A method, device and computer-readable medium for screening mesenchymal stem cells

Abstract

The present invention provides a method, device, and computer-readable medium for screening mesenchymal stem cells. The method first collects microscopic images of mesenchymal stem cells from each sample, with the microscopic images being accompanied by sample numbers and timestamps. The method then removes invalid areas from the microscopic images and performs differential calculations on valid areas to generate a preprocessed image. The preprocessed image is then divided into multiple image blocks, and the image information entropy of the image blocks is calculated. The image information entropy is then calculated by aggregation, and a time series sample of the image information entropy and the timestamp of the preprocessed image is constructed. The time series sample is then assigned an image category, input into a prediction model, and a prediction result is calculated. Finally, the cell samples are screened based on the prediction result. This method allows the final state of stem cells to be assessed in their early stages, thereby pre-screening samples with better conditions, effectively avoiding waste during the culture process, reducing labor and material resources, and improving experimental accuracy.

Description

Technical field

[0001] The present invention relates to the field of biotechnology, and in particular to a method and device for screening mesenchymal stem cells and a computer-readable medium. [Background Technology]

[0002] Mesenchymal stem cells (MSCs) are a type of multipotent stem cell derived from adult tissues that can differentiate into a variety of cell types, such as muscle cells, neurons, and cartilage cells. They are found in many different tissues, including bone marrow, adipose tissue, placenta, liver, and lung. These cells can be cultured and expanded in vitro from tissue samples for further research and therapeutic applications.

[0003] The cultivation of mesenchymal stem cells is a challenging process, with the following technical difficulties:

[0004] 1. Maintaining stable cell growth and proliferation. This requires optimizing the formulation of cell culture medium, including adding appropriate growth factors and nutrients, and controlling culture conditions such as temperature, oxygen levels, and CO2 levels.

[0005] 2. Maintain cell pluripotency. Mesenchymal stem cells tend to lose their stem cell properties during in vitro culture, so it is necessary to control culture conditions and time to maintain their pluripotency and differentiation ability.

[0006] 3. During cell culture, attention should be paid to cell purity and sterility to avoid problems such as cell contamination and cell death. This can be achieved by selecting appropriate cell culture media and disinfection measures.

[0007] Therefore, a large number of samples in poor condition will appear in the culture process of mesenchymal stem cells, and these will eventually be eliminated. If a method can be invented to evaluate the final state of stem cells based on their early state, samples with better conditions can be screened out in advance, effectively avoiding waste in the culture process, reducing the consumption of manpower and material resources, and improving the accuracy of the experiment. [Summary of the invention]

[0008] In view of this, the present invention provides a method, device and computer readable medium for screening mesenchymal stem cells.

[0009] The present application provides a method for screening mesenchymal stem cells, which comprises the following steps:

[0010] S1. Collecting a microscopic image of mesenchymal stem cells of each sample, wherein the microscopic image is accompanied by a sample number and a time stamp;

[0011] S2, after eliminating invalid areas of the microscopic image of the mesenchymal stem cells, performing differential calculation on the valid areas to generate a preprocessed image;

[0012] S3, after dividing the pre-processed image into a plurality of image blocks, calculating the image information entropy of the image blocks, calculating the image information entropy of the pre-processed image by aggregation, and then constructing a time series sample of the image information entropy and the timestamp of the pre-processed image;

[0013] S4, adding image categories to the time series samples, inputting them into the prediction model, and calculating the prediction results;

[0014] S5. Screening the mesenchymal stem cell samples according to the prediction results.

[0015] Furthermore, the microscopic image of the mesenchymal stem cells is generated in the following manner:

[0016] Initial microscopic images of all samples at multiple moments are collected according to a preset period. After each initial microscopic image is assigned a sample number and a timestamp, the initial microscopic images of the same sample at each moment are combined into a microscopic image group.

[0017] Furthermore, the S2 specifically includes the following steps:

[0018] S21, selecting an image to be processed from the initial microscopic image group and obtaining the three-dimensional coordinates of the processed image, the three-dimensional coordinates carrying RGB information; using the three-dimensional coordinates at the center of the image to be processed as the standard three-dimensional coordinates;

[0019] S22, performing a difference calculation between the RGB information of the other three-dimensional coordinates and the RGB information of the standard three-dimensional coordinates, identifying the three-dimensional coordinates whose difference exceeds a preset threshold as invalid areas, selecting and eliminating all invalid areas, and then identifying the remaining areas as valid areas;

[0020] S23, performing sum and difference calculations on the three-dimensional coordinates of the valid area and the standard three-dimensional coordinate information to obtain the sum and difference of the three-dimensional coordinates;

[0021] S24 , obtaining three-dimensional coordinate difference information based on the three-dimensional coordinate difference value and the three-dimensional coordinate sum value, and generating a preprocessed image according to the coordinate difference information.

[0022] Furthermore, the S3 specifically includes the following steps:

[0023] S31, after setting the image block size, divide the pre-processed image into N overlapping image blocks, and calculate the image blocks by the formula H(x)=α·Φ(x)F[exp(P(y)+V(y))]-β·Γ(x)·F -1 [exp(P(y)+V(y))] 2 / ωCalculate the information entropy of each image block, where H(x) is the image information entropy of the image block, α is the first empirical parameter, β is the second empirical parameter, Φ(x) is the Goode-Turing smoothing process, Γ(x) is the Gaussian filter smoothing process, F is the Fourier transform and F -1 is the inverse Fourier transform, P(y) is the probability distribution of the pixel values of the preprocessed image, and V(y) is the phase spectrum of the preprocessed image;

[0024] S32, aggregating the image information entropy of each image block by calculating a weighted average value to obtain the image information entropy of the preprocessed image;

[0025] S33. Construct a time series sample based on the image information entropy and timestamp of the preprocessed image.

[0026] Furthermore, the image category is generated specifically through the following steps:

[0027] Set the size of the input image to W×H×C, where W represents width, H represents height, and C represents the number of channels;

[0028] The convolutional neural network structure is used to extract the features of each channel and output a tensor whose size is expressed as W′×H′×D c , where W′ and H′ are the width and height of the output feature map, D c is the characteristic dimension of channel C;

[0029] Output the feature vector fij corresponding to each position (i, j) of the feature map c Compressed into a single feature vector sij c ;

[0030] The single feature tensors of all channels are stacked to generate a stacked tensor S. The size of the stacked tensor is expressed as W′×H′×C, where each position (i, j) corresponds to a C-dimensional vector sij;

[0031] The C-dimensional vector sij corresponding to each position (i, j) is input into the fully connected layer and the output tensor P is obtained. The tensor P represents the probability that each position belongs to each category, thereby obtaining the final image category.

[0032] Furthermore, the prediction model is constructed through the following steps:

[0033] The time series samples are represented by vector s t =[f1,f2,...,f n ] is represented as a state, and each element represents the characteristics of a homology group;

[0034] The prediction method is represented by vector a t=[g1,g2,...,g m ] are represented as actions, and each element is a characteristic of a processing method;

[0035] The growth status of mesenchymal stem cells is used as the reward function r t ;

[0036] Construct Q function Q(s t ,a t )←(1-α)Q(s t ,a t )+α(r t +γmax a Q(s t+1 ,a)), where α represents the learning rate, r t Indicates taking action a in the current state t The immediate reward obtained after γ represents the loss factor, max a Q(s t+1 ,a) means in state s t+1 The maximum Q value corresponding to all actions;

[0037] By a*=argmax(Q(s t ,a t ))Select the optimal strategy, a* represents the optimal action;

[0038] Constructing a reward function

[0039] Among them, Q t In state s t Next take action a t The corresponding Q value; o represents an element in the input data set η; ψ represents the set of action weights; K represents the number of quantiles; τ represents the temperature parameter; E is the number of iterations; Indicates that under the optimal strategy, state s t Next take action a t corresponding losses;

[0040] The model is iteratively trained to the preset accuracy through training samples to obtain the prediction model.

[0041] Furthermore, after the prediction model is constructed, the method further includes:

[0042] Distribute the constructed prediction model on each node;

[0043] The node uploads the acquired time series samples of the additional image categories to the blockchain network, and after verification and authorization through the smart contract, requests other nodes to assist in prediction;

[0044] After receiving requests from other nodes, the node performs model calculations based on the prediction model and outputs the prediction results;

[0045] When multiple nodes provide prediction results, the contributions and rewards of the nodes are recorded through smart contracts;

[0046] The nodes elected by the blockchain consensus verify and confirm the prediction results and output the final prediction results.

[0047] On the other hand, the present invention also provides a mesenchymal stem cell screening device, which comprises:

[0048] an acquisition module, configured to acquire a microscopic image of the mesenchymal stem cells of each sample, wherein the microscopic image is accompanied by a sample number and a timestamp;

[0049] A preprocessing module is used to remove invalid areas of the microscopic image of mesenchymal stem cells and perform differential calculation on the valid areas to generate a preprocessed image;

[0050] a calculation module, configured to divide the pre-processed image into a plurality of image blocks, calculate the image information entropy of the image blocks, calculate the image information entropy of the pre-processed image by aggregation, and then construct a time series sample of the image information entropy and the timestamp of the pre-processed image;

[0051] A prediction module, configured to add image categories to the time series samples, input the samples into a prediction model, and calculate a prediction result;

[0052] A screening module is used to screen the mesenchymal stem cell samples according to the prediction results.

[0053] The device includes at least one processor and a memory, wherein the at least one processor is coupled to the memory and is configured to read and execute instructions in the memory to perform any one of the above methods.

[0054] On the other hand, the present invention further provides a computer-readable medium, characterized in that the computer-readable medium stores program code, and when the computer program code runs on a computer, the computer executes any of the above methods.

[0055] One of the above technical solutions has the following beneficial effects:

[0056] The present invention provides a method for screening mesenchymal stem cells. First, microscopic images of mesenchymal stem cells of each sample are collected, and the microscopic images are accompanied by sample numbers and timestamps. Then, after eliminating invalid areas of the microscopic images of the mesenchymal stem cells, a differential calculation is performed on the valid areas to generate a pre-processed image. Then, after dividing the pre-processed image into multiple image blocks, the image information entropy of the image blocks is calculated. The image information entropy of the pre-processed image is calculated by aggregation, and then a time series sample of the image information entropy of the pre-processed image and the timestamp is constructed. Subsequently, after adding image categories to the time series samples, the model is input into a prediction model and the prediction results are calculated. Finally, the mesenchymal stem cell samples are screened according to the prediction results. In this way, the final state of the stem cells can be evaluated in the early state of the stem cells, thereby screening out samples with better conditions in advance, effectively avoiding waste in the culture process, reducing the consumption of manpower and material resources, and improving the accuracy of the experiment.

Brief Description of the Drawings

[0057] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0058] Figure 1 This is a schematic flow chart of a method for screening mesenchymal stem cells provided in an embodiment of the present invention;

[0059] Figure 2 This is a structural framework diagram of a mesenchymal stem cell screening device provided by an embodiment of the present invention;

[0060] Figure 3 The figure is a schematic diagram of the hardware structure of a mesenchymal stem cell screening device provided by an embodiment of the present invention. [Specific implementation method]

[0061] In order to better understand the technical solution of the present invention, the embodiments of the present invention are described in detail below.

[0062] It should be understood that the embodiments described are only a portion of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative work are within the scope of protection of the present invention.

[0063] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The singular forms "a", "an", "the" and "the" used in the embodiments of the present invention and the appended claims are also intended to include plural forms unless the context clearly indicates otherwise.

[0064] It should be understood that the term "and / or" as used herein is merely a description of the relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, or B exists alone. Furthermore, the character " / " in this document generally indicates that the associated objects are in an "or" relationship.

[0065] Please refer to Figure 1 The present application provides a method for culturing and screening mesenchymal stem cells, the screening method comprising the following steps:

[0066] S1. Collecting a microscopic image of mesenchymal stem cells of each sample, wherein the microscopic image is accompanied by a sample number and a time stamp;

[0067] S2, after eliminating invalid areas of the microscopic image of the mesenchymal stem cells, performing differential calculation on the valid areas to generate a preprocessed image;

[0068] S3, after dividing the pre-processed image into a plurality of image blocks, calculating the image information entropy of the image blocks, calculating the image information entropy of the pre-processed image by aggregation, and then constructing a time series sample of the image information entropy and the timestamp of the pre-processed image;

[0069] S4, adding image categories to the time series samples, inputting them into the prediction model, and calculating the prediction results;

[0070] S5. Screening the mesenchymal stem cell samples according to the prediction results.

[0071] The present invention provides a method for screening mesenchymal stem cells. First, microscopic images of mesenchymal stem cells of each sample are collected, and the microscopic images are accompanied by sample numbers and timestamps. Then, after eliminating invalid areas of the microscopic images of the mesenchymal stem cells, a differential calculation is performed on the valid areas to generate a pre-processed image. Then, after dividing the pre-processed image into multiple image blocks, the image information entropy of the image blocks is calculated. The image information entropy of the pre-processed image is calculated by aggregation, and then a time series sample of the image information entropy of the pre-processed image and the timestamp is constructed. Subsequently, after adding image categories to the time series samples, the model is input into a prediction model and the prediction results are calculated. Finally, the mesenchymal stem cell samples are screened according to the prediction results. In this way, the final state of the stem cells can be evaluated in the early state of the stem cells, thereby screening out samples with better conditions in advance, effectively avoiding waste in the culture process, reducing the consumption of manpower and material resources, and improving the accuracy of the experiment.

[0072] The microscopic image of the mesenchymal stem cells in the embodiment of the present application is generated in the following manner:

[0073] Initial microscopic images of all samples at multiple moments are collected according to a preset period. After each initial microscopic image is assigned a sample number and a timestamp, the initial microscopic images of the same sample at each moment are combined into a microscopic image group.

[0074] Specifically, the system is in T1, T2, ..., T n Initial microscopic images of each culture sample are collected at multiple moments. After attaching the sample number and time stamp to each image, the initial microscopic images of the same sample at each moment are combined into a microscopic image group. Multiple culture samples form multiple initial microscopic image groups.

[0075] Specifically, the embodiment of the present application further refines S2, which specifically includes the following steps:

[0076] S21, selecting an image to be processed from the initial microscopic image group and obtaining the three-dimensional coordinates of the processed image, the three-dimensional coordinates carrying RGB information; using the three-dimensional coordinates at the center of the image to be processed as the standard three-dimensional coordinates;

[0077] S22, performing a difference calculation between the RGB information of the other three-dimensional coordinates and the RGB information of the standard three-dimensional coordinates, identifying the three-dimensional coordinates whose difference exceeds a preset threshold as invalid areas, selecting and eliminating all invalid areas, and then identifying the remaining areas as valid areas;

[0078] S23, performing sum and difference calculations on the three-dimensional coordinates of the valid area and the standard three-dimensional coordinate information to obtain the sum and difference of the three-dimensional coordinates;

[0079] S24 , obtaining three-dimensional coordinate difference information based on the three-dimensional coordinate difference value and the three-dimensional coordinate sum value, and generating a preprocessed image according to the coordinate difference information.

[0080] Specifically, the embodiment of the present application further refines S3, which specifically includes the following steps:

[0081] S31, after setting the image block size, divide the pre-processed image into N overlapping image blocks, and calculate the image blocks by the formula H(x)=α·Φ(x)F[exp(P(y)+V(y))]-β·Γ(x)·F -1 [exp(P(y)+V(y))] 2 / ω Calculate the information entropy of each image block, where H(x) is the image information entropy of the image block, α is the first empirical parameter, β is the second empirical parameter, Φ(x) is the Goode-Turing smoothing process, Γ(x) is the Gaussian filter smoothing process, F is the Fourier transform and F -1 is the inverse Fourier transform, P(y) is the probability distribution of the pixel values of the preprocessed image, and V(y) is the phase spectrum of the preprocessed image;

[0082] S32, aggregating the image information entropy of each image block by calculating a weighted average value to obtain the image information entropy of the preprocessed image;

[0083] S33. Construct a time series sample based on the image information entropy and timestamp of the preprocessed image.

[0084] Furthermore, the image category is generated specifically through the following steps:

[0085] Set the size of the input image to W×H×C, where W represents width, H represents height, and C represents the number of channels;

[0086] The convolutional neural network structure is used to extract the features of each channel and output a tensor whose size is expressed as W′×H′×D c , where W′ and H′ are the width and height of the output feature map, D c is the characteristic dimension of channel C;

[0087] Output the feature vector fij corresponding to each position (i, j) of the feature map c Compressed into a single feature vector sij c ;

[0088] The single feature tensors of all channels are stacked to generate a stacked tensor S. The size of the stacked tensor is expressed as W′×H′×C, where each position (i, j) corresponds to a C-dimensional vector sij;

[0089] The C-dimensional vector sij corresponding to each position (i, j) is input into the fully connected layer and the output tensor P is obtained. The tensor P represents the probability that each position belongs to each category, thereby obtaining the final image category.

[0090] Specifically, the prediction model is constructed through the following steps:

[0091] The time series samples are represented by vector s t =[f1,f2,...,f n ] is represented as a state, and each element represents the characteristics of a homology group;

[0092] The prediction method is represented by vector a t =[g1,g2,...,g m ] are represented as actions, and each element is a characteristic of a processing method;

[0093] The growth status of mesenchymal stem cells is used as the reward function r t ;

[0094] Construct Q function Q(s t ,a t)←(1-α)Q(s t ,a t )+α(r t +γmax a Q(s t+1 ,a)), where α represents the learning rate, r t Indicates taking action a in the current state t The immediate reward obtained after γ represents the loss factor, max a Q(s t+1 ,a) means that in state s t+1 The maximum Q value corresponding to all actions;

[0095] By a*=argmax(Q(s t ,a t ))Select the optimal strategy, a* represents the optimal action;

[0096] Constructing a reward function

[0097] Among them, Q t In state s t Next take action a t The corresponding Q value; o represents an element in the input data set η; ψ represents the set of action weights; K represents the number of quantiles; τ represents the temperature parameter; E is the number of iterations; Indicates that under the optimal strategy, state s t Next take action a t corresponding losses;

[0098] The model is iteratively trained to the preset accuracy through training samples to obtain the prediction model.

[0099] Furthermore, after the prediction model is constructed, the method further includes:

[0100] Distribute the constructed prediction model on each node;

[0101] The node uploads the acquired time series samples of the additional image categories to the blockchain network, and after verification and authorization through the smart contract, requests other nodes to assist in prediction;

[0102] After receiving requests from other nodes, the node performs model calculations based on the prediction model and outputs the prediction results;

[0103] When multiple nodes provide prediction results, the contributions and rewards of the nodes are recorded through smart contracts;

[0104] The nodes elected by the blockchain consensus verify and confirm the prediction results and output the final prediction results.

[0105] The present invention applies the above-mentioned cell state prediction model in a distributed environment to achieve more secure and reliable cell state prediction. Specifically, the technical solution of the present invention deploys the cell state prediction model in a blockchain network to realize a decentralized cell state prediction service.

[0106] In a blockchain network, every node can participate in the process of cell state prediction. Each time a prediction is made, a node uploads its observed cell state information to the blockchain network and requests assistance from other nodes. Other nodes then provide predictions based on their own data and models, ultimately reaching a consensus and obtaining the predicted result.

[0107] The decentralized and tamper-proof nature of blockchain technology ensures the security and reliability of cell state predictions. Furthermore, blockchain's consensus mechanism and smart contracts enable automated verification of prediction results and reward distribution.

[0108] This cell state prediction method based on blockchain technology can provide more reliable and efficient prediction services for the cell culture field, and also provides new ideas and directions for the application of blockchain technology in the field of life sciences.

[0109] The present invention screens out samples of mesenchymal stem cells with good growth and development, retains them, and continues to culture them. Specifically, it determines whether the prediction result reaches a preset threshold. If so, the mesenchymal stem cells in this group of samples are cultured; if not, the mesenchymal stem cells in this group of samples are cultured.

[0110] The present invention also provides a mesenchymal stem cell screening device, which comprises:

[0111] An acquisition module 210 is used to acquire a microscopic image of the mesenchymal stem cells of each sample, wherein the microscopic image is accompanied by a sample number and a timestamp;

[0112] A pre-processing module 220 is configured to remove invalid areas of the microscopic image of mesenchymal stem cells and perform differential calculation on valid areas to generate a pre-processed image;

[0113] A calculation module 230 is configured to divide the pre-processed image into a plurality of image blocks, calculate the image information entropy of the image blocks, calculate the image information entropy of the pre-processed image by aggregation, and then construct a time series sample of the image information entropy and the timestamp of the pre-processed image;

[0114] The prediction module 240 is used to add image categories to the time series samples, input them into the prediction model, and calculate the prediction results;

[0115] The screening module 250 is configured to screen the mesenchymal stem cell samples according to the prediction results.

[0116] Please refer to Figure 3 , which is a schematic diagram of the hardware structure of a mesenchymal stem cell screening device implemented in the present invention. The data prediction device includes at least one processor and memory. The at least one processor is coupled to the memory and is configured to read and execute instructions in the memory to perform the data prediction method provided in an embodiment of the present invention.

[0117] In a third aspect, an embodiment of the present invention provides a computer-readable medium having a program code stored therein, which, when executed on a computer, causes the computer to execute the data prediction method provided by an embodiment of the present invention.

[0118] The systems, devices, modules, or units described in the above embodiments may be implemented by computer chips or entities, or by products having certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0119] For the convenience of description, the above device is described as being divided into various units or modules according to their functions. Of course, when implementing the present invention, the functions of each unit or module can be implemented in the same or multiple software and / or hardware.

[0120] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0121] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0122] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0123] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0124] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0125] Memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0126] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0127] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0128] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0129] The present invention may be described in the general context of computer-executable instructions, such as program modules, executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, and the like that perform specific tasks or implement specific abstract data types. The present invention may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media, including storage devices.

[0130] The various embodiments of the present invention are described in a progressive manner. Similar portions between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences between the other embodiments. In particular, the system embodiment is generally similar to the method embodiment, so its description is relatively simple. For relevant portions, refer to the description of the method embodiment.

[0131] The foregoing is merely an embodiment of the present invention and is not intended to limit the present invention. It will be apparent to those skilled in the art that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention are intended to be included within the scope of the claims of the present invention.

Claims

1. A method for screening mesenchymal stem cells, characterized in that: The method comprises the following steps: S1. Collecting a microscopic image of mesenchymal stem cells of each sample, wherein the microscopic image is accompanied by a sample number and a time stamp; S2, after eliminating invalid areas of the microscopic image of the mesenchymal stem cells, performing differential calculation on the valid areas to generate a preprocessed image; S3, after dividing the pre-processed image into a plurality of image blocks, calculating the image information entropy of the image blocks, calculating the image information entropy of the pre-processed image by aggregation, and then constructing a time series sample of the image information entropy and the timestamp of the pre-processed image; S4, adding image categories to the time series samples, inputting them into the prediction model, and calculating the prediction results; S5. screening the mesenchymal stem cell sample according to the prediction result; The microscopic image of the mesenchymal stem cells is specifically generated by the following method: Collect initial microscopic images of all samples at multiple times according to a preset period, attach a sample number and a timestamp to each initial microscopic image, and combine the initial microscopic images of the same sample at different times into a microscopic image group; The S2 specifically includes the following steps: S21, selecting an image to be processed from the initial microscopic image group and obtaining the three-dimensional coordinates of the processed image, the three-dimensional coordinates carrying RGB information; using the three-dimensional coordinates at the center of the image to be processed as the standard three-dimensional coordinates; S22, performing a difference calculation between the RGB information of the other three-dimensional coordinates and the RGB information of the standard three-dimensional coordinates, identifying the three-dimensional coordinates whose difference exceeds a preset threshold as invalid areas, selecting and eliminating all invalid areas, and then identifying the remaining areas as valid areas; S23, performing sum and difference calculations on the three-dimensional coordinates of the valid area and the standard three-dimensional coordinate information to obtain the sum and difference of the three-dimensional coordinates; S24, obtaining three-dimensional coordinate difference information based on the three-dimensional coordinate difference value and the three-dimensional coordinate sum, and generating a preprocessed image according to the coordinate difference information; The S3 specifically includes the following steps: S31, after setting the image block size, divide the pre-processed image into N overlapping image blocks, and calculate the image blocks by the formula H(x)=α·Φ(x)F[exp(P(y)+V(y))]-β·Γ(x)·F -1 [exp(P(y)+V(y))] 2 / ω Calculate the information entropy of each image block, where H(x) is the image information entropy of the image block, α is the first empirical parameter, β is the second empirical parameter, Φ(x) is the Goode-Turing smoothing process, Γ(x) is the Gaussian filter smoothing process, F is the Fourier transform and F -1 is the inverse Fourier transform, P(y) is the probability distribution of the pixel values of the preprocessed image, and V(y) is the phase spectrum of the preprocessed image; S32, aggregating the image information entropy of each image block by calculating a weighted average value to obtain the image information entropy of the preprocessed image; S33, constructing a time series sample based on image information entropy and timestamp of the preprocessed image; The image category is specifically generated through the following steps: Set the size of the input image to W×H×C, where W represents width, H represents height, and C represents the number of channels; The convolutional neural network structure is used to extract the features of each channel and output a tensor whose size is expressed as W′×H′×D c , where W′ and H′ are the width and height of the output feature map, D c is the characteristic dimension of channel C; Output the feature vector fij corresponding to each position (i, j) of the feature map c Compressed into a single feature vector sij c ; The single feature tensors of all channels are stacked to generate a stacked tensor S. The size of the stacked tensor is expressed as W′×H′×C, where each position (i, j) corresponds to a C-dimensional vector sij; The C-dimensional vector sij corresponding to each position (i, j) is input into the fully connected layer and the output tensor P is obtained. The tensor P represents the probability of each position belonging to each category, thereby obtaining the final image category. The prediction model is constructed through the following steps: The time series samples are represented by vector s t =[f1,f2,...,f n ] is represented as a state, and each element represents the characteristics of a homology group; The prediction method is represented by vector a t =[g1,g2,...,g m ] are represented as actions, and each element is a characteristic of a processing method; The growth status of mesenchymal stem cells is used as the reward function r t ; Construct Q function Q(s t ,a t )←(1-α)Q(s t ,a t )+α(r t +γmax a Q(s t+1 ,a)), where α represents the learning rate, r t Indicates taking action a in the current state t The immediate reward obtained after γ represents the loss factor, max a Q(s t+1 ,a) means in state s t+1 The maximum Q value corresponding to all actions; By a*=argmax(Q(s t ,a t ))Select the optimal strategy, a* represents the optimal action; Constructing a reward function Among them, Q t In state s t Next take action a t The corresponding Q value; o represents an element in the input data set η; ψ represents the set of action weights; K represents the number of quantiles; τ represents the temperature parameter; E is the number of iterations; Indicates that under the optimal strategy, state s t Next take action a t corresponding losses; The model is iteratively trained to the preset accuracy through training samples to obtain the prediction model.

2. The method for screening mesenchymal stem cells according to claim 1, wherein: After the prediction model is built, the method further includes: Distribute the constructed prediction model on each node; The node uploads the acquired time series samples of the additional image categories to the blockchain network, and after verification and authorization through the smart contract, requests other nodes to assist in prediction; After receiving requests from other nodes, the node performs model calculations based on the prediction model and outputs the prediction results; When multiple nodes provide prediction results, the contributions and rewards of the nodes are recorded through smart contracts; The nodes elected by the blockchain consensus verify and confirm the prediction results and output the final prediction results.

3. A mesenchymal stem cell screening device using the method according to any one of claims 1 and 2, characterized in that: The device includes: an acquisition module, configured to acquire a microscopic image of the mesenchymal stem cells of each sample, wherein the microscopic image is accompanied by a sample number and a timestamp; A preprocessing module is used to remove invalid areas of the microscopic image of mesenchymal stem cells and perform differential calculation on the valid areas to generate a preprocessed image; a calculation module, configured to divide the pre-processed image into a plurality of image blocks, calculate the image information entropy of the image blocks, calculate the image information entropy of the pre-processed image by aggregation, and then construct a time series sample of the image information entropy and the timestamp of the pre-processed image; A prediction module, configured to add image categories to the time series samples, input the samples into a prediction model, and calculate a prediction result; A screening module is used to screen the mesenchymal stem cell samples according to the prediction results.

4. A mesenchymal stem cell screening device, characterized in that: The system comprises at least one processor and a memory, wherein the at least one processor is coupled to the memory and configured to read and execute instructions in the memory to perform the method according to any one of claims 1 and 2.

5. A computer-readable medium, characterized in that The computer-readable medium stores a program code, and when the computer program code is run on a computer, the computer is caused to perform the method according to any one of claims 1 and 2 .

Images