Cell stress analysis method based on deep learning and its application

Abstract

The present application is applicable to the field of cell analysis technology, and in particular to a deep learning-based cell stress analysis method and application thereof, the method comprising: obtaining stress state values, cell environment values ​​and cell morphology characteristic values ​​of target analysis cells to provide data support for subsequent analysis; determining multiple stress analysis reference trajectories of the target analysis cells based on the cell environment values ​​and cell morphology characteristic values ​​of the target analysis cells; determining multiple stress analysis groups of the target analysis cells based on the multiple stress analysis reference trajectories and stress state values ​​of the target analysis cells; inputting each stress analysis group of the target analysis cells into a stress analysis model to obtain stress analysis data corresponding to each stress analysis group, thereby providing a quantitative evaluation of the cell stress state; determining the cell stress analysis results of the target analysis cells based on the stress analysis data corresponding to all stress analysis groups, thereby quickly, accurately and intuitively determining the cell stress performance.

Description

Technical Field

[0001] The present application belongs to the field of cell analysis technology, and in particular relates to a cell stress analysis method based on deep learning and its application. Background Art

[0002] Cell stress analysis is an important field of biomedical research, which aims to evaluate the response of cells to different environmental factors. These environmental factors include mechanical force, temperature change, chemicals and hypoxia, which usually trigger adaptive responses of cells, such as proliferation, differentiation or apoptosis. There are many analysis methods, including morphological observation, microscopic examination of cell morphological changes, detection of expression levels of specific biomarkers such as stress proteins and reactive oxygen species in cells, analysis of changes in gene expression using qPCR or RNA-seq technology, and evaluation of cell cycle and apoptosis using flow cytometry. 。

[0003] At present, the stress analysis of cells is still based on morphological observation. The researchers examine the changes in cell morphology through a microscope and detect specific biomarkers in cells. They often need to record and process data manually. The lack of effective analysis auxiliary tools limits the efficiency and accuracy of cell stress analysis. Summary of the invention

[0004] The embodiments of the present application provide a deep learning-based cell stress analysis method and its application, which can solve the problem in the prior art that the efficiency and accuracy of cell stress analysis are limited due to the lack of efficient analysis auxiliary tools.

[0005] In a first aspect, an embodiment of the present application provides a cell stress analysis method based on deep learning, comprising:

[0006] Obtain stress state values, cell environment values ​​and cell morphology characteristic values ​​of target cells for analysis;

[0007] Determining a plurality of stress analysis reference trajectories of the target analysis cell based on the cell environment value and the cell morphology characteristic value of the target analysis cell;

[0008] Determining a plurality of stress analysis groups of the target analysis cell according to the plurality of stress analysis reference trajectories and the stress state values ​​of the target analysis cell;

[0009] Input each of the stress analysis groups of the target analysis cells into a stress analysis model to obtain stress analysis data corresponding to each of the stress analysis groups; wherein the stress analysis model is a model obtained through deep learning;

[0010] The cell stress analysis result of the target analysis cell is determined according to the stress analysis data corresponding to all the stress analysis groups.

[0011] The above technical solutions in the embodiments of the present application have at least the following technical effects:

[0012] The cell stress analysis method based on deep learning provided in the embodiment of the present application provides data support for subsequent analysis by obtaining the stress state value, cell environment value and cell morphological characteristic value of the target analysis cell. Based on the cell environment value and cell morphological characteristic value of the target analysis cell, multiple stress analysis reference trajectories of the target analysis cell are determined to better reflect the stress response of different regions of the cell. Based on the multiple stress analysis reference trajectories and stress state values ​​of the target analysis cell, multiple stress analysis groups of the target analysis cell are determined to deeply understand the stress response mechanism of the cell. Each stress analysis group of the target analysis cell is input into the stress analysis model to obtain the stress analysis data corresponding to each stress analysis group, providing a quantitative assessment of the cell stress state. According to the stress analysis data corresponding to all stress analysis groups, the cell stress analysis results of the target analysis cell are determined, and the cell stress performance is analyzed quickly, accurately and intuitively.

[0013] In a second aspect, the present application embodiment provides a cell stress analysis system based on deep learning, including:

[0014] An acquisition unit, used for acquiring stress state values, cell environment values ​​and cell morphology characteristic values ​​of target cells for analysis;

[0015] A trajectory unit, configured to determine a plurality of stress analysis reference trajectories of the target analysis cell based on the cell environment value and the cell morphology characteristic value of the target analysis cell;

[0016] an aggregation unit, configured to determine a plurality of stress analysis groups of the target analysis cell according to the plurality of stress analysis reference trajectories and the stress state values ​​of the target analysis cell;

[0017] An analysis unit, used for inputting each of the stress analysis groups of the target analysis cells into a stress analysis model to obtain stress analysis data corresponding to each of the stress analysis groups; wherein the stress analysis model is a model obtained through deep learning;

[0018] The result unit is used to determine the cell stress analysis result of the target analysis cell according to the stress analysis data corresponding to all the stress analysis groups.

[0019] In the third aspect, an embodiment of the present application provides a deep learning-based cell stress analysis device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method described in any one of the above aspects when executing the computer program.

[0020] In a fourth aspect, an embodiment of the present application provides a computer program product, which, when executed on a terminal device, enables the terminal device to execute the method of any one of the above aspects.

[0021] It can be understood that the beneficial effects of the second to fourth aspects mentioned above can be found in the relevant descriptions in the above aspects and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0023] Figure 1 It is a flowchart of a cell stress analysis method based on deep learning provided in one embodiment of the present application;

[0024] Figure 2 is a schematic diagram of the operation of a cell stress analysis method based on deep learning provided in an embodiment of the present application;

[0025] Figure 3 is a flowchart of step S230 in the cell stress analysis method based on deep learning provided in one embodiment of the present application;

[0026] Figure 4 is a flow chart of step S300 in a cell stress analysis method based on deep learning provided in an embodiment of the present application;

[0027] Figure 5 is a schematic diagram of the structure of a cell stress analysis system based on deep learning provided in one embodiment of the present application;

[0028] Figure 6 It is a structural schematic diagram of a deep learning-based cell stress analysis device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0029] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0030] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0031] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0032] As used in the specification of this application and the appended claims, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting" depending on the context. Similarly, the phrase "if it is determined" or "if the described condition or event is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce the described condition or event is detected" or "in response to detecting the described condition or event" depending on the context.

[0033] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0034] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0035] At present, the stress analysis of cells is still based on morphological observation. The researchers examine the changes in cell morphology through a microscope and detect specific biomarkers in cells. They often need to record and process data manually. The lack of effective analysis auxiliary tools limits the efficiency and accuracy of cell stress analysis.

[0036] To solve the above problems, the embodiments of the present application provide a cell stress analysis method, system and device based on deep learning. In this method, by obtaining the stress state value, cell environment value and cell morphological characteristic value of the target analysis cell, data support is provided for subsequent analysis. Based on the cell environment value and cell morphological characteristic value of the target analysis cell, multiple stress analysis reference trajectories of the target analysis cell are determined to better reflect the stress response of different regions of the cell. Based on the multiple stress analysis reference trajectories and stress state values ​​of the target analysis cell, multiple stress analysis groups of the target analysis cell are determined, so as to deeply understand the stress response mechanism of the cell. Each stress analysis group of the target analysis cell is input into the stress analysis model to obtain the stress analysis data corresponding to each stress analysis group, which provides a quantitative evaluation of the cell stress state. According to the stress analysis data corresponding to all stress analysis groups, the cell stress analysis results of the target analysis cell are determined, and the cell stress performance is analyzed quickly, accurately and intuitively.

[0037] The deep learning-based cell stress analysis method provided in the embodiment of the present application can be applied to a terminal device. In this case, the terminal device is the executor of the deep learning-based cell stress analysis method provided in the embodiment of the present application. The embodiment of the present application does not impose any restrictions on the specific type of the terminal device.

[0038] For example, the terminal device may be an ultra-mobile personal computer (UMPC), a netbook, a personal digital assistant (PDA), a desktop computer, a computer, a laptop computer, a communication device, a computing device, a satellite wireless device, etc.

[0039] In order to better understand the deep learning-based cell stress analysis method provided in the embodiment of the present application, the specific implementation process of the deep learning-based cell stress analysis method provided in the embodiment of the present application is exemplarily introduced below.

[0040] Figure 1 A schematic flow chart of a cell stress analysis method based on deep learning provided in an embodiment of the present application is shown. Figure 2 The operation flow chart of the cell stress analysis method based on deep learning provided in the embodiment of the present application is shown. The cell stress analysis method based on deep learning includes:

[0041] S100, obtaining stress state values, cell environment values ​​and cell morphology characteristic values ​​of target cells for analysis.

[0042] It can be understood that stress state values ​​usually reflect the mechanical or biological stress that cells are subjected to under specific conditions. Cell environment values ​​describe external factors that affect cell morphology, including temperature, pH value, ion concentration, nutrient supply, etc. Cell morphological characteristic values ​​describe the biological responses of cells under specific environmental conditions, which may include cell proliferation, migration, apoptosis, etc. Stress state values ​​directly measure the degree of deformation of the cell membrane or the pressure changes inside the cell through microscopy techniques (such as atomic force microscopy, laser confocal microscopy, etc.), or use fluorescent probes or photosensitive probes to monitor changes in certain indicators in the cell (such as calcium ion concentration, pH value, etc.), which can indirectly reflect the stress state of the cell. Cell environment values ​​can use chemical sensors or biosensors to monitor the physical and chemical properties of the environment around the cell in real time. For example, a pH sensor is used to monitor the pH of the culture medium, or an oxygen sensor is used to monitor the dissolved oxygen level in the culture medium. To obtain cell morphological characteristic values, a high-resolution camera can be used to record the dynamic changes of cells during the culture process, and the movement trajectory and morphological pattern of the cells can be analyzed through image processing technology. Flow cytometry is used to analyze the biological characteristics of cells under flow conditions, such as cell size, shape, and surface marker expression, which can be associated with the morphological characteristic values ​​of the cells. The obtained stress state values, cell environment values, and cell morphological characteristic value data can be stored in a related database and obtained through an API interface or data upload.

[0043] S200, determining a plurality of stress analysis reference trajectories of the target analysis cell based on the cell environment value and the cell morphology characteristic value of the target analysis cell.

[0044] It can be understood that the stress analysis reference trajectory refers to the path or curve established in the process of cell stress analysis to describe the stress response of cells under specific conditions. The stress analysis reference trajectory reflects the stress path of cells under different stress states and is an important indicator of the dynamic response of cells. The data of cell environmental values ​​and cell morphological characteristic values ​​can be integrated to determine the characteristic information such as the morphology of the cells, and the target analysis cells can be rasterized according to the characteristic information to divide the spatial position of the cells into multiple grid regions. The features of each grid region will be marked according to its environmental value and morphological characteristics, and computer vision technology, such as image segmentation, can be used to identify different regions of the cell. In each grid region, multiple stress analysis reference points are identified. The selection of reference points can be based on feature extraction methods, such as using K-means clustering to identify areas that perform significantly under stress conditions. Each reference point should be representative and able to reflect the stress characteristics of its region. Based on the selected stress analysis reference points, multiple stress analysis reference trajectories are defined.

[0045] In a possible implementation, S200, based on the cell environment value and the cell morphology characteristic value of the target analysis cell, determines a plurality of stress analysis reference trajectories of the target analysis cell, including:

[0046] S210, based on the cell environment value and the cell morphology characteristic value of the target analysis cell, rasterizing the target analysis cell to obtain a plurality of rasterized regions of the target analysis cell.

[0047] It can be understood that rasterization is to divide a continuous spatial area into several small, uniform cells (or grids), each grid cell represents a certain spatial position and feature value. In cell analysis, the rasterized area can convert the complex morphology and environmental characteristics of cells into a format that is easier to process.

[0048] Optionally, S210, based on the cell environment value and the cell morphology characteristic value of the target analysis cell, a gridding process is performed on the target analysis cell to obtain a plurality of gridding regions of the target analysis cell, including:

[0049] S211, establishing a cell structure model of the target analysis cell based on the cell environment value and cell morphology characteristic value of the target analysis cell.

[0050] It can be understood that a preliminary 3D structure can be pre-generated using a 3D reconstruction algorithm (such as Delaunay triangulation), and optimized based on the cell environment value and cell morphological feature value of the cell to obtain a cell structure model of the target analysis cell. This can be achieved through computer graphics software, such as using a 3D modeling tool to map the cell environment value and morphological feature value to the parameters of the 3D model, which can include the size, shape, surface features, etc. of the cell. The cell structure model can reflect the actual morphology of the cell, including the relative position and size of the cell membrane, nucleus and other organelles. The cell environment value and morphological feature value can be integrated into the geometric model. The dynamic behavior of the cell can be determined by defining a parameterized model and taking the environmental features as the input of the model.

[0051] S212, performing rasterization processing on the target analysis cell according to the cell structure model of the target analysis cell to obtain a plurality of rasterized regions of the target analysis cell.

[0052] It can be understood that rasterization refers to decomposing a complex three-dimensional model into multiple small, discrete parts, usually called "grids" or "grids". The rasterized areas can be more easily numerically calculated, analyzed and visualized. The rasterization algorithm of the cell structure model can be pre-set. The rasterization algorithm can include octree, grid generation algorithm (such as Delaunay triangulation) and voxelization. The cell structure model can be processed by the selected rasterization algorithm to generate multiple rasterized areas. Each rasterized area will represent a small area inside the cell. The rasterized area and its corresponding environmental values ​​and morphological feature values ​​can be stored in a database for subsequent analysis and access. For example, a spatial data structure (such as an R-tree or a KD-tree) can be used to quickly retrieve data in a specific area.

[0053] S220, determining a plurality of stress analysis reference points of the target analysis cell according to the plurality of gridded regions of the target analysis cell.

[0054] It can be understood that the stress analysis reference point refers to a specific position selected in the cell structure model, which can reflect the stress state of the cell under different conditions. High-density areas can be automatically detected by using a density clustering algorithm, which usually contains important stress information, and one or more center points can be selected from each high-density area as reference points. You can choose to cluster the data points into K clusters and select the centroid of each cluster as a reference point. You can also select a certain number of points uniformly distributed in each grid as stress analysis reference points through a uniform distribution method.

[0055] S230, determining multiple stress analysis reference trajectories of the target analysis cell based on the multiple stress analysis reference points of the target analysis cell.

[0056] It can be understood that according to the stress analysis reference points of the target analysis cells, the coordinate information of the stress analysis reference points can be used to generate trajectories. Each stress analysis reference point represents the stress state of a certain position of the cell under specific conditions. By analyzing the spatial distribution and mutual relationship of these points, the distribution trend of stress in the cell can be derived. Interpolation methods or curve fitting techniques can be included to create smooth stress analysis reference trajectories between stress analysis reference points. Stress analysis reference trajectories are not just the connection of points, they also reflect the dynamic changes of cells under different stress states. Stress analysis reference trajectories can represent the propagation path of stress in different regions of the cell and reveal the response mechanism of the cell to external stress sources. For example, a trajectory may show stress concentration in a certain area, while another trajectory may reflect the dispersion trend of stress, which is helpful to understand the stress response of different regional positions of the cell.

[0057] Optionally, see Figure 3 , S230, based on the multiple stress analysis reference points of the target analysis cell, determining multiple stress analysis reference trajectories of the target analysis cell, including:

[0058] S231, based on multiple stress analysis reference points of the target analysis cell, determine the coordinate information of each stress analysis reference point of the target analysis cell; wherein the coordinate information is used to reflect the coordinate position of the stress analysis reference point in the cell structure model.

[0059] It can be understood that the coordinate information provides an accurate spatial position for each reference point, which helps to understand the stress distribution inside the cell. The coordinate information can be used to simulate and calculate the stress state and visualize the data. The geometric data in the cell structure model can be used to calculate the coordinate values ​​of the reference point and describe it using the coordinate system of the model. The calculation method can use methods such as center of mass calculation and boundary point calculation so that the obtained coordinates can represent the actual position of the stress analysis reference point in the cell model.

[0060] S232, screening is performed according to the coordinate information of each stress analysis reference point of the target analysis cell to obtain a plurality of starting stress analysis reference points and a plurality of ending stress analysis reference points of the target analysis cell.

[0061] It is understood that by analyzing the spatial relationship between stress analysis reference points, it is possible to determine which points can be considered as starting and ending points. The definitions of the starting stress analysis reference point and the ending stress analysis reference point are relative. The starting reference point generally indicates the area where the stress analysis process begins, while the ending reference point indicates the ending position of the stress analysis. By performing detailed calculations and analyses on the coordinates of each stress analysis reference point, starting and ending points that meet specific criteria can be identified. The analysis can include changes in stress values, relative relationships of positions, and other biological or physical considerations. Geometric algorithms or data mining techniques can be used to screen these points. For example, by analyzing the changes in the stress state of each reference point, it can be determined which points are the starting points of stress transfer and which are the ending points of stress concentration.

[0062] S233, determining multiple stress analysis reference trajectories of the target analysis cell according to multiple starting stress analysis reference points and multiple ending stress analysis reference points of the target analysis cell.

[0063] It can be understood that a suitable path generation algorithm (such as Dijkstra algorithm, A* algorithm) can be used to connect the start and end points to obtain multiple stress analysis reference trajectories of the target analysis cell, and the generated stress analysis reference trajectories are stored in a database for subsequent analysis and visualization.

[0064] S300, determining a plurality of stress analysis groups of target analysis cells according to a plurality of stress analysis reference trajectories and stress state values ​​of the target analysis cells.

[0065] It can be understood that stress analysis groups refer to the division of cells into multiple groups with similar stress characteristics according to the stress state and morphology of cells under specific conditions during the cell analysis process. The stress analysis group of the target analysis cell is constructed by integrating multiple stress analysis reference trajectories and the corresponding stress state values. The stress state value is the key data describing the stress level experienced by the cell under specific conditions, and exists in the form of multiple data points, which reflect the stress changes of the cell at different time and spatial positions. By combining these stress state values ​​with the pre-defined stress analysis reference trajectories, the stress distribution patterns of different regions within the cell can be identified, and then the cells can be divided into multiple stress analysis groups. Each stress analysis group represents a cell region that may exhibit similar stress characteristics under specific conditions, which is helpful for subsequent detailed analysis and research.

[0066] In one possible implementation, see Figure 4 S300, the stress state value includes a plurality of stress state data points, and a plurality of stress analysis groups of the target analysis cell are determined according to a plurality of stress analysis reference trajectories and stress state values ​​of the target analysis cell, including:

[0067] S310, calculating the correlation between each stress state data point and each stress analysis reference trajectory according to the multiple stress analysis reference trajectories and the multiple stress state data points of the target analysis cell.

[0068] It can be understood that correlation refers to the association between two variables. The correlation between each stress state data point and each stress analysis reference trajectory can be used to determine their mutual relationship in stress analysis. Appropriate correlation measurement methods can be selected, including Pearson correlation coefficient, Spearman rank correlation coefficient and cosine similarity. Pearson correlation coefficient is used to measure linear correlation, and the value range is [-1,,1]. The closer the value is to 1 or -1, the stronger the correlation. Spearman rank correlation coefficient is applicable to nonlinear relationships and is calculated based on the ranking of data. Cosine similarity is used to evaluate the similarity of two vectors and is often used for data comparison in high-dimensional space.

[0069] Optionally, S310, based on multiple stress analysis reference trajectories and multiple stress state data points of the target analysis cell, respectively calculating the correlation between each stress state data point and each stress analysis reference trajectory, includes:

[0070] S311, based on multiple stress analysis reference trajectories and multiple stress state data points of the target analysis cell, respectively calculate the relative position information of each stress state data point and each stress analysis reference trajectory; wherein the relative position information is used to reflect the relative coordinate position of the stress state data point and the stress analysis reference trajectory in the cell structure model.

[0071] It is understood that a standard coordinate system can be selected to describe the position in the cell structure model, such as using a three-dimensional coordinate system (X,, Y,, Z) to represent the coordinates inside the cell. For each stress state data point, its distance to each point in each stress analysis reference trajectory can be calculated, such as using the Euclidean distance to measure the relative position.

[0072] S312, obtaining the correlation between each stress state data point and each stress analysis reference trajectory according to the relative position information between each stress state data point and each stress analysis reference trajectory.

[0073] It is understood that the correlation can be calculated using different methods, such as Euclidean distance, Pearson correlation coefficient or cosine similarity. By analyzing the relative position of each data point relative to the reference trajectory, its importance in the analysis model can be determined. For example, this can be achieved by converting the Euclidean distance into correlation. For example, if a maximum distance D is set, the correlation can be defined as: , , where S is the correlation between the stress state data point and the stress analysis reference trajectory, D is the preset maximum distance constant, and d is the Euclidean distance. The distance calculation results can be recorded in a matrix, where the rows represent the stress state data points and the columns represent the stress analysis reference trajectory. The correlation calculation results can be recorded in a matrix, where the rows represent the stress state data points and the columns represent the stress analysis reference trajectory. If the distance d between a stress state data point and a trajectory is small, it can be considered that the correlation S between them is high, otherwise the correlation is low, which helps to understand the performance of different stress states in cell structures, thereby providing a basis for subsequent analysis.

[0074] S320 , according to the correlation between each stress state data point and each stress analysis reference trajectory, each stress state data point is matched with the stress analysis reference trajectory to obtain a plurality of initial stress analysis trajectory groups of the target analysis cell.

[0075] It can be understood that each stress state data point can be matched with the stress analysis reference trajectory to obtain multiple initial stress analysis trajectory groups of the target analysis cell. The stress state data point is matched and divided with the stress analysis reference trajectory with the highest correlation to form multiple initial stress analysis trajectory groups. At this time, the initial stress analysis trajectory group includes the actual data point and the stress trajectory to be analyzed, which can provide basic data for subsequent analysis.

[0076] S330, clustering multiple initial stress analysis trajectory groups of the target analysis cell to determine multiple stress analysis groups of the target analysis cell.

[0077] It can be understood that cluster analysis is to group similar trajectories in the initial stress analysis trajectory group, thereby fusing similar initial stress analysis trajectory groups to obtain different stress analysis groups. Stress analysis group refers to a classification set formed by clustering multiple initial stress analysis trajectory groups in cell stress analysis. Each stress analysis group represents a group of trajectories with similar characteristics and response patterns under a specific stress state. The stress analysis group is the result of matching and dividing based on the correlation between the stress state data points and the stress analysis reference trajectory, which is used to reveal the stress response of different regions of the cell. A variety of clustering algorithms, such as K-means clustering, hierarchical clustering, or DBSCAN, can be used to classify according to the similarity and correlation between trajectories. The goal of clustering is to group trajectories with similar stress characteristics into the same group, simplify and merge the analysis, and effectively reveal the response patterns and morphological characteristics of cells under different stress conditions. For example, by analyzing the center point and distribution characteristics of each trajectory group, the biological significance of cells under different stress states can be understood.

[0078] S400, input each stress analysis group of the target analysis cell into the stress analysis model to obtain stress analysis data corresponding to each stress analysis group; wherein the stress analysis model is a model obtained through deep learning.

[0079] It can be understood that stress analysis data can reflect the stress intensity distribution of various regions within the cell, show which regions are subjected to greater mechanical or environmental stress, and help identify potential weaknesses or damaged areas. By analyzing the stress analysis group, the stress changes of cells in different regions can be determined, revealing the adaptability and recovery ability of cells. Stress analysis data can be combined with the morphological characteristics of cells to reflect the physiological responses of cells under specific stress states, such as migration, proliferation or apoptosis. The constructed stress analysis group can be passed to the stress analysis model. The stress analysis model calculates the input through its internal algorithms and parameters to generate corresponding stress analysis data. The stress analysis data output by the stress analysis model can be recorded so that the stress analysis group can correspond to its corresponding stress analysis data. Stress analysis data can include stress values, stress distribution, and other related analysis indicators.

[0080] Exemplarily, the sample data of the stress analysis group can be collated and input into the training set, and the sample data includes the stress analysis group itself and the stress analysis data corresponding to the stress analysis group sample. The sample data can be preprocessed, including removing noise, standardizing values, and processing missing data. After preprocessing, the sample data is divided into a training set and a validation set for performance evaluation during the model training process. A suitable algorithm can be selected according to the nature of the task, such as a deep learning model, a support vector machine, or a decision tree. During the training process, the model will be iterated multiple times on the training set, and each iteration will calculate the output through forward propagation, and compare it with the true label, and use the loss function to evaluate the model performance. The back propagation algorithm is used to update the model parameters to reduce prediction errors. After the training is completed, the model will be evaluated on the validation set to test its generalization ability. This can be done by calculating indicators such as accuracy, recall, and F1 score. After multiple rounds of training and verification, when the performance of the model reaches expectations, it can be applied to actual stress analysis tasks. The example training set is as follows:

[0081]

[0082] S500, determining a cell stress analysis result of a target analysis cell according to stress analysis data corresponding to all stress analysis groups.

[0083] It can be understood that the stress analysis group is aggregated based on multi-dimensional data such as cell microstructure, environmental changes, and cell morphological characteristics. The stress analysis data can be combined with the cell structure model. By combining the cell structure model, the distribution of stress inside the cell can be visualized, and the performance of cells under different stress conditions can be further analyzed. It can not only be used to simulate the response of cells under external stress, but also provide guidance for future experimental design and clinical applications. At the end of this stage, the obtained cell stress analysis results will be presented through visualization tools to facilitate a more intuitive understanding of the performance of cells under different stress states.

[0084] In a possible implementation, S500, determining the cell stress analysis result of the target analysis cell according to the stress analysis data corresponding to all stress analysis groups, includes:

[0085] S510: Determine the correlation between the stress analysis data and the cell structure model according to the stress analysis data corresponding to all stress analysis groups.

[0086] It can be understood that the stress analysis group is the result of matching and dividing based on the correlation between the stress state data points and the stress analysis reference trajectory, which is used to reveal the stress response of different regions of the cell. The obtained stress analysis data can be combined with the three-dimensional structural characteristics of the cell to reveal the stress response mechanism of the cell in different regions. The stress analysis data corresponding to all stress analysis groups contain multiple stress state data points, revealing that different stress characteristics are reflected in different cell regions. These stress analysis data can be mapped to the structural model of the cell to identify the stress distribution in different cell regions. The regions or grids divided according to the cell structure model can be mapped to establish a mapping relationship, and the stress state data points in the stress analysis group corresponding to the stress analysis data can be matched with the corresponding cell regions. For example, the value of a stress analysis group may correspond to the stress state of a certain region of the cell, which can be achieved through spatial coordinates or regional indexes. Statistical analysis can be used to evaluate the degree of correlation between stress data and cell structure regions, find out the actual manifestation of stress state in cell structure, and help understand the adaptation mechanism of cells under different environmental conditions.

[0087] S520, determining a stress analysis distribution map of the target analysis cell based on the correlation between the stress analysis data and the cell structure model, and determining the stress analysis distribution map as a cell stress analysis result of the target analysis cell.

[0088] It can be understood that the stress analysis distribution map is a visual representation of the stress state in the cell, which can intuitively show the stress distribution of the cell in different regions. The correlation between the stress data and the cell structure can be determined, and the stress analysis data can be mapped to the three-dimensional model of the cell. The stress data can be accurately located to a specific position in the cell model according to its spatial coordinates. The process of determining the stress analysis distribution map can include the application of data interpolation and visualization technology. For example, grid interpolation (such as Kriging interpolation, linear interpolation, etc.) is included to infer the stress state of the unknown position through the surrounding known data points, thereby generating a continuous stress distribution map. The stress analysis distribution map shows the stress intensity in different regions inside the cell, and the change of color or pattern can intuitively reflect the height and distribution characteristics of the stress. After the stress analysis distribution map is generated, it can be established as the cell stress analysis result of the target analysis cell, which can provide a comprehensive understanding of the stress response of the cell under specific conditions.

[0089] Optionally, S520, based on the correlation between the stress analysis data and the cell structure model, determining a stress analysis distribution map of the target analysis cell, and determining the stress analysis distribution map as a cell stress analysis result of the target analysis cell, includes:

[0090] S521, based on the correlation between the stress analysis data and the cell structure model, determine the position distribution of the stress analysis data in the cell structure model.

[0091] It can be understood that the stress analysis group corresponding to each stress analysis data contains the association between the stress state data point and the cell structure model, and the stress analysis reference trajectory and the specific spatial mapping coordinates of the stress state data point can be determined, corresponding to the geometric structure of the cell model. Through these spatial coordinates, the stress state data can be accurately mapped to a specific area of ​​the cell structure model, that is, the position distribution of the stress analysis data in the cell structure model is determined.

[0092] S522, determining a stress analysis distribution map of the target analysis cell according to the position distribution of the stress analysis data in the cell structure model, and determining the stress analysis distribution map as a cell stress analysis result of the target analysis cell.

[0093] It can be understood that the stress state data is combined with the geometric structure of the cell model according to the position distribution of the stress analysis data in the cell structure model. Each stress state data point corresponds to a specific coordinate position of the cell model. Graphics software or data visualization tools can be used to display the stress levels of different regions in the form of color gradients, heat maps, etc. For example, low stress areas may be represented by blue, while high stress areas are represented by red, thus forming a clear stress distribution map. The stress analysis distribution map is not just a simple visualization of the data. It represents the comprehensive analysis results of the internal stress state of the cell, realizes the rapid, accurate and intuitive determination of the cell stress performance, and provides a powerful auxiliary capability for understanding the stress performance of the cell.

[0094] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0095] Corresponding to the deep learning-based cell stress analysis method described in the above embodiment, the embodiment of the present application also provides a deep learning-based cell stress analysis system, and each unit of the system can implement each step of the deep learning-based cell stress analysis method. Figure 5 A structural block diagram of a deep learning-based cell stress analysis system provided in an embodiment of the present application is shown. For ease of explanation, only the parts related to the embodiment of the present application are shown.

[0096] Reference Figure 5 , the deep learning-based cell stress analysis system includes:

[0097] An acquisition unit, used for acquiring stress state values, cell environment values ​​and cell morphology characteristic values ​​of target cells for analysis;

[0098] A trajectory unit, configured to determine a plurality of stress analysis reference trajectories of the target analysis cell based on the cell environment value and the cell morphology characteristic value of the target analysis cell;

[0099] an aggregation unit, configured to determine a plurality of stress analysis groups of the target analysis cell according to the plurality of stress analysis reference trajectories and the stress state values ​​of the target analysis cell;

[0100] An analysis unit, used for inputting each of the stress analysis groups of the target analysis cells into a stress analysis model to obtain stress analysis data corresponding to each of the stress analysis groups; wherein the stress analysis model is a model obtained through deep learning;

[0101] The result unit is used to determine the cell stress analysis result of the target analysis cell according to the stress analysis data corresponding to all the stress analysis groups.

[0102] It should be noted that the information interaction, execution process, etc. between the above-mentioned systems / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0103] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit module can exist physically separately, or two or more unit modules can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0104] The present application also provides a cell stress analysis device based on deep learning. Figure 6 This is a schematic diagram of the structure of a cell stress analysis device based on deep learning provided in one embodiment of the present application. Figure 6 As shown, the cell stress analysis device 6 based on deep learning of this embodiment includes: at least one processor 60 ( Figure 6 Only one is shown), at least one memory 61 ( Figure 6 Only one is shown in the figure) and a computer program 62 stored in the at least one memory 61 and executable on the at least one processor 60. When the processor 60 executes the computer program 62, the deep learning-based cell stress analysis device 6 implements the steps in any of the above-mentioned deep learning-based cell stress analysis method embodiments, or the deep learning-based cell stress analysis device 6 implements the functions of each unit in the above-mentioned system embodiments.

[0105] Exemplarily, the computer program 62 may be divided into one or more units, which are stored in the memory 61 and executed by the processor 60 to complete the present application. The one or more units may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program 62 in the deep learning-based cell stress analysis device 6.

[0106] The cell stress analysis device 6 based on deep learning can be a computing device or terminal device such as a desktop computer, a notebook, a PDA, and a cloud server. The cell stress analysis device based on deep learning can include, but is not limited to, a processor 60 and a memory 61. Those skilled in the art can understand that Figure 6 It is only an example of the deep learning-based cell stress analysis device 6 and does not constitute a limitation of the deep learning-based cell stress analysis device 6. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components, for example, it may also include input and output devices, network access devices, buses, etc.

[0107] The processor 60 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0108] In some embodiments, the memory 61 may be an internal storage unit of the deep learning-based cell stress analysis device 6, such as a hard disk or memory of the deep learning-based cell stress analysis device 6. In other embodiments, the memory 61 may also be an external storage device of the deep learning-based cell stress analysis device 6, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the deep learning-based cell stress analysis device 6. Further, the memory 61 may also include both an internal storage unit and an external storage device of the deep learning-based cell stress analysis device 6. The memory 61 is used to store an operating system, an application, a boot loader (BootLoader), data, and other programs, such as the program code of the computer program, etc. The memory 61 may also be used to temporarily store data that has been output or is to be output.

[0109] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in any of the above method embodiments are implemented.

[0110] An embodiment of the present application provides a computer program product. When the computer program product runs on a deep learning-based cell stress analysis device, the deep learning-based cell stress analysis device implements the steps in any of the above-mentioned method embodiments.

[0111] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device that can carry the computer program code to a deep learning-based cell stress analysis device, a recording medium, a computer memory, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), an electric carrier signal, a telecommunication signal, and a software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0112] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0113] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0114] In the embodiments provided in the present application, it should be understood that the disclosed deep learning-based cell stress analysis system / deep learning-based cell stress analysis device and method can be implemented in other ways. For example, the deep learning-based cell stress analysis system / deep learning-based cell stress analysis device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0115] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0116] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A cell stress analysis method based on deep learning, characterized in that: include: Obtain stress state values, cell environment values ​​and cell morphology characteristic values ​​of target cells for analysis; Determining a plurality of stress analysis reference trajectories of the target analysis cell based on the cell environment value and the cell morphology characteristic value of the target analysis cell; Determining a plurality of stress analysis groups of the target analysis cell according to the plurality of stress analysis reference trajectories and the stress state values ​​of the target analysis cell; Inputting each of the stress analysis groups of the target analysis cells into a stress analysis model to obtain stress analysis data corresponding to each of the stress analysis groups; The stress analysis model is a deep learning model trained with sample data of the stress analysis group; The sample data includes the stress analysis group itself and the stress analysis data corresponding to the stress analysis group sample; The cell stress analysis result of the target analysis cell is determined according to the stress analysis data corresponding to all the stress analysis groups.

2. The cell stress analysis method based on deep learning according to claim 1, characterized in that: The determining of a plurality of stress analysis reference trajectories of the target analysis cell based on the cell environment value and the cell morphology characteristic value of the target analysis cell comprises: Based on the cell environment value and the cell morphology characteristic value of the target analysis cell, performing a rasterization process on the target analysis cell to obtain a plurality of rasterized regions of the target analysis cell; Determining a plurality of stress analysis reference points of the target analysis cell according to the plurality of gridded regions of the target analysis cell; Based on the multiple stress analysis reference points of the target analysis cell, multiple stress analysis reference trajectories of the target analysis cell are determined.

3. The cell stress analysis method based on deep learning according to claim 2, characterized in that: The step of performing a rasterization process on the target analysis cell based on the cell environment value and the cell morphology characteristic value of the target analysis cell to obtain a plurality of rasterized regions of the target analysis cell includes: Establishing a cell structure model of the target analysis cell based on the cell environment value and the cell morphology characteristic value of the target analysis cell; The target analysis cell is subjected to a rasterization process according to the cell structure model of the target analysis cell to obtain a plurality of rasterized regions of the target analysis cell.

4. The cell stress analysis method based on deep learning according to claim 3, characterized in that: The determining of a plurality of stress analysis reference trajectories of the target analysis cell based on the plurality of stress analysis reference points of the target analysis cell comprises: Based on the multiple stress analysis reference points of the target analysis cell, determine the coordinate information of each stress analysis reference point of the target analysis cell; wherein the coordinate information is used to reflect the coordinate position of the stress analysis reference point in the cell structure model; Screening is performed according to the coordinate information of each stress analysis reference point of the target analysis cell to obtain a plurality of starting stress analysis reference points and a plurality of ending stress analysis reference points of the target analysis cell; A plurality of stress analysis reference trajectories of the target analysis cell are determined according to the plurality of starting stress analysis reference points and the plurality of ending stress analysis reference points of the target analysis cell.

5. The cell stress analysis method based on deep learning according to claim 4, characterized in that: The stress state value includes a plurality of stress state data points, and determining a plurality of stress analysis groups of the target analysis cell according to the plurality of stress analysis reference trajectories and the stress state value of the target analysis cell includes: According to the plurality of stress analysis reference trajectories and a plurality of stress state data points of the target analysis cell, respectively calculating the correlation between each stress state data point and each stress analysis reference trajectory; According to the correlation between each stress state data point and each stress analysis reference trajectory, each stress state data point is matched with a stress analysis reference trajectory to obtain a plurality of initial stress analysis trajectory groups of the target analysis cell; Clustering is performed on the multiple initial stress analysis trajectory groups of the target analysis cell to determine multiple stress analysis groups of the target analysis cell.

6. The cell stress analysis method based on deep learning according to claim 5, characterized in that: The step of calculating the correlation between each stress state data point and each stress analysis reference trajectory based on the multiple stress analysis reference trajectories and multiple stress state data points of the target analysis cell comprises: According to the multiple stress analysis reference trajectories and multiple stress state data points of the target analysis cell, respectively calculate the relative position information of each stress state data point and each stress analysis reference trajectory; wherein the relative position information is used to reflect the relative coordinate position of the stress state data point and the stress analysis reference trajectory in the cell structure model; According to the relative position information between each stress state data point and each stress analysis reference trajectory, the correlation between each stress state data point and each stress analysis reference trajectory is obtained.

7. The cell stress analysis method based on deep learning as claimed in claim 3, characterized in that: Determining the cell stress analysis result of the target analysis cell according to the stress analysis data corresponding to all the stress analysis groups includes: Determining, according to the stress analysis data corresponding to all the stress analysis groups, a correlation relationship between the stress analysis data and the cell structure model; Based on the correlation between the stress analysis data and the cell structure model, a stress analysis distribution map of the target analysis cell is determined, and the stress analysis distribution map is determined as a cell stress analysis result of the target analysis cell.

8. The cell stress analysis method based on deep learning according to claim 7, characterized in that: The step of determining the stress analysis distribution map of the target analysis cell based on the correlation between the stress analysis data and the cell structure model, and determining the stress analysis distribution map as the cell stress analysis result of the target analysis cell, comprises: Based on the correlation between the stress analysis data and the cell structure model, determining the position distribution of the stress analysis data in the cell structure model; According to the position distribution of the stress analysis data in the cell structure model, a stress analysis distribution map of the target analysis cell is determined, and the stress analysis distribution map is determined as a cell stress analysis result of the target analysis cell.

9. A cell stress analysis system based on deep learning, characterized in that: include: An acquisition unit, used for acquiring stress state values, cell environment values ​​and cell morphology characteristic values ​​of target cells for analysis; A trajectory unit, configured to determine a plurality of stress analysis reference trajectories of the target analysis cell based on the cell environment value and the cell morphology characteristic value of the target analysis cell; an aggregation unit, configured to determine a plurality of stress analysis groups of the target analysis cell according to the plurality of stress analysis reference trajectories and the stress state values ​​of the target analysis cell; An analysis unit, used for inputting each of the stress analysis groups of the target analysis cells into a stress analysis model to obtain stress analysis data corresponding to each of the stress analysis groups; The stress analysis model is a deep learning model trained with sample data of the stress analysis group; The sample data includes the stress analysis group itself and the stress analysis data corresponding to the stress analysis group sample; The result unit is used to determine the cell stress analysis result of the target analysis cell according to the stress analysis data corresponding to all the stress analysis groups.

10. A cell stress analysis device based on deep learning, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 8 is implemented.

Images