user side and the information processing module cooperate to realize preliminary screening of sperm donation volunteers, personalized customization of sperm donation plans and normalization processing of physical examination parameters, so that the precision and accuracy of sperm quality evaluation of the evaluation management module are improved, and a more comprehensive and more accurate sperm evaluation result is provided.

Description

Technical Field

[0001] The present invention belongs to the technical field of information management, and in particular relates to a sperm bank sperm supply management system, method and storage medium. Background Art

[0002] Human sperm banks, with the goal of treating infertility and preventing genetic and sexually transmitted diseases, utilize ultra-low temperature freezing technology to collect, test, preserve, and provide sperm. Currently, sperm bank management relies primarily on manual labor, but sperm bank management involves multiple steps, including donor screening, semen collection, storage, testing, provision, and follow-up. This not only increases staff workload but can also reduce overall operational efficiency. Furthermore, manual operations are slow when querying and compiling statistical data, making it difficult to provide accurate data in a timely manner, impacting decision-making and management efficiency.

[0003] Chinese patent CN112992323B discloses a sperm bank sperm donation management system, including a reception module, a freezing module, a file management module and a supply module; the reception module is used to enter the identity information and physical examination information of first-time sperm donors, identify the identities of repeat sperm donors, encrypt the identity information of sperm donors for the first time, and calculate the amount of sperm donation subsidies; the freezing module is used to encrypt the identity information of sperm donors after the first encryption for the second time, record the semen analysis data of sperm donors and information on semen in and out of the storage; the file management module is used to manage the electronic files of qualified sperm donors and record feedback information from sperm-using institutions. However, the existing system's reception module is unable to identify and evaluate multi-source heterogeneous physical examination information from multiple dimensions when screening the physical examination information of sperm donors, which reduces the efficiency and accuracy of sperm quality testing and affects the efficiency and success rate of sperm donation. To address the above problems, we propose a sperm bank sperm donation management system, method and storage medium. Summary of the Invention

[0004] The purpose of the present invention is to address the shortcomings of the existing technology and provide a sperm bank sperm donation management system, method and storage medium, which solves the problem that the existing system's reception module is unable to identify and evaluate multi-source heterogeneous physical examination information from multiple dimensions when screening the physical examination information of sperm donors, thereby reducing the efficiency and accuracy of sperm quality detection.

[0005] The present invention is implemented as follows: a sperm bank sperm supply management system, the sperm bank sperm supply management system comprising:

[0006] The user end is used to obtain basic information of sperm donor volunteers, customize sperm donation plans based on the basic information of sperm donor volunteers, and obtain multi-source heterogeneous physical examination parameters of volunteers;

[0007] The information processing module uses edge computing to normalize multi-source heterogeneous physical examination parameters, generate a standard parameter set, upload the standard parameter set to the sperm database, pre-build a sperm quality assessment model, traverse the sperm database, and the sperm database issues a modeling dataset. The sperm quality assessment model is iteratively trained using the modeling dataset until convergence, and the sperm quality assessment model is output;

[0008] The evaluation management module loads the standard parameter set in real time, executes the sperm quality evaluation model with the standard parameter set as input, outputs the sperm quality evaluation value, determines whether the sperm quality evaluation value meets the preset quality threshold, and if the sperm quality evaluation value meets the preset quality threshold, assigns the sperm quality evaluation value to the volunteer identification code, and uploads the sperm quality evaluation value and the volunteer identification code to the sperm database;

[0009] The supply management module is used to obtain sperm demand information, trigger sperm supply instructions based on the sperm demand information and sperm quality assessment value, generate sperm supply strategies based on genetic algorithms in response to sperm supply instructions, and upload the sperm supply strategies to the sperm database.

[0010] The sperm database is used to store the basic information of sperm donation volunteers, standard parameter sets, sperm quality assessment values, volunteer identification codes, sperm demand information and sperm supply strategies, and to enter, query, update and delete the basic information of sperm donation volunteers, standard parameter sets, sperm quality assessment values ​​and volunteer identification codes.

[0011] Preferably, the user terminal includes:

[0012] An information collection unit is used to collect basic information of sperm donor volunteers and create a volunteer identification code based on the basic information of sperm donor volunteers;

[0013] The plan customization unit is used to obtain the basic information of sperm donation volunteers and customize the volunteer sperm donation plan based on the basic information of sperm donation volunteers;

[0014] The parameter indexing unit traverses the sperm database based on the volunteer identification code, obtains the multi-source heterogeneous physical examination parameters of the volunteers, and standardizes the multi-source heterogeneous physical examination parameters of the volunteers.

[0015] Preferably, the method for customizing a sperm donation plan for a sperm donor based on the basic information of the sperm donor specifically includes:

[0016] The volunteer identification code is sent to the information collection unit, which verifies the volunteer identification code. If the code matches, the verification is successful, and the basic information of the sperm donor is obtained from the information collection unit. The basic information of the sperm donor includes name, age, contact information, height, weight, smoking and drinking habits, history of exposure to harmful substances, past medical history, education, and work information.

[0017] Based on the Delphi method, the basic information of sperm donors is weighted to obtain the basic weight set. The Hit-att network is used as the basic feature extractor to extract features from the basic weight set and obtain the weight matrix corresponding to the basic attributes in the basic weight set.

[0018] Load the basic weight set, normalize the attention weights of the basic attributes based on the softmax function, and obtain the attribute normalization coefficient;

[0019] The attribute normalization coefficient is calculated using the following formula:

[0020]

[0021] e j =tanh(W j h e +b j )(2)

[0022] Among them, α j is the attribute normalization coefficient, e j The attention vector of the j-th attribute, W j Represents the weight matrix input representation of the basic attributes, h e is the input representation of the basic weighted set, b j is the bias vector;

[0023] Obtain the attribute normalization coefficient corresponding to the basic weighted set, use the cross-fusion method to weight the basic attributes in the basic weighted set, and introduce mask coefficients for the basic attributes missing in the basic weighted set;

[0024] The output vector of the weighted fusion of basic attributes and the normalization coefficient are spliced ​​together to output the overall attribute representation vector, and the target pass rate is calculated;

[0025] The target pass rate is calculated using the following formula:

[0026]

[0027] Among them, C i,j is the target qualified rate of the output, A is the number of basic attributes j, is the overall attribute representation vector, mask(j) represents the mask coefficient;

[0028] Obtaining a target pass rate, and determining whether the target pass rate meets the pass threshold based on a preset pass threshold;

[0029] If the qualified threshold is met, the corresponding sperm donation volunteer is determined to be qualified, and the abstinence period of the sperm donation volunteer is determined based on the target qualified rate;

[0030] Obtain the abstinence time of the sperm donation volunteers, trigger a plan generation template based on the abstinence time of the sperm donation volunteers, enter the abstinence time of the sperm donation volunteers into the plan generation template, and integrate to obtain a personalized sperm donation plan.

[0031] Preferably, the method for normalizing multi-source heterogeneous physical examination parameters using edge computing specifically includes:

[0032] Load multi-source heterogeneous physical examination parameters, extract semen concentration, hepatitis B, hepatitis C, HIV, syphilis screening parameters, and chromosome examination parameters from the multi-source heterogeneous physical examination parameters, and merge semen concentration, hepatitis B, hepatitis C, HIV, syphilis screening parameters, and chromosome examination parameters into a test data set;

[0033] Process missing values ​​and outliers in the test data set, decompose the test data set by Hilbert transform to obtain analytical data, and calculate the unilateral spectrum of the analytical data;

[0034] Among them, the one-sided spectrum is expressed as:

[0035]

[0036] f(d) is the unilateral spectrum output representation of the analytical data, h is the data dimension, x(d) represents the analytical data, χ is the data label coefficient, δ(x) represents the impulse function, and γ is the bandwidth of the impulse function;

[0037] Load the analytical data and the unilateral spectrum corresponding to the analytical data, reconstruct and update the analytical data based on Fourier transform combined with the unilateral spectrum corresponding to the analytical data, and obtain the assay reconstruction set;

[0038] The data label coefficient is used as a quadratic penalty term, and the Lagrange multiplier is introduced to perform weighted summation on the assay reconstruction set. The weighted summation result is used as the prior coefficient of the sperm quality assessment model for sperm image analysis.

[0039] Prior coefficients and sperm image information from multi-source heterogeneous physical examination parameters are loaded and integrated into a standard parameter set.

[0040] Preferably, the method of iteratively training the sperm quality assessment model using the modeling data set until convergence specifically includes:

[0041] The convolutional neural network model EfficientNetV2 is used as the backbone network of the sperm quality assessment model. The backbone network introduces an attention mechanism, a priori module based on the Bayesian algorithm, and a multilayer perceptron. The multilayer perceptron consists of a set of input layers, a set of output layers, and three sets of hidden layers. Each layer is nested with a Gelu function as the activation function. The input size of the EfficientNetV2 backbone network is 112*112, and the number of channels is 128.

[0042] Obtain the modeling dataset, read the modeling dataset and its corresponding prior coefficients, divide the modeling dataset into a training set and a test set, and set the training rounds, learning rate, and loss function of the sperm quality assessment model;

[0043] The training set is grayscale processed and converted into a grayscale image. The image size is adjusted to 112*112. The training set is used as input to perform iterative training of the sperm quality assessment model. Each time a sperm quality assessment model is trained, the recognition accuracy and loss value corresponding to the sperm quality assessment model are output;

[0044] Obtain a test set, use the test set to test the sperm quality assessment model, and output the test accuracy and test loss value corresponding to the test set;

[0045] The test accuracy and test loss value corresponding to the test set are compared with the recognition accuracy and loss value corresponding to the training set. The sperm quality assessment model with the highest recognition accuracy is saved as the final model, and the converged sperm quality assessment model is output.

[0046] Preferably, the method for executing the sperm quality assessment model using the standard parameter set as input specifically comprises:

[0047] Obtain a standard parameter set, and use the EfficientNetV2 backbone network to extract features from the standard parameter set, outputting a feature extraction set containing sperm morphology features, sperm quantity features, and sperm concentration features;

[0048] The sperm morphology feature vector is weighted based on the attention mechanism, and the weighted morphology weighted vector is input into the multi-layer perceptron, which calculates the sperm motility value;

[0049] The sperm motility value is calculated by the following formula:

[0050]

[0051] Among them, h j represents the sperm motility value, m is the sperm motility coefficient, v0 is the sperm movement rate, Δt represents the shooting time interval, and σ represents the number of random hyperplanes of the attention mechanism local sensitive hash function;

[0052] The prior module based on the Bayesian algorithm assigns the prior coefficient to the feature extraction set, combines the hidden layer of the multi-layer perceptron to comprehensively evaluate the sperm quality, outputs the sperm quality assessment value, and assigns the sperm quality assessment value to the volunteer identification code;

[0053]

[0054] z i is the sperm quality assessment value, j con is the sperm concentration, n is the number of sperm, are the input representation and mean of the feature extraction set, respectively.

[0055] Preferably, the supply management module specifically includes:

[0056] A demand acquisition unit is used to obtain sperm demand information and trigger a sperm use instruction based on the sperm demand information and sperm quality assessment value;

[0057] The instruction response unit generates an insemination strategy based on a genetic algorithm in response to the sperm use instruction, and uploads the insemination strategy to the sperm database.

[0058] On the other hand, the present invention also provides a sperm bank sperm supply management method, the sperm bank sperm supply management method specifically comprising:

[0059] Obtain basic information of sperm donor volunteers, customize sperm donation plans based on the basic information of sperm donor volunteers, and obtain multi-source heterogeneous physical examination parameters of volunteers;

[0060] Edge computing is used to normalize multi-source heterogeneous physical examination parameters to generate a standard parameter set, which is uploaded to the sperm database. A sperm quality assessment model is pre-built, the sperm database is traversed, and the sperm database issues a modeling dataset. The sperm quality assessment model is iteratively trained using the modeling dataset until convergence, and the sperm quality assessment model is output.

[0061] Load the standard parameter set in real time, use the standard parameter set as input, execute the sperm quality assessment model, and output the sperm quality assessment value;

[0062] Determine whether the sperm quality assessment value meets the preset quality threshold;

[0063] If the sperm quality assessment value meets the preset quality threshold, the sperm quality assessment value is assigned to the volunteer identification code, and the sperm quality assessment value and the volunteer identification code are uploaded to the sperm database;

[0064] Acquire sperm demand information, trigger sperm use instructions based on the sperm demand information and sperm quality assessment value, generate sperm supply strategies based on genetic algorithms in response to the sperm use instructions, and upload the sperm supply strategies to the sperm database.

[0065] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed, the sperm bank sperm supply management method is implemented.

[0066] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0067] In an embodiment of the present invention, a user terminal is provided for customizing sperm donation plans of sperm donors based on their basic information, and an information processing module is provided for normalizing multi-source heterogeneous physical examination parameters using edge computing and pre-building a sperm quality assessment model. The user terminal and the information processing module cooperate to realize the initial screening of sperm donors, personalized customization of sperm donation plans, and normalization of physical examination parameters, thereby improving the precision and accuracy of sperm quality assessment in the assessment management module, making it possible to comprehensively assess sperm quality from multiple dimensions, and providing more comprehensive and accurate sperm assessment results. This overcomes the problem that the existing system's reception module cannot identify and assess multi-source heterogeneous physical examination information from multiple dimensions when screening the physical examination information of sperm donors, thereby reducing the efficiency and accuracy of sperm quality detection.

[0068] In an embodiment of the present invention, the basic information of sperm donor volunteers is weighted based on the Delphi method, and the basic weighted set is extracted with the Hit-att network. At the same time, the attention weights of the basic attributes are normalized using the softmax function. This enables an accurate assessment of the target qualification rate based on the basic information of sperm donor volunteers, provides decision support for the customization of sperm donation plans and the initial screening of volunteers, and improves the efficiency of sperm bank sperm donation management.

[0069] In this embodiment of the present invention, edge computing is used to normalize multi-source heterogeneous physical examination parameters, thereby providing a priori coefficients for the sperm quality assessment model to evaluate sperm images. Furthermore, Lagrange multipliers are introduced to weight the sum of the assay reconstruction set. This weighted summation serves as the a priori coefficient for the sperm quality assessment model to analyze sperm images, thus providing an objective and quantitative standard for sperm quality assessment and helping to reduce human error and subjective bias. Furthermore, the weight distribution and assessment model can be adjusted as needed to accommodate different application scenarios and requirements.

[0070] In an embodiment of the present invention, the sperm quality assessment model is built on the EfficientNetV2 backbone network, and an attention mechanism, a priori module based on the Bayesian algorithm, and a multi-layer perceptron are introduced into the backbone network. This effectively integrates the prior coefficients with sperm image information, improves the accuracy of sperm quality identification, and can extract useful information from complex sperm images, providing a solid foundation for subsequent analysis and evaluation. Through convolution operations at different stages and techniques such as depthwise separable convolution, it effectively captures the multi-scale features in the image. The multi-layer perceptron has a strong nonlinear mapping capability and can fit complex functional relationships, which is very important for complex tasks such as sperm quality assessment. It can map the input feature space to the output quality grade space, achieving accurate classification and regression prediction of sperm quality.

[0071] In an embodiment of the present invention, a sperm quality assessment model comprehensively calculates a sperm quality assessment value based on sperm morphological characteristics, sperm count characteristics, sperm concentration characteristics, and sperm motility values. By assessing the sperm count of sperm donors, it can be determined whether they have an adequate sperm supply. By comprehensively calculating the sperm quality assessment value, an objective evaluation index can be provided for sperm banks, helping them to better screen and manage sperm samples from sperm donors. This helps to improve the overall quality and reputation of sperm banks and meet more sperm donation requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 It is a structural diagram of the sperm bank sperm supply management system provided by the present invention.

[0073] Figure 2 The present invention shows a schematic diagram of the implementation process of a method for customizing a sperm donation plan for a sperm donor based on the basic information of the sperm donor.

[0074] Figure 3 The figure shows a schematic diagram of the implementation process of the method for normalizing multi-source heterogeneous physical examination parameters using edge computing.

[0075] Figure 4 A schematic diagram of the implementation process of a method for iteratively training a sperm quality assessment model to convergence using a modeling dataset is shown.

[0076] Figure 5 A schematic diagram of the sperm quality assessment model method is shown with a standard parameter set as input.

[0077] Figure 6 The figure shows a schematic diagram of the implementation process of the sperm bank sperm supply management method.

[0078] In the figure: 100-user terminal, 110-information collection unit, 120-plan customization unit, 130-parameter index unit, 200-information processing module, 300-evaluation management module, 400-supply management module, 410-demand acquisition unit, 420-command response unit, 500-sperm database. DETAILED DESCRIPTION

[0079] Unless otherwise defined, all technical and scientific terms used herein have the same meanings as commonly understood by those skilled in the art to which this application belongs. The terms used in the specification of the application are only for the purpose of describing specific embodiments and are not intended to limit this application. The terms "including" and "having" and any variations thereof in the specification and claims of this application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of this application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0080] The existing system's reception module cannot identify and evaluate multi-source heterogeneous physical examination information from multiple dimensions when screening the physical examination information of sperm donors, which reduces the efficiency and accuracy of sperm quality detection. In response to the above problems, we proposed a sperm bank sperm supply management system, method and storage medium. The sperm bank sperm supply management system consists of a user terminal 100, an information processing module 200, an evaluation management module 300, a supply management module 400 and a sperm database 500. When the sperm bank sperm supply management system is working, the user terminal 100 first obtains the basic information of the sperm donor volunteers, customizes the volunteer sperm donation plan based on the basic information of the sperm donor volunteers, obtains the multi-source heterogeneous physical examination parameters of the volunteers, and then the information processing module 200 is used to process the sperm donation information. The processing module 200 uses edge computing to normalize multi-source heterogeneous physical examination parameters, uses the modeling data set to iteratively train the sperm quality assessment model until convergence, and outputs the sperm quality assessment model. Then the assessment management module 300 uses the standard parameter set as input, executes the sperm quality assessment model, outputs the sperm quality assessment value, assigns the sperm quality assessment value to the volunteer identification code, and uploads the sperm quality assessment value and the volunteer identification code to the sperm database 500. Finally, the supply management module 400 triggers the sperm use instruction based on the sperm demand information and the sperm quality assessment value. In response to the sperm use instruction, it generates a sperm supply strategy based on the genetic algorithm and uploads the sperm supply strategy to the sperm database 500. In an embodiment of the present invention, a user terminal 100 is provided for customizing sperm donation plans of sperm donors based on their basic information, and an information processing module 200 is provided for normalizing multi-source heterogeneous physical examination parameters using edge computing and pre-building a sperm quality assessment model. The user terminal 100 and the information processing module 200 cooperate to realize the initial screening of sperm donors, personalized customization of sperm donation plans, and normalization of physical examination parameters, thereby improving the precision and accuracy of the sperm quality assessment of the assessment management module 300, making it possible to comprehensively assess sperm quality from multiple dimensions, providing more comprehensive and accurate sperm assessment results, and overcoming the problem that the existing system's reception module cannot identify and assess multi-source heterogeneous physical examination information from multiple dimensions when screening the physical examination information of sperm donors, thereby reducing the efficiency and accuracy of sperm quality detection.

[0081] The embodiment of the present invention provides a sperm bank sperm supply management system, Figure 1 The schematic diagram shows the sperm bank sperm supply management system, which specifically includes:

[0082] The user terminal 100 is used to obtain basic information of sperm donor volunteers, customize sperm donation plans based on the basic information of sperm donor volunteers, and obtain multi-source heterogeneous physical examination parameters of volunteers;

[0083] The information processing module 200 uses edge computing to normalize multi-source heterogeneous physical examination parameters, generates a standard parameter set, uploads the standard parameter set to the sperm database 500, pre-builds a sperm quality assessment model, traverses the sperm database 500, and the sperm database 500 issues a modeling dataset. The sperm quality assessment model is iteratively trained using the modeling dataset until convergence, and the sperm quality assessment model is output;

[0084] The evaluation management module 300 loads the standard parameter set in real time, executes the sperm quality evaluation model with the standard parameter set as input, outputs a sperm quality evaluation value, determines whether the sperm quality evaluation value meets a preset quality threshold, and if so, assigns the sperm quality evaluation value to a volunteer identification code, and uploads the sperm quality evaluation value and the volunteer identification code to the sperm database 500;

[0085] The supply management module 400 is used to obtain sperm demand information, trigger sperm use instructions based on the sperm demand information and sperm quality assessment values, generate sperm supply strategies based on genetic algorithms in response to the sperm use instructions, and upload the sperm supply strategies to the sperm database 500.

[0086] The sperm database 500 is used to store basic information of sperm donors, standard parameter sets, sperm quality assessment values, volunteer identification codes, sperm demand information, and sperm supply strategies, and to input, query, update, and delete basic information of sperm donors, standard parameter sets, sperm quality assessment values, and volunteer identification codes.

[0087] In this embodiment, the sperm database 500 uses the Microsoft SQL Server 2019 database. The application of this database provides important technical support for the later system database development, and the sperm database 500 can extract data from related hospital information systems such as EHR, HIS, PACS, etc. The sperm database 500 can store multi-source heterogeneous data such as text files, web-josh, XML, tabular data, and images.

[0088] In this embodiment, when the sperm bank sperm supply management system is working, first the user terminal 100 obtains the basic information of the sperm donation volunteers, customizes the volunteer sperm donation plan based on the basic information of the sperm donation volunteers, and obtains the multi-source heterogeneous physical examination parameters of the volunteers. Then the information processing module 200 uses edge computing to normalize the multi-source heterogeneous physical examination parameters, uses the modeling data set to iteratively train the sperm quality assessment model until convergence, and outputs the sperm quality assessment model. Then the assessment management module 300 uses the standard parameter set as input, executes the sperm quality assessment model, outputs the sperm quality assessment value, assigns the sperm quality assessment value to the volunteer identification code, and uploads the sperm quality assessment value and the volunteer identification code to the sperm database 500. Finally, the supply management module 400 triggers the sperm use instruction in combination with the sperm demand information and the sperm quality assessment value. In response to the sperm use instruction, it generates a sperm supply strategy based on the genetic algorithm and uploads the sperm supply strategy to the sperm database 500.

[0089] In an embodiment of the present invention, a user terminal 100 is provided for customizing sperm donation plans of sperm donors based on their basic information, and an information processing module 200 is provided for normalizing multi-source heterogeneous physical examination parameters using edge computing and pre-building a sperm quality assessment model. The user terminal 100 and the information processing module 200 cooperate to realize the initial screening of sperm donors, personalized customization of sperm donation plans, and normalization of physical examination parameters, thereby improving the precision and accuracy of the sperm quality assessment of the assessment management module 300, making it possible to comprehensively assess sperm quality from multiple dimensions, providing more comprehensive and accurate sperm assessment results, and overcoming the problem that the existing system's reception module cannot identify and assess multi-source heterogeneous physical examination information from multiple dimensions when screening the physical examination information of sperm donors, thereby reducing the efficiency and accuracy of sperm quality detection.

[0090] In this embodiment, the user terminal 100 includes:

[0091] The information collection unit 110 is used to collect basic information of sperm donor volunteers and create a volunteer identification code based on the basic information of sperm donor volunteers;

[0092] A plan customization unit 120 is used to obtain basic information of sperm donation volunteers and customize a sperm donation plan for the volunteers based on the basic information of the sperm donation volunteers;

[0093] The parameter indexing unit 130 traverses the sperm database 500 based on the volunteer identification code, obtains the multi-source heterogeneous physical examination parameters of the volunteer, and standardizes the multi-source heterogeneous physical examination parameters of the volunteer.

[0094] Meanwhile, the supply management module 400 specifically includes:

[0095] The demand acquisition unit 410 is used to acquire sperm demand information and trigger a sperm use instruction based on the sperm demand information and the sperm quality assessment value;

[0096] The instruction response unit 420 generates an insemination strategy based on a genetic algorithm in response to the sperm use instruction, and uploads the insemination strategy to the sperm database 500 .

[0097] In this embodiment, the goal of generating a sperm supply strategy based on a genetic algorithm is to maximize the overall pregnancy success rate, minimize costs, or maximize a specific benefit, while meeting the fertilization requirements. For example, if the goal is to maximize the pregnancy success rate, then the objective function can be defined as a function related to the pregnancy success rate, which may be related to factors such as the sperm quality assessment value and the required quantity in the sperm demand information. Each individual in the initial population is evaluated based on the objective function and its fitness value is calculated. The fitness value is used to measure the degree to which an individual is good at solving a problem. For individuals with a sperm supply strategy, the fitness value can be calculated based on the simulation results of the pregnancy success rate under the corresponding sperm supply strategy or other relevant indicators.

[0098] It should be noted that the user terminal 100 can be a laptop computer, a personal digital assistant (PDA), a mobile phone or other communication device. At the same time, sperm donation volunteers can automatically register and log in through the user terminal 100. After verification, the user terminal 100 generates a volunteer identification code based on hash encryption. Generating the volunteer identification code based on hash encryption can ensure the privacy of the volunteer and avoid the risk of information leakage. At the same time, the basic information of the sperm donation volunteers can be automatically collected and obtained through the information collection unit 110, which can reduce the burden of manual information entry and improve the efficiency of initial screening. The plan customization unit 120 can customize the volunteer sperm donation plan based on the basic information of the sperm donation volunteers. It does not require subjective judgment by medical staff and sperm donation volunteers to determine whether the sperm donation volunteers meet the initial screening standards, and can customize the sperm donation plan based on the physical condition of the sperm donation volunteers, thereby facilitating the efficiency of sperm donation collection.

[0099] The embodiment of the present invention provides a method for customizing a sperm donation plan for a sperm donor based on the basic information of the sperm donor. Figure 2 The following is a schematic diagram of a method for customizing a sperm donation plan based on the basic information of sperm donors. The method specifically includes:

[0100] Step S101: Send the volunteer identification code to the information collection unit 110. The information collection unit 110 verifies the volunteer identification code. If the code matches, the verification is successful, and the basic information of the sperm donation volunteer is obtained from the information collection unit 110.

[0101] It should be noted that the basic information of sperm donation volunteers includes but is not limited to name, age, contact information, height, weight, smoking and drinking habits, history of exposure to harmful substances, past medical history, education, and work information;

[0102] The basic information data of sperm donation volunteers includes many aspects, which are crucial to ensuring the quality and safety of sperm. By obtaining the height and weight of the volunteers and calculating the BMI (body mass index), their health status is assessed. The volunteers' smoking and drinking habits and history of exposure to harmful substances will affect sperm quality. Past medical history includes family history of genetic diseases, mental illness, infectious diseases, sexually transmitted diseases, hemophilia, and thalassemia.

[0103] Step S102: weighting the basic information of sperm donors based on the Delphi method to obtain a basic weighted set, using the Hit-att network as a basic feature extractor to extract features from the basic weighted set to obtain a weight matrix corresponding to the basic attributes in the basic weighted set;

[0104] In this embodiment, the Delphi method is a systematic method for collecting and synthesizing expert opinions. It solicits expert opinions through multiple rounds of anonymous questionnaires and gradually reaches a consensus. It is widely used in prediction, decision support, and social science research. When weighting the basic information of sperm donors, the weight of height and weight can be 1, the weight of smoking and drinking habits can be 3, the weight of harmful substance exposure history can be 4, and the weight of previous medical history can be 6. After the weighting is completed, the Hit-att network is used as the basic feature extractor to quickly extract the features of the basic attributes in the basic weighting set. The Hit-Att network (also known as the Hit-Attentive Network) is a neural network structure based on the attention mechanism, which is generally used to process long-distance dependencies in sequence data or image data. The Hit-Att network can be composed of an encoder and a decoder.

[0105] Step S103: load the basic weighted set, normalize the attention weights of the basic attributes based on the softmax function, and obtain the attribute normalization coefficient;

[0106] The attribute normalization coefficient is calculated using the following formula:

[0107]

[0108] e j =tanh(W j h e +b j )(2)

[0109] Among them, α j is the attribute normalization coefficient, e j The attention vector of the j-th attribute, W j Represents the weight matrix input representation of the basic attributes, h e is the input representation of the basic weighted set, bj is the bias vector;

[0110] Step S104: obtaining attribute normalization coefficients corresponding to the basic weighted set, weighting and fusing the basic attributes in the basic weighted set using a cross-fusion method, and introducing mask coefficients for the basic attributes missing from the basic weighted set;

[0111] In this embodiment of the present invention, by introducing masking coefficients for the confirmed basic attributes, it is possible to avoid directly deleting these records or using simple interpolation methods that may cause information loss. By introducing masking coefficients, it is also possible to mark which attributes are missing, thereby maintaining data integrity in subsequent processing.

[0112] Step S105: The output vector of the weighted fusion of basic attributes and the normalization coefficient are spliced ​​and output as the overall attribute representation vector, and the target pass rate is calculated;

[0113] The target pass rate is calculated using the following formula:

[0114]

[0115] Among them, C i,j is the target qualified rate of the output, A is the number of basic attributes j, is the overall attribute representation vector, mask(j) represents the mask coefficient;

[0116] Step S106, obtaining a target qualified rate, and determining whether the target qualified rate meets the qualified threshold based on a preset qualified threshold. In this embodiment, the qualified threshold may be 0.7-0.75;

[0117] Step S107: If the qualified threshold is met, the corresponding sperm donor is determined to be qualified, and the abstinence period of the sperm donor is determined based on the target qualified rate;

[0118] Step S108: If the qualified threshold is not met, the corresponding sperm donation volunteer is determined to be unqualified and the initial screening ends.

[0119] Step S109, obtaining the abstinence time of the sperm donation volunteer, triggering a plan generation template based on the abstinence time of the sperm donation volunteer, entering the abstinence time of the sperm donation volunteer into the plan generation template, and integrating to obtain a personalized sperm donation plan.

[0120] In an embodiment of the present invention, the basic information of sperm donor volunteers is weighted based on the Delphi method, and the basic weighted set is extracted with the Hit-att network. At the same time, the attention weights of the basic attributes are normalized using the softmax function. This enables an accurate assessment of the target qualification rate based on the basic information of sperm donor volunteers, provides decision support for the customization of sperm donation plans and the initial screening of volunteers, and improves the efficiency of sperm bank sperm donation management.

[0121] The embodiment of the present invention provides a method for normalizing multi-source heterogeneous physical examination parameters using edge computing. Figure 3 The following is a schematic diagram of the implementation process of a method for normalizing multi-source heterogeneous physical examination parameters using edge computing. The method for normalizing multi-source heterogeneous physical examination parameters using edge computing specifically includes:

[0122] Step S201, loading multi-source heterogeneous physical examination parameters, extracting semen concentration, hepatitis B, hepatitis C, HIV, syphilis screening parameters, and chromosome examination parameters from the multi-source heterogeneous physical examination parameters, and merging the semen concentration, hepatitis B, hepatitis C, HIV, syphilis screening parameters, and chromosome examination parameters into a test data set;

[0123] Step S202 , processing missing values ​​and outliers on the test data set, decomposing the test data set by Hilbert transform to obtain analytical data, and calculating the unilateral spectrum of the analytical data;

[0124] In this embodiment of the present invention, considering that physical examination parameters come from different systems, different medical examination institutions, and hospital data systems, and that these multi-source heterogeneous physical examination parameters are important indicators and prior knowledge for assessing sperm quality, the Hilbert transform is used to decompose the test data set to obtain analytical data. The single-sided spectrum of the analytical data is then calculated, thereby converting the real signal into a complex signal, enabling deeper analysis on the complex plane. This conversion not only preserves all the information of the original signal but also provides additional phase characteristic information.

[0125] Among them, the one-sided spectrum is expressed as:

[0126]

[0127] f(d) is the unilateral spectrum output representation of the parsed data, h is the data dimension. In this embodiment, the data dimension is 1-6, x(d) represents the parsed data, χ is the data label coefficient, which refers to the weight of the source data corresponding to the data. The label coefficient can be 0.1-0.8, δ(x) represents the impulse function, and γ is the bandwidth of the impulse function;

[0128] Step S203, loading the analytical data and the unilateral spectrum corresponding to the analytical data, reconstructing and updating the analytical data based on Fourier transform combined with the unilateral spectrum corresponding to the analytical data, and obtaining an assay reconstruction set;

[0129] Step S204: using the data label coefficient as a quadratic penalty term, introducing a Lagrange multiplier, and weighted summing the assay reconstruction set. The weighted summation result is used as the prior coefficient of the sperm quality assessment model for sperm image analysis.

[0130] It should be noted that in sperm image analysis, the prior coefficient can be used as one of the inputs of the evaluation model to help the model more accurately identify and classify sperm images, thereby improving the accuracy and reliability of sperm quality assessment.

[0131] Step S205 , loading the a priori coefficients and sperm image information in the multi-source heterogeneous physical examination parameters and integrating them into a standard parameter set.

[0132] In this embodiment of the present invention, edge computing is used to normalize multi-source heterogeneous physical examination parameters, thereby providing a priori coefficients for the sperm quality assessment model to evaluate sperm images. Furthermore, Lagrange multipliers are introduced to weight the sum of the assay reconstruction set. This weighted summation serves as the a priori coefficient for the sperm quality assessment model to analyze sperm images, thus providing an objective and quantitative standard for sperm quality assessment and helping to reduce human error and subjective bias. Furthermore, the weight distribution and assessment model can be adjusted as needed to accommodate different application scenarios and requirements.

[0133] The embodiment of the present invention provides a method for iteratively training a sperm quality assessment model to convergence using a modeling data set. Figure 4 The present invention shows a schematic diagram of a method for iteratively training a sperm quality assessment model to convergence using a modeling dataset. The method for iteratively training a sperm quality assessment model to convergence using a modeling dataset specifically includes:

[0134] Step S301, using the convolutional neural network model EfficientNetV2 as the backbone network of the sperm quality assessment model, and introducing an attention mechanism, a priori module based on the Bayesian algorithm, and a multilayer perceptron into the backbone network, wherein the multilayer perceptron includes a set of input layers, a set of output layers, and three sets of hidden layers, and each layer is nested with a Ge Lu function as an excitation function. The input size of the EfficientNetV2 backbone network is 112*112, and the number of channels is 128;

[0135] It should be noted that in the Bayesian algorithm, the fusion of prior coefficients and features is achieved through Bayes' theorem. Specifically, Bayes' theorem allows us to calculate the posterior probability through the prior probability (i.e., the prior coefficient) and the likelihood function (the probability calculated based on the features and new data). This process is actually a dynamic update process, in which the prior coefficient is used as the initial estimate, and the new data is used to update this estimate to obtain a more accurate posterior probability. Fusion of the prior coefficients with features means that we have found a balance between utilizing existing knowledge and new data. This fusion not only improves the model's ability to predict unknown data, but also enables the model to better adapt to different application scenarios and data distributions. In addition, in this way, we can reduce our dependence on large amounts of training data and improve the efficiency and generalization ability of the model.

[0136] Step S302: Obtain a modeling data set, read the modeling data set and its corresponding prior coefficients, divide the modeling data set into a training set and a test set, set the training rounds, learning rate, and loss function of the sperm quality assessment model. In this embodiment, the training rounds of the sperm quality assessment model can be 100-500 times, the learning rate can be 0.0002, and the loss function can be the Huber loss function.

[0137] Step S303: grayscale processing is performed on the training set, and the grayscale image is converted into a grayscale image. The image size is adjusted to 112*112. The training set is used as input to perform iterative training of the sperm quality assessment model. Each time a sperm quality assessment model is trained, the recognition accuracy and loss value corresponding to the sperm quality assessment model are output;

[0138] Step S304: obtaining a test set, using the test set to test the sperm quality assessment model, and outputting the test accuracy and test loss value corresponding to the test set;

[0139] Step S305 : Compare the test accuracy and test loss value corresponding to the test set with the recognition accuracy and loss value corresponding to the training set, save the sperm quality assessment model with the highest recognition accuracy as the final model, and output the converged sperm quality assessment model.

[0140] In an embodiment of the present invention, the sperm quality assessment model is built on the EfficientNetV2 backbone network, and an attention mechanism, a priori module based on the Bayesian algorithm, and a multi-layer perceptron are introduced into the backbone network. This effectively integrates the prior coefficients with sperm image information, improves the accuracy of sperm quality identification, and can extract useful information from complex sperm images, providing a solid foundation for subsequent analysis and evaluation. Through convolution operations at different stages and techniques such as depthwise separable convolution, it effectively captures the multi-scale features in the image. The multi-layer perceptron has a strong nonlinear mapping capability and can fit complex functional relationships, which is very important for complex tasks such as sperm quality assessment. It can map the input feature space to the output quality grade space, achieving accurate classification and regression prediction of sperm quality.

[0141] The embodiment of the present invention provides a method for executing a sperm quality assessment model using a standard parameter set as input. Figure 5 A schematic diagram of a method for executing a sperm quality assessment model using a standard parameter set as input is shown. The method for executing a sperm quality assessment model using a standard parameter set as input specifically includes:

[0142] Step S401: Obtain a standard parameter set, extract features from the standard parameter set using the EfficientNetV2 backbone network, and output a feature extraction set containing sperm morphology features, sperm quantity features, and sperm concentration features;

[0143] Step S402: weighting the sperm morphology feature vector based on the attention mechanism, inputting the weighted morphology weighted vector into a multi-layer perceptron, and calculating the sperm motility value by the multi-layer perceptron;

[0144] The sperm motility value is calculated by the following formula:

[0145]

[0146] Among them, h j represents the sperm motility value, m is the sperm motility coefficient, v0 is the sperm movement rate, Δt represents the shooting time interval, σ represents the number of random hyperplanes of the attention mechanism local sensitive hash function. In this embodiment, the number of random hyperplanes can be 3-15, and the sperm motility coefficient is 1-3, where 1 represents the motion mode of immobility (IM), 2 represents the motion mode of slow or non-forward motion (NP), and 3 represents the motion mode of fast forward motion (PR).

[0147] Step S403: The a priori module based on the Bayesian algorithm assigns the a priori coefficient to the feature extraction set, combines the hidden layer of the multi-layer perceptron to comprehensively evaluate the sperm quality, outputs the sperm quality evaluation value, and assigns the sperm quality evaluation value to the volunteer identification code;

[0148]

[0149] z i is the sperm quality assessment value, j con is the sperm concentration, n is the number of sperm, are the input representation and mean of the feature extraction set, respectively.

[0150] In an embodiment of the present invention, a sperm quality assessment model comprehensively calculates a sperm quality assessment value based on sperm morphological characteristics, sperm count characteristics, sperm concentration characteristics, and sperm motility values. By assessing the sperm count of sperm donors, it can be determined whether they have an adequate sperm supply. By comprehensively calculating the sperm quality assessment value, an objective evaluation index can be provided for sperm banks, helping them to better screen and manage sperm samples from sperm donors. This helps to improve the overall quality and reputation of sperm banks and meet more sperm donation requirements.

[0151] The embodiment of the present invention provides a sperm bank sperm supply management method, Figure 6 The figure shows a schematic diagram of the implementation process of the sperm bank sperm supply management method, which specifically includes:

[0152] Step S10, obtaining basic information of sperm donor volunteers, customizing a sperm donation plan based on the basic information of sperm donor volunteers, and obtaining multi-source heterogeneous physical examination parameters of the volunteers;

[0153] Step S20: normalize the multi-source heterogeneous physical examination parameters using edge computing to generate a standard parameter set, upload the standard parameter set to the sperm database 500, pre-build a sperm quality assessment model, traverse the sperm database 500, and the sperm database 500 issues a modeling dataset. The sperm quality assessment model is iteratively trained using the modeling dataset until convergence, and the sperm quality assessment model is output;

[0154] Step S30, loading the standard parameter set in real time, executing the sperm quality assessment model with the standard parameter set as input, and outputting the sperm quality assessment value;

[0155] Step S40, determining whether the sperm quality assessment value meets a preset quality threshold;

[0156] Step S50: if the sperm quality assessment value meets the preset quality threshold, the sperm quality assessment value is assigned to the volunteer identification code, and the sperm quality assessment value and the volunteer identification code are uploaded to the sperm database 500;

[0157] Step S60 : If the sperm quality assessment value does not meet the preset quality threshold, the sperm quality assessment value is set to unqualified, and the volunteer identification code is uploaded to the sperm database 500 .

[0158] Step S70: obtaining sperm demand information, triggering a sperm use instruction based on the sperm demand information and the sperm quality assessment value, generating a sperm supply strategy based on a genetic algorithm in response to the sperm use instruction, and uploading the sperm supply strategy to the sperm database 500.

[0159] It is understandable that the sperm bank sperm supply management method provided in the embodiment of the present invention corresponds to the above-mentioned sperm bank sperm supply management system. The explanation, examples, beneficial effects, etc. of its relevant contents can refer to the corresponding contents in the sperm bank sperm supply management system and will not be repeated here.

[0160] On the other hand, an embodiment of the present invention also provides a computer device, which includes a display screen, a memory, a processor, and a computer program, wherein the memory stores the computer program, and when the computer program is executed by the processor, the processor executes the steps of the sperm bank sperm donation management method.

[0161] On the other hand, the present invention provides a schematic diagram of a readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the processor executes the steps of the sperm bank sperm donation management method.

[0162] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor. The processor is the control center of the terminal device and connects various parts of the entire user terminal using various interfaces and lines.

[0163] The memory can be used to store computer programs and / or modules. The processor implements the various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory and accessing the data stored in the memory. The memory can primarily include a program storage area and a data storage area. The program storage area can store an operating system, at least one application required for a function, and the like; the data storage area can store data generated based on the use of the sperm bank's sperm supply management system. Furthermore, the memory can include high-speed random access memory and non-volatile memory, such as a hard disk, internal memory, a plug-in hard disk, a smart memory card (SMC), a secure digital (SD) card, a flash memory card, at least one disk storage device, a flash memory device, or other volatile solid-state storage device.

[0164] Finally, it should be noted that the computer-readable storage medium (e.g., memory) herein may be a volatile memory or a non-volatile memory, or may include both a volatile memory and a non-volatile memory. By way of example and not limitation, the non-volatile memory may include a read-only memory (ROM), a programmable ROM (PROM), an electrically programmable ROM (EPROM), an electrically erasable programmable ROM (EEPROM), or a flash memory. The volatile memory may include a random access memory (RAM), which may act as an external cache memory. By way of example and not limitation, RAM may be obtained in a variety of forms, such as synchronous RAM (DRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct Rambus RAM (DRRAM). The storage devices of the disclosed aspects are intended to include, but are not limited to, these and other suitable types of memory.

[0165] In summary, the present invention provides a sperm bank sperm supply management system, method and storage medium. When the sperm bank sperm supply management system is working, the user terminal 100 first obtains the basic information of the sperm donation volunteers, customizes the volunteer sperm donation plan based on the basic information of the sperm donation volunteers, and obtains the multi-source heterogeneous physical examination parameters of the volunteers. Then, the information processing module 200 uses edge computing to normalize the multi-source heterogeneous physical examination parameters, uses the modeling data set to iteratively train the sperm quality assessment model until convergence, and outputs the sperm quality assessment model. Then, the assessment management module 300 uses the standard parameter set as input, executes the sperm quality assessment model, outputs the sperm quality assessment value, assigns the sperm quality assessment value to the volunteer identification code, and uploads the sperm quality assessment value and the volunteer identification code to the sperm database 500. Finally, the supply management module 400 triggers the sperm use instruction in combination with the sperm demand information and the sperm quality assessment value. In response to the sperm use instruction, it generates a sperm supply strategy based on the genetic algorithm and uploads the sperm supply strategy to the sperm database 500.

[0166] In an embodiment of the present invention, a user terminal 100 is provided for customizing sperm donation plans of sperm donors based on their basic information, and an information processing module 200 is provided for normalizing multi-source heterogeneous physical examination parameters using edge computing and pre-building a sperm quality assessment model. The user terminal 100 and the information processing module 200 cooperate to realize the initial screening of sperm donors, personalized customization of sperm donation plans, and normalization of physical examination parameters, thereby improving the precision and accuracy of the sperm quality assessment of the assessment management module 300, making it possible to comprehensively assess sperm quality from multiple dimensions, providing more comprehensive and accurate sperm assessment results, and overcoming the problem that the existing system's reception module cannot identify and assess multi-source heterogeneous physical examination information from multiple dimensions when screening the physical examination information of sperm donors, thereby reducing the efficiency and accuracy of sperm quality detection.

[0167] It should be noted that for the aforementioned embodiments, for simplicity of description, they are all expressed as a series of action combinations. However, those skilled in the art should be aware that the present invention is not limited by the order of the actions described, because according to the present invention, certain steps may be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in this specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0168] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the scope of protection of the invention. Obviously, the embodiments described are only some embodiments of the present invention, rather than all embodiments. Based on these embodiments, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in this field can still combine, add, delete or make other adjustments to the features in the various embodiments of the present invention according to the circumstances without conflict, without making creative work, so as to obtain different other technical solutions that do not deviate from the concept of the present invention in essence, and these technical solutions also fall within the scope of protection of the present invention.

Claims

1. A sperm bank sperm supply management system, characterized in that: The sperm bank sperm supply management system comprises: The user end is used to obtain basic information of sperm donation volunteers, customize sperm donation plans based on the basic information of sperm donation volunteers, and obtain multi-source heterogeneous physical examination parameters of volunteers; The information processing module uses edge computing to normalize multi-source heterogeneous physical examination parameters, generate a standard parameter set, upload the standard parameter set to the sperm database, pre-build a sperm quality assessment model, traverse the sperm database, and the sperm database sends a modeling data set. The sperm quality assessment model is iteratively trained using the modeling data set until convergence, and the sperm quality assessment model is output; The evaluation management module loads the standard parameter set in real time, takes the standard parameter set as input, executes the sperm quality evaluation model, outputs the sperm quality evaluation value, determines whether the sperm quality evaluation value meets the preset quality threshold, and if the sperm quality evaluation value meets the preset quality threshold, assigns the sperm quality evaluation value to the volunteer identification code, and uploads the sperm quality evaluation value and the volunteer identification code to the sperm database; The supply management module is used to obtain sperm demand information, trigger sperm use instructions based on the sperm demand information and sperm quality assessment value, generate sperm supply strategies based on genetic algorithms in response to sperm use instructions, and upload the sperm supply strategies to the sperm database.

2. The sperm bank sperm supply management system according to claim 1, characterized in that: The sperm bank sperm supply management system also includes: The sperm database is used to store the basic information of sperm donation volunteers, standard parameter sets, sperm quality assessment values, volunteer identification codes, sperm demand information and sperm supply strategies, and to enter, query, update and delete the basic information of sperm donation volunteers, standard parameter sets, sperm quality assessment values ​​and volunteer identification codes.

3. The sperm bank sperm supply management system according to claim 1, characterized in that: The user terminal comprises: An information collection unit, used to collect basic information of sperm donation volunteers and create a volunteer identification code based on the basic information of sperm donation volunteers; A plan customization unit is used to obtain basic information of sperm donation volunteers and customize sperm donation plans based on the basic information of sperm donation volunteers; The parameter indexing unit traverses the sperm database based on the volunteer identification code, obtains the multi-source heterogeneous physical examination parameters of the volunteers, and standardizes the multi-source heterogeneous physical examination parameters of the volunteers.

4. The sperm bank sperm supply management system according to claim 3, characterized in that: The method for customizing a sperm donation plan for a volunteer based on basic information of the sperm donation volunteer specifically includes: Send the volunteer identification code to the information collection unit, the information collection unit verifies the volunteer identification code, if it matches, the verification is passed, and the basic information of the sperm donation volunteer is obtained from the information collection unit, where the basic information of the sperm donation volunteer includes name, age, contact information, height, weight, smoking and drinking habits, history of exposure to harmful substances, past medical history, education, and work information; Based on the Delphi method, the basic information of sperm donation volunteers is weighted to obtain the basic weighted set. The Hit-att network is used as the basic feature extractor to extract features from the basic weighted set and obtain the weight matrix corresponding to the basic attributes in the basic weighted set. Load the basic weighted set, normalize the attention weights of the basic attributes based on the softmax function, and obtain the attribute normalization coefficient; Among them, the attribute normalization coefficient is calculated by the following formula: e j =tanh(W j h e +b j )(2) Among them, α j is the attribute normalization coefficient, e j The attention vector of the j-th attribute, W j The weight matrix input representation of the basic attributes, h e is the input representation of the basic weighted set, b j is the bias vector; Obtain the attribute normalization coefficient corresponding to the basic weighted set, use the cross-fusion method to weightedly fuse the basic attributes in the basic weighted set, and introduce mask coefficients for the missing basic attributes in the basic weighted set; The output vector of the weighted fusion of basic attributes and the normalization coefficient are spliced ​​and output as the overall attribute representation vector, and the target pass rate is calculated; Among them, the target pass rate is calculated by the following formula: Among them, C i,j is the target qualified rate of the output, A is the number of basic attributes j, is the overall attribute representation vector, mask(j) represents the mask coefficient; Obtaining a target qualified rate, and determining whether the target qualified rate meets the qualified threshold based on a preset qualified threshold; If the qualified threshold is met, the corresponding sperm donation volunteer is determined to be qualified, and the abstinence period of the sperm donation volunteer is determined based on the target qualified rate; The abstinence time of sperm donation volunteers is obtained, and a plan generation template is triggered based on the abstinence time of sperm donation volunteers. The abstinence time of sperm donation volunteers is entered into the plan generation template to integrate and obtain a personalized sperm donation plan.

5. The sperm bank sperm supply management system according to claim 1, characterized in that: The method for normalizing multi-source heterogeneous physical examination parameters using edge computing specifically includes: Load multi-source heterogeneous physical examination parameters, extract semen concentration, hepatitis B, hepatitis C, HIV, syphilis screening parameters, and chromosome examination parameters from the multi-source heterogeneous physical examination parameters, and merge semen concentration, hepatitis B, hepatitis C, HIV, syphilis screening parameters, and chromosome examination parameters into a test data set; Process missing values ​​and outliers on the test data set, decompose the test data set by Hilbert transform to obtain analytical data, and calculate the single-sided spectrum of the analytical data; Among them, the one-sided spectrum is expressed as: f(d) is the single-sided spectrum output representation of the analytical data, h is the data dimension, x(d) represents the analytical data, χ is the data label coefficient, δ(x) represents the impulse function, and γ is the bandwidth of the impulse function; Load the analytical data and the single-sided spectrum corresponding to the analytical data, reconstruct and update the analytical data based on Fourier transform combined with the single-sided spectrum corresponding to the analytical data, and obtain the assay reconstruction set; The data label coefficient is used as the quadratic penalty term, the Lagrange multiplier is introduced, and the weighted sum of the assay reconstruction set is taken as the prior coefficient of the sperm quality assessment model for sperm image analysis. Priori coefficients and sperm image information from multi-source heterogeneous physical examination parameters are loaded and integrated into a standard parameter set.

6. The sperm bank sperm supply management system according to claim 5, characterized in that: The method of iteratively training the sperm quality assessment model using the modeling data set until convergence specifically includes: The convolutional neural network model EfficientNetV2 is used as the backbone network of the sperm quality assessment model, and the attention mechanism, the prior module based on the Bayesian algorithm, and the multilayer perceptron are introduced into the backbone network. The multilayer perceptron includes a set of input layers, a set of output layers, and three sets of hidden layers. The layers are nested with Gelu functions as the excitation functions. The input size of the EfficientNetV2 backbone network is 112*112, and the number of channels is 128. Obtain the modeling data set, read the modeling data set and its corresponding prior coefficients, divide the modeling data set into a training set and a test set, and set the training rounds, learning rate, and loss function of the sperm quality assessment model; The training set is gray-processed and converted into a gray-scale image. The image size is adjusted to 112*112. The training set is used as input to perform iterative training of the sperm quality assessment model. Each time a sperm quality assessment model is trained, the recognition accuracy and loss value corresponding to the sperm quality assessment model are output; Obtain a test set, use the test set to test the sperm quality assessment model, and output the test accuracy and test loss value corresponding to the test set; The test accuracy and test loss value corresponding to the test set are compared with the recognition accuracy and loss value corresponding to the training set, and the sperm quality assessment model with the highest recognition accuracy is saved as the final model, and the converged sperm quality assessment model is output.

7. The sperm bank sperm supply management system according to claim 6, characterized in that: The method for executing the sperm quality assessment model using the standard parameter set as input specifically includes: Obtain the standard parameter set, and use the EfficientNetV2 backbone network to extract features from the standard parameter set, and output a feature extraction set containing sperm morphology features, sperm quantity features, and sperm concentration features; The sperm morphology feature vector is weighted based on the attention mechanism, and the weighted morphology weighted vector is input into the multi-layer perceptron, which calculates the sperm motility value; Among them, the sperm motility value is calculated by the following formula: Among them, h j represents the sperm motility value, m is the sperm motility coefficient, v0 is the sperm movement rate, Δt represents the shooting time interval, and σ represents the number of random hyperplanes of the local sensitive hash function of the attention mechanism; The prior module based on the Bayesian algorithm assigns the prior coefficient to the feature extraction set, combines the hidden layer of the multi-layer perceptron to comprehensively evaluate the sperm quality, outputs the sperm quality evaluation value, and assigns the sperm quality evaluation value to the volunteer identification code; z i is the sperm quality assessment value, j con is the sperm concentration, n is the number of sperm, x p , are the input representation and mean of the feature extraction set, respectively.

8. The sperm bank sperm supply management system according to claim 1, characterized in that: The supply management module specifically includes: A demand acquisition unit, used to acquire sperm demand information, and trigger a sperm use instruction in combination with the sperm demand information and sperm quality assessment value; The instruction response unit generates an insemination strategy based on a genetic algorithm in response to the sperm use instruction, and uploads the insemination strategy to the sperm database.

9. A sperm bank sperm supply management method, implemented by using the sperm bank sperm supply management system according to any one of claims 1 to 8, characterized in that: The sperm bank sperm supply management method specifically comprises: Obtain basic information of sperm donation volunteers, customize sperm donation plans based on the basic information of sperm donation volunteers, and obtain multi-source heterogeneous physical examination parameters of volunteers; Edge computing is used to normalize multi-source heterogeneous physical examination parameters, generate a standard parameter set, upload the standard parameter set to the sperm database, pre-build a sperm quality assessment model, traverse the sperm database, and the sperm database sends a modeling data set. The sperm quality assessment model is iteratively trained using the modeling data set until convergence, and the sperm quality assessment model is output; Load the standard parameter set in real time, use the standard parameter set as input, execute the sperm quality assessment model, and output the sperm quality assessment value; Determine whether the sperm quality assessment value meets the preset quality threshold; If the sperm quality assessment value meets the preset quality threshold, the sperm quality assessment value is assigned to the volunteer identification code, and the sperm quality assessment value and the volunteer identification code are uploaded to the sperm database; The sperm demand information is obtained, and the sperm demand information and sperm quality assessment value are combined to trigger the sperm use instruction. In response to the sperm use instruction, a sperm supply strategy is generated based on a genetic algorithm, and the sperm supply strategy is uploaded to the sperm database.

10. A computer-readable storage medium, characterized in that: Computer program instructions are stored, and when the computer program instructions are executed, the sperm bank sperm supply management method according to claim 9 is implemented.

Images