A method and system for culturing kidney and liver cells based on multimodal information

Abstract

The present invention relates to the field of cell culture technology, and in particular to a kidney and liver cell culture method and system based on multimodal information. The actual multimodal culture data of kidney and liver cells in a culture device are obtained at a preset culture time node, and the culture state of the kidney and liver cells at the preset culture time node is analyzed according to the final multimodal simulation culture data and the multimodal culture actual data; if the culture state of the kidney and liver cells at the preset culture time node is normal, the culture device is not regulated; if the culture state of the kidney and liver cells at the preset culture time node is abnormal, the culture device is regulated. The present invention realizes dynamic monitoring and precise regulation of the kidney and liver cell culture process, and improves the success rate, quality and stability of cell culture.

Description

Technical Field

[0001] The present invention relates to the technical field of cell culture, and in particular to a method and system for culturing kidney and liver cells based on multimodal information. Background Art

[0002] Kidney and liver are important metabolic and detoxification organs in the human body, and their cell culture technology has important application value in biomedical research, drug screening, disease model construction, and regenerative medicine. With the rapid development of multimodal information technology, analysis methods combining multi-source data such as imaging, genomics, proteomics, and metabolomics have provided new ideas for cell culture. Multimodal information can comprehensively reflect the morphology, function, metabolic state of cells and their interaction with the environment, providing a scientific basis for optimizing cell culture conditions and improving cell function expression. However, the current kidney and liver cell culture methods based on multimodal information are still in the exploratory stage, lacking a systematic technical framework and efficient implementation methods. Especially in the combination of simulation culture and actual culture, how to use multimodal simulation data to guide the actual culture process, and to monitor and dynamically regulate in real time during the culture process, is still a technical problem that needs to be solved urgently. In addition, the existing technology also has shortcomings in the integration and analysis of multimodal data, the construction accuracy of simulation models, and the real-time evaluation and regulation strategies of culture status. Therefore, developing a kidney and liver cell culture method based on multimodal information and realizing the precision, intelligence and efficiency of the cell culture process through the organic combination of simulation culture and actual culture has important scientific significance and application value. Summary of the invention

[0003] The present invention overcomes the deficiencies of the prior art and provides a method and system for culturing kidney and liver cells based on multimodal information.

[0004] To achieve the above-mentioned purpose, the technical solution adopted by the present invention is:

[0005] The present invention discloses a method for culturing kidney and liver cells based on multimodal information, comprising the following steps:

[0006] Obtaining several preset culture schemes for kidney and liver cells to be cultured, importing each preset culture scheme into a kidney and liver cell simulation culture system for simulation culture, and screening out a final culture scheme and a final multimodal simulation culture dataset based on the multimodal simulation culture datasets of each preset culture scheme;

[0007] Controlling the culture equipment to actually culture the kidney and liver cells to be cultured according to the final culture plan;

[0008] In the actual culture process, actual multimodal culture data of kidney and liver cells in the culture device are obtained at a preset culture time node, wherein the actual multimodal culture data includes actual cell morphology data, actual cell function data, and actual cell proliferation data;

[0009] Extracting the final multimodal simulation culture data corresponding to the preset culture time node from the final multimodal simulation culture data set; analyzing the culture state of kidney and liver cells at the preset culture time node according to the final multimodal simulation culture data and the multimodal culture actual data;

[0010] If the culture state of the kidney and liver cells is normal at the preset culture time node, the culture equipment is not regulated; if the culture state of the kidney and liver cells is abnormal at the preset culture time node, the culture equipment is regulated.

[0011] Specifically, the steps of constructing the kidney and liver cell simulation culture system are as follows:

[0012] Acquire target multi-source biological data through a big data network, wherein the target multi-source biological data includes tissue samples of kidney and liver cells, gene sequences, protein expression profiles, and metabolites;

[0013] Extracting features from the target multi-source biological data to obtain key feature information of kidney and liver cells in a physiological state, wherein the key feature information includes structural composition, functional units, intercellular communication mechanisms, material and energy metabolism rules, and growth and differentiation logic;

[0014] According to the key feature information, based on biological principles, the structural composition is constructed as a basic framework, the functional units are set as modules, the signal transduction pathway is established according to the intercellular communication mechanism, the material exchange and conversion process is set according to the material and energy metabolism rules, and the cell state transition rules are set according to the growth and differentiation logic to initialize the cell culture simulation model;

[0015] Obtaining a functional specification of the culture equipment, and setting a working range of physical and chemical environmental parameters in the cell culture simulation model according to the functional specification, wherein the working range of the physical and chemical environmental parameters includes the type and concentration gradient of nutrients, temperature range, pH fluctuation, and gas exchange conditions;

[0016] Obtaining the initial state of the kidney and liver cells to be cultured, and setting the type, quantity, distribution pattern and initial functional state of the cells in the cell culture simulation model according to the initial state of the kidney and liver cells to be cultured;

[0017] After the setting is completed, the model parameters of the cell culture simulation model are saved, and the cell culture simulation model is connected and integrated with external input and output devices to form a kidney and liver cell simulation culture system.

[0018] Specifically, each preset culture scheme is introduced into the kidney and liver cell simulation culture system for simulation culture, and the final culture scheme and the final multimodal simulation culture data set are screened out according to the multimodal simulation culture data set of each preset culture scheme, specifically:

[0019] Obtaining several preset culture schemes for kidney and liver cells to be cultured, and introducing each preset culture scheme into the kidney and liver cell simulation culture system;

[0020] Based on each preset culture scheme, the kidney and liver cell simulation culture system is controlled to perform simulation culture work, and the simulated culture cell morphology data, simulated culture cell function data and simulated culture cell proliferation data of the kidney and liver cells in the kidney and liver cell simulation culture system at several preset culture time nodes after culture by each preset culture scheme are obtained, and a multimodal simulation culture data set corresponding to each preset culture scheme is generated;

[0021] Obtaining the culture task requirements of the target kidney and liver cells, obtaining the standard cell morphology data, standard cell function data, and standard cell proliferation data of the kidney and liver cells at each preset culture time node according to the culture task requirements, and generating a multimodal standard culture data set corresponding to the culture target kidney and liver cells;

[0022] Introducing a cosine similarity algorithm, and calculating the cosine similarity value between the multimodal simulation training data set corresponding to each preset training scheme and the multimodal standard training data set based on the cosine similarity algorithm;

[0023] Comparing the cosine similarity value between the multimodal simulation training data set corresponding to each preset training scheme and the multimodal standard training data set with a preset threshold;

[0024] If there is at least one cosine similarity value greater than a preset threshold, the cosine similarity values ​​greater than the preset threshold are sorted by size to extract the maximum cosine similarity value;

[0025] The preset training program corresponding to the maximum cosine similarity value is marked as the final training program, and the multimodal simulation training data set corresponding to the final training program is marked as the final multimodal simulation training data set; the final training program and its final multimodal simulation training data set are output.

[0026] The following steps are also included:

[0027] If all cosine similarity values ​​are not greater than the preset threshold, all preset training plans are marked as invalid training plans, and a warning message that the preset training plans need to be re-planned is generated;

[0028] The re-planned preset culture scheme is obtained, and the re-planned preset culture scheme is continued to be simulated based on the kidney and liver cell simulation culture system until the cosine similarity value between the multimodal simulation culture data set corresponding to at least one re-planned preset culture scheme and the multimodal standard culture data set is greater than a preset threshold.

[0029] Specifically, the culture state of kidney and liver cells at the preset culture time node is analyzed according to the final multimodal simulation culture data and the multimodal culture actual data, specifically:

[0030] Acquiring actual multimodal culture data of kidney and liver cells in a culture device at a preset culture time node, and obtaining actual multimodal culture data of kidney and liver cells at the preset culture time node;

[0031] Extracting the final multimodal simulation culture data of kidney and liver cells at the preset culture time node from the final multimodal simulation culture data set;

[0032] Calculate the cosine similarity value between the actual multimodal culture data of kidney and liver cells at the preset culture time node and the final multimodal simulation culture data;

[0033] If the cosine similarity value between the actual multimodal culture data of the kidney and liver cells at the preset culture time node and the final multimodal simulation culture data is greater than the preset similarity threshold, the culture state of the kidney and liver cells at the preset culture time node is marked as a normal state;

[0034] If the cosine similarity value between the actual multimodal culture data of the kidney and liver cells at the preset culture time node and the final multimodal simulation culture data is not greater than the preset similarity threshold, the culture state of the kidney and liver cells at the preset culture time node will be marked as an abnormal state.

[0035] Specifically, if the culture state of the kidney and liver cells is abnormal at the preset culture time node, the culture device is regulated and processed, specifically:

[0036] If the culture state of the kidney and liver cells is abnormal at the preset culture time node, then obtaining various actual working parameters of the culture device at the current preset culture time node based on various preset sensors in the culture device;

[0037] and obtaining various preset working parameters of the cultivation equipment at the current preset cultivation time node in the final cultivation plan;

[0038] Calculate the difference between various actual working parameters of the cultivation equipment at the current preset cultivation time node and the corresponding preset working parameters;

[0039] Compare the difference between each actual working parameter and the corresponding preset working parameter with the preset difference threshold value respectively;

[0040] If the difference between a certain actual working parameter and the corresponding preset working parameter is greater than the preset difference threshold, the actual working parameter is marked as an abnormal working parameter, and the offset amplitude of the abnormal working parameter is determined according to the difference between the actual working parameter and the corresponding preset working parameter, so as to obtain the abnormal working parameter and its offset amplitude of the cultivation equipment at the current preset cultivation time node;

[0041] If the difference between a certain actual working parameter and the corresponding preset working parameter is not greater than the preset difference threshold, the actual working parameter is marked as a normal working parameter.

[0042] Specifically, if the culture state of the kidney and liver cells is abnormal at the preset culture time node, the culture device is regulated and processed, and the following steps are also included:

[0043] Formulate in advance the control measures corresponding to the various deviation amplitude ranges of various working parameters of the culture equipment during operation;

[0044] Constructing a knowledge index library, importing the control measures corresponding to various deviation amplitude ranges of various working parameters of the cultivation equipment during the working process into the knowledge index library; and regularly updating the knowledge index library;

[0045] Obtaining abnormal working parameters and offset amplitudes of the cultivation equipment at the current preset cultivation time node;

[0046] Importing the abnormal working parameters and the offset amplitudes of the cultivation equipment at the current preset cultivation time node into the knowledge index library for matching, and obtaining corresponding control measures;

[0047] The acquired corresponding control measures are transmitted to the control terminal of the culture device, so that the control terminal performs control processing on the abnormal working parameters of the culture device based on the corresponding control measures.

[0048] The present invention also discloses a kidney and liver cell culture system based on multimodal information, the kidney and liver cell culture system comprising a memory and a processor, the memory storing a kidney and liver cell culture method program, and when the kidney and liver cell culture method program is executed by the processor, any one of the kidney and liver cell culture method steps is implemented.

[0049] The present invention solves the technical defects existing in the background technology, and has the following beneficial effects: the kidney and liver cell culture method based on multimodal information can screen out the optimal culture scheme through simulated culture, provide a scientific basis for actual culture, and in the actual culture process, through the collection and comparative analysis of multimodal data, the culture state of the cells can be accurately judged; according to the culture state, it is determined whether to regulate the culture equipment, thereby realizing dynamic monitoring and precise regulation of the kidney and liver cell culture process, and improving the success rate, quality and stability of cell culture. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, 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 invention. For ordinary technicians in this field, drawings of other embodiments can be obtained based on these drawings without paying creative work.

[0051] Figure 1 The figure is a flowchart of the overall method of culturing kidney and liver cells based on multimodal information;

[0052] Figure 2 A partial flowchart of a method for culturing kidney and liver cells based on multimodal information;

[0053] Figure 3 This is a system block diagram of a kidney and liver cell culture system based on multimodal information. DETAILED DESCRIPTION

[0054] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0055] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0056] like Figure 1 As shown, the present invention discloses a method for culturing kidney and liver cells based on multimodal information, comprising the following steps:

[0057] S102, obtaining several preset culture schemes for kidney and liver cells to be cultured, importing each preset culture scheme into a kidney and liver cell simulation culture system for simulation culture, and screening out a final culture scheme and a final multimodal simulation culture dataset based on the multimodal simulation culture datasets of each preset culture scheme;

[0058] S104, controlling the culture equipment to actually culture the kidney and liver cells to be cultured according to the final culture plan;

[0059] S106. During the actual culture process, obtaining actual multimodal culture data of kidney and liver cells in the culture device at a preset culture time node, wherein the actual multimodal culture data includes actual cell morphology data, actual cell function data, and actual cell proliferation data;

[0060] S108, extracting the final multimodal simulation culture data corresponding to the preset culture time node from the final multimodal simulation culture data set; analyzing the culture state of the kidney and liver cells at the preset culture time node according to the final multimodal simulation culture data and the multimodal culture actual data;

[0061] S110. If the culture state of the kidney and liver cells is normal at the preset culture time node, the culture equipment is not regulated; if the culture state of the kidney and liver cells is abnormal at the preset culture time node, the culture equipment is regulated.

[0062] It should be noted that this kidney and liver cell culture method based on multimodal information can screen out the optimal culture plan through simulated culture, providing a scientific basis for actual culture, and in the actual culture process, through the collection and comparative analysis of multimodal data, the culture state of the cells can be accurately judged; according to the culture state, it is decided whether to regulate the culture equipment, thereby realizing dynamic monitoring and precise regulation of the kidney and liver cell culture process, and improving the success rate, quality and stability of cell culture.

[0063] Specifically, the steps of constructing the kidney and liver cell simulation culture system are as follows:

[0064] Acquire target multi-source biological data through a big data network, wherein the target multi-source biological data includes tissue samples of kidney and liver cells, gene sequences, protein expression profiles, and metabolites;

[0065] Extracting features from the target multi-source biological data to obtain key feature information of kidney and liver cells in a physiological state, wherein the key feature information includes structural composition, functional units, intercellular communication mechanisms, material and energy metabolism rules, and growth and differentiation logic;

[0066] According to the key feature information, based on biological principles, the structural composition is constructed as a basic framework, the functional units are set as modules, the signal transduction pathway is established according to the intercellular communication mechanism, the material exchange and conversion process is set according to the material and energy metabolism rules, and the cell state transition rules are set according to the growth and differentiation logic to initialize the cell culture simulation model;

[0067] Obtaining a functional specification of the culture equipment, and setting a working range of physical and chemical environmental parameters in the cell culture simulation model according to the functional specification, wherein the working range of the physical and chemical environmental parameters includes the type and concentration gradient of nutrients, temperature range, pH fluctuation, and gas exchange conditions;

[0068] Obtaining the initial state of the kidney and liver cells to be cultured, and setting the type, quantity, distribution pattern and initial functional state of the cells in the cell culture simulation model according to the initial state of the kidney and liver cells to be cultured;

[0069] After the setting is completed, the model parameters of the cell culture simulation model are saved, and the cell culture simulation model is connected and integrated with external input and output devices to form a kidney and liver cell simulation culture system.

[0070] It should be noted that the target multi-source biological data is first obtained through the big data network. This step covers information on multiple levels of kidney and liver cells, from the macro level of tissue samples to the micro level of gene sequences, protein expression profiles, and metabolites. The acquisition of this multi-source data lays the foundation for a comprehensive understanding of kidney and liver cells, because different types of data reflect the characteristics of cells from different angles. Feature extraction is performed on the acquired multi-source biological data to obtain the key feature information of kidney and liver cells in physiological states. These key feature information includes the structural composition of cells, functional units, intercellular communication mechanisms, material and energy metabolism rules, and growth and differentiation logic. These features are the core elements of the normal physiological functions of cells. By extracting these features, it is possible to clarify how the various parts of the cell work together and how the cell interacts with the outside world.

[0071] Construct a cell culture simulation model based on key feature information. Based on biological principles, the structural composition is used as the basic framework, and the functional units are set as modules, which is similar to building a system with different functional areas. Establish signal transduction pathways according to the intercellular communication mechanism, which is the key to simulating how information is transmitted between cells. Set the material exchange and conversion process according to the rules of material and energy metabolism, which can reflect the dynamic changes of materials and energy in cells. Set the cell state transition rules according to the growth and differentiation logic to initialize the cell culture simulation model, so that a model that can simulate the physiological activities of kidney and liver cells is constructed from multiple aspects.

[0072] Obtain the functional specification of the culture equipment, and set the working range of the physical and chemical environmental parameters in the cell culture simulation model according to the specification. This step links the function of the actual culture equipment with the simulation model, so that the environmental parameters in the simulation model (such as nutrients, temperature, pH value, gas exchange, etc.) match the actual culture environment, improving the authenticity of the simulation. Obtain the initial state of the kidney and liver cells to be cultured, and set the type, quantity, distribution pattern and initial functional state of the cells in the cell culture simulation model according to the initial state. This makes the starting state of the simulation model consistent with the actual cells to be cultured. Finally, save the model parameters, and connect and integrate the model with external input and output devices to form a complete kidney and liver cell simulation culture system, so that it can easily interact with external devices to perform operations such as data input and result output.

[0073] The above-mentioned method for constructing a kidney and liver cell simulation culture system helps to simulate the scheme, optimize the culture conditions, and predict the culture results before actual culture, thereby improving the success rate and efficiency of kidney and liver cell culture, and can provide a reliable preliminary simulation foundation for subsequent cell culture-based research (such as drug screening, disease model construction, etc.).

[0074] Specifically, each preset culture scheme is introduced into the kidney and liver cell simulation culture system for simulation culture, and the final culture scheme and the final multimodal simulation culture data set are screened out according to the multimodal simulation culture data set of each preset culture scheme, specifically:

[0075] Obtaining several preset culture schemes for kidney and liver cells to be cultured, and introducing each preset culture scheme into the kidney and liver cell simulation culture system;

[0076] Based on each preset culture scheme, the kidney and liver cell simulation culture system is controlled to perform simulation culture work, and the simulated culture cell morphology data, simulated culture cell function data and simulated culture cell proliferation data of the kidney and liver cells in the kidney and liver cell simulation culture system at several preset culture time nodes after culture by each preset culture scheme are obtained, and a multimodal simulation culture data set corresponding to each preset culture scheme is generated;

[0077] Obtaining the culture task requirements of the target kidney and liver cells, obtaining the standard cell morphology data, standard cell function data, and standard cell proliferation data of the kidney and liver cells at each preset culture time node according to the culture task requirements, and generating a multimodal standard culture data set corresponding to the culture target kidney and liver cells;

[0078] Introducing a cosine similarity algorithm, and calculating the cosine similarity value between the multimodal simulation training data set corresponding to each preset training scheme and the multimodal standard training data set based on the cosine similarity algorithm;

[0079] Comparing the cosine similarity value between the multimodal simulation training data set corresponding to each preset training scheme and the multimodal standard training data set with a preset threshold;

[0080] If there is at least one cosine similarity value greater than a preset threshold, the cosine similarity values ​​greater than the preset threshold are sorted by size to extract the maximum cosine similarity value;

[0081] The preset training program corresponding to the maximum cosine similarity value is marked as the final training program, and the multimodal simulation training data set corresponding to the final training program is marked as the final multimodal simulation training data set; the final training program and its final multimodal simulation training data set are output.

[0082] The preset culture scheme is a series of culture plans or strategies pre-set by relevant technicians for the kidney and liver cells to be cultured. The scheme includes the settings of various conditions and operations that may be involved in the culture process of kidney and liver cells.

[0083] Among them, the culture task requirements are the goal orientation of the entire kidney and liver cell culture. It is set based on the normal physiological functions of kidney and liver cells in the body, tissue repair needs, or specific research purposes (such as building disease models, drug screening, etc.). For example, if the purpose is to build a liver disease model, the culture task requirements may be set to induce specific pathological changes in liver cells within a specific time, such as simulating abnormal deposition of extracellular matrix in the process of liver fibrosis, progressive damage to liver cell function, etc.; if it is for drug screening, the culture task requirements may include requirements for cell sensitivity and tolerance to drugs, as well as specific changes in cell function and morphology under the action of drugs.

[0084] The following steps are also included:

[0085] If all cosine similarity values ​​are not greater than the preset threshold, all preset training plans are marked as invalid training plans, and a warning message that the preset training plans need to be re-planned is generated;

[0086] The re-planned preset culture scheme is obtained, and the re-planned preset culture scheme is continued to be simulated based on the kidney and liver cell simulation culture system until the cosine similarity value between the multimodal simulation culture data set corresponding to at least one re-planned preset culture scheme and the multimodal standard culture data set is greater than a preset threshold.

[0087] It should be noted that several preset culture schemes for kidney and liver cells to be cultured are first obtained, and then these schemes are respectively imported into the kidney and liver cell simulation culture system. Based on each preset culture scheme, the system performs simulation culture. In this process, the system obtains the morphological data, functional data and proliferation data of simulated cultured cells at several preset culture time nodes, thereby generating a multimodal simulation culture data set corresponding to each preset culture scheme. The significance of this step is to comprehensively record the performance of cells in a simulation environment under different preset culture schemes from multiple angles (morphology, function, proliferation). According to the culture task requirements of the target kidney and liver cells, the standard cell morphology, function and proliferation data at each preset culture time node are obtained to generate a multimodal standard culture data set. This standard data set is a reference for measuring whether the preset culture scheme meets the requirements. It represents the state that kidney and liver cells should achieve during the culture process under ideal conditions.

[0088] The cosine similarity algorithm is introduced to calculate the similarity between the multimodal simulation training data set of each preset training program and the multimodal standard training data set. The cosine similarity algorithm can quantify the similarity between two data sets. In this case, it can be used to evaluate the degree of closeness between the performance of the preset training program in the simulation training and the ideal state. The calculated cosine similarity value is then compared with the preset threshold. If there is at least one cosine similarity value greater than the preset threshold, it means that a preset training program meets the requirements to a certain extent. These cosine similarity values ​​greater than the preset threshold are sorted by size, and the largest cosine similarity value is extracted. The corresponding preset training program is marked as the final training program, and the corresponding multimodal simulation training data set is the final multimodal simulation training data set. This process can select the program closest to the ideal state from multiple feasible preset training programs.

[0089] If the cosine similarity values ​​of all the preset training programs are not greater than the preset threshold, it means that all the preset training programs do not meet the requirements. At this time, these programs are marked as invalid training programs and a warning message is generated. Then the re-planned preset training programs are obtained and simulation processing is continued in the simulation training system until at least one preset training program has a cosine similarity value greater than the preset threshold, ensuring that a training program that meets the requirements can be obtained in the end.

[0090] Through the above steps, multiple preset culture schemes can be comprehensively and systematically evaluated and screened using the simulation culture system before kidney and liver cell culture. Through the collection of multimodal data (cell morphology, function, proliferation) and the use of the cosine similarity algorithm, the degree of proximity of each preset culture scheme to the ideal culture requirements can be accurately measured. This method can efficiently determine the final culture scheme that best meets the requirements of the culture task and its corresponding multimodal simulation culture data set from a large number of preset culture schemes, providing the best program guidance for actual kidney and liver cell culture. At the same time, preset culture schemes that do not meet the requirements can be discovered and replanned in a timely manner, ensuring the effectiveness and reliability of the final culture scheme, thereby improving the success rate, quality and fit with the expected goals of kidney and liver cell culture.

[0091] Specifically, the culture state of the kidney and liver cells at the preset culture time node is analyzed according to the final multimodal simulation culture data and the multimodal culture actual data, such as Figure 2 As shown, specifically:

[0092] S202, acquiring actual multimodal culture data of kidney and liver cells in a culture device at a preset culture time node, and obtaining actual multimodal culture data of kidney and liver cells at the preset culture time node;

[0093] S204, extracting the final multimodal simulation culture data of kidney and liver cells at the preset culture time node from the final multimodal simulation culture data set;

[0094] S206, calculating the cosine similarity value between the actual multimodal culture data of kidney and liver cells at a preset culture time node and the final multimodal simulation culture data;

[0095] S208, if the cosine similarity value between the actual multimodal culture data of the kidney and liver cells at the preset culture time node and the final multimodal simulation culture data is greater than the preset similarity threshold, the culture state of the kidney and liver cells at the preset culture time node is marked as a normal state;

[0096] S210. If the cosine similarity value between the actual multimodal culture data of the kidney and liver cells at the preset culture time node and the final multimodal simulation culture data is not greater than the preset similarity threshold, the culture state of the kidney and liver cells at the preset culture time node is marked as an abnormal state.

[0097] It should be noted that, first, the actual multimodal culture data of kidney and liver cells are obtained from the culture equipment at the preset culture time node, and these actual data reflect the status of cells in the real culture environment. At the same time, the final multimodal simulation culture data corresponding to the preset culture time node are extracted from the final multimodal simulation culture data set, which is the data under ideal conditions obtained by simulation culture before.

[0098] If the calculated cosine similarity value is greater than the preset similarity threshold, it means that the actual culture situation is close to the simulated culture situation. At this time, the culture state of the kidney and liver cells at the preset culture time node is marked as normal, which means that the performance of the cells in the actual culture process in terms of morphology, function, proliferation, etc. is in line with expectations, and the culture process proceeds as planned. On the contrary, if the cosine similarity value is not greater than the preset similarity threshold, it indicates that there is a large difference between the actual culture situation and the simulated culture situation, and the culture state of the kidney and liver cells is marked as abnormal, which may imply that there are some problems in the actual culture process, such as deviations in the culture environment or interference of cells by unknown factors. This method can timely discover problems in the actual culture process, so that corresponding measures can be taken in time during the cell culture process (such as regulating and processing the culture equipment or adjusting the culture scheme, etc.), thereby improving the success rate and quality of kidney and liver cell culture, ensuring that the cell culture process develops in the expected direction as much as possible, and helping to better understand the differences between actual culture and simulated culture, providing a reference for further optimizing the culture scheme.

[0099] Specifically, if the culture state of the kidney and liver cells is abnormal at the preset culture time node, the culture device is regulated and processed, specifically:

[0100] If the culture state of the kidney and liver cells is abnormal at the preset culture time node, then obtaining various actual working parameters of the culture device at the current preset culture time node based on various preset sensors in the culture device;

[0101] and obtaining various preset working parameters of the cultivation equipment at the current preset cultivation time node in the final cultivation plan;

[0102] Calculate the difference between various actual working parameters of the cultivation equipment at the current preset cultivation time node and the corresponding preset working parameters;

[0103] Compare the difference between each actual working parameter and the corresponding preset working parameter with the preset difference threshold value respectively;

[0104] If the difference between a certain actual working parameter and the corresponding preset working parameter is greater than the preset difference threshold, the actual working parameter is marked as an abnormal working parameter, and the offset amplitude of the abnormal working parameter is determined according to the difference between the actual working parameter and the corresponding preset working parameter, so as to obtain the abnormal working parameter and its offset amplitude of the cultivation equipment at the current preset cultivation time node;

[0105] If the difference between a certain actual working parameter and the corresponding preset working parameter is not greater than the preset difference threshold, the actual working parameter is marked as a normal working parameter.

[0106] It should be noted that when the culture state of kidney and liver cells is detected to be abnormal at the preset culture time node, the actual working parameters of the current preset culture time node are first obtained from various preset sensors in the culture device. These actual working parameters reflect the current actual operating conditions of the device, such as the actual temperature value obtained by the temperature sensor, the actual pH value obtained by the pH sensor, etc. At the same time, the preset working parameters at the current preset culture time node are obtained from the final culture program. These preset working parameters are ideal parameter values ​​set according to the optimal culture program previously screened.

[0107] If the difference between an actual working parameter and the corresponding preset working parameter is greater than the preset difference threshold, the actual working parameter is marked as an abnormal working parameter. And the offset amplitude of the abnormal working parameter is determined according to the difference between the actual working parameter and the corresponding preset working parameter. If the difference between an actual working parameter and the corresponding preset working parameter is not greater than the preset difference threshold, the actual working parameter is marked as a normal working parameter, indicating that the parameter is within an acceptable range and does not need to be adjusted.

[0108] Specifically, if the culture state of the kidney and liver cells is abnormal at the preset culture time node, the culture device is regulated and processed, and the following steps are also included:

[0109] Formulate in advance the control measures corresponding to the various deviation amplitude ranges of various working parameters of the culture equipment during operation;

[0110] Constructing a knowledge index library, importing the control measures corresponding to various deviation amplitude ranges of various working parameters of the cultivation equipment during the working process into the knowledge index library; and regularly updating the knowledge index library;

[0111] Obtaining abnormal working parameters and offset amplitudes of the cultivation equipment at the current preset cultivation time node;

[0112] Importing the abnormal working parameters and the offset amplitudes of the cultivation equipment at the current preset cultivation time node into the knowledge index library for matching, and obtaining corresponding control measures;

[0113] The acquired corresponding control measures are transmitted to the control terminal of the culture device, so that the control terminal performs control processing on the abnormal working parameters of the culture device based on the corresponding control measures.

[0114] It should be noted that first, control measures corresponding to different offset amplitude ranges of various working parameters of the culture equipment during operation should be formulated in advance. A knowledge index library is constructed and these control measures are imported into it, and the knowledge index library is updated regularly. This knowledge index library is like a knowledge base for dealing with abnormal equipment conditions. Regular updates can ensure that the control measures in it adapt to new equipment conditions, research findings, or technical improvements.

[0115] After obtaining the abnormal working parameters and their offset amplitudes of the culture equipment at the current preset culture time node, they are imported into the knowledge index library for matching. This step uses the search function of the knowledge index library to find the corresponding control measures based on the actual abnormal parameters and offset amplitudes. For example, if the abnormal working parameters and their offset amplitudes of the pH value are found, by searching in the knowledge index library, specific control measures for the abnormal pH value can be obtained.

[0116] The corresponding control measures obtained are transmitted to the control terminal of the culture device, so that the control terminal can control and process the abnormal working parameters of the device based on these control measures. This ensures that the device can be adjusted according to the pre-established strategy to correct the abnormal working parameters, thereby improving the cell culture environment and improving the success rate and quality of kidney and liver cell culture.

[0117] Through the above steps, during the culture process of kidney and liver cells, when the culture state is detected to be abnormal, the culture equipment can be quickly and accurately regulated. By pre-establishing control measures and building a knowledge index library, the function of automatically matching and obtaining corresponding control measures according to the actual abnormal working parameters and their offset amplitudes is realized, avoiding the uncertainty and time-consuming nature of artificial temporary formulation of control strategies. This helps to correct abnormal parameters of cell culture equipment in a timely and effective manner, optimize the cell culture environment, ensure that kidney and liver cell culture can proceed smoothly according to the expected plan, improve the stability and reliability of cell culture, and the regular update of the knowledge index library can adapt to the evolving needs of cell culture technology.

[0118] In addition, the method further comprises the following steps:

[0119] Obtaining a fault event log of the culture device, and obtaining various fault operating condition events of the culture device in the fault event log;

[0120] and obtaining equipment working parameters of the training equipment when various fault conditions occur, and obtaining environmental characteristic parameters of the training equipment when various fault conditions occur;

[0121] A graph structure is constructed based on the equipment working parameters and environmental characteristic parameters of the training equipment when various fault conditions occur, wherein the equipment working parameters and environmental characteristic parameters are used as nodes in the graph, and the connection relationship between the nodes represents the correlation between the parameters;

[0122] The graph neural network is used to extract features from the constructed graph structure. The dependency relationship and global features between nodes are captured through the propagation mechanism of the graph neural network, and the correlation features between different fault conditions and equipment working parameters and environmental characteristic parameters are obtained.

[0123] Generate a correlation feature matrix based on correlation features between different fault conditions and equipment operating parameters and environmental characteristic parameters;

[0124] Constructing a prediction model, importing the associated feature matrix into the prediction model for coding learning, and outputting a culture equipment failure prediction model when all model parameters of the prediction model meet preset requirements;

[0125] After the culture device is regulated, the culture status of the kidney and liver cells is continuously obtained at the next preset culture time node;

[0126] If the culture state of the kidney and liver cells is still abnormal at the next preset culture time node, obtaining actual equipment working parameters and actual environmental characteristic parameters of the culture equipment;

[0127] Importing actual equipment working parameters and actual environmental characteristic parameters of the culture equipment into the culture equipment fault prediction model for prediction;

[0128] The failure probability of the culture device is obtained through prediction; if the failure probability is greater than a preset probability value, a failure warning message is generated, and a cell storage protection mechanism of the culture device is started.

[0129] It should be noted that the equipment operating parameters (such as temperature, pH value, gas composition, etc.) and environmental characteristic parameters (such as light intensity, humidity, etc.) are obtained when each fault condition event occurs. These parameters serve as the basic data for subsequent analysis. The equipment operating parameters and environmental characteristic parameters are constructed into a graph structure, in which the equipment operating parameters and environmental characteristic parameters are used as nodes in the graph, and the connection relationship between the nodes represents the correlation between the parameters. This graph structure can intuitively display the relationship between the parameters, which is convenient for subsequent feature extraction and analysis.

[0130] Through the above steps, during the culture process, it is possible to construct a graph structure using fault event logs and parameter data, extract features through graph neural networks, generate a correlation feature matrix, and then build a fault prediction model. This model can accurately predict the failure probability of the equipment, and generate early warning information when the failure probability exceeds the preset value, and activate the protection mechanism, so as to timely discover and deal with potential failure risks, improve the success rate and stability of cell culture, and ensure the reliability and safety of the cell culture process.

[0131] In addition, the method further comprises the following steps:

[0132] Acquire actual cell proliferation data of kidney and liver cells in several sub-culture areas during the culture process at preset culture time points;

[0133] Introducing a Gaussian process regression algorithm, and performing regression calculation on the actual cell proliferation data of each sub-culture area obtained based on the Gaussian process regression algorithm to obtain an actual dynamic regression curve of the actual cell proliferation data of each sub-culture area of ​​kidney and liver cells;

[0134] Acquiring a culture task requirement for kidney and liver cells, and acquiring a preset cell proliferation data range value for each sub-culture area at a preset culture time point according to the culture task requirement;

[0135] Defining a range domain in the actual dynamic regression curve according to a preset cell proliferation data range value of each sub-culture area;

[0136] Calculate the total length of the segment of the actual dynamic regression curve outside the range domain and the total length of the segment of the actual dynamic regression curve within the range domain;

[0137] The total length of the actual dynamic regression curve outside the range is processed by ratio with the total length of the actual dynamic regression curve within the range, so as to obtain the proliferation deviation of kidney and liver cells;

[0138] It is determined whether the proliferation deviation is greater than a preset deviation value, and if so, a fault warning message is generated.

[0139] It should be noted that the proliferation deviation can be used to evaluate the difference between the actual performance of each sub-culture area at the preset culture time point and the expected target. When the proliferation deviation is large, it means that there is a large deviation between the actual cell proliferation data and the preset range value, and it is necessary to adjust the culture conditions or take other measures to optimize the cell proliferation process.

[0140] Through the above steps, real-time monitoring and evaluation of cell proliferation during kidney and liver cell culture can be achieved. Through the Gaussian process regression algorithm, a dynamic regression curve of cell proliferation in each sub-culture area can be obtained, thereby capturing the trend and changes of cell proliferation. When the proliferation deviation exceeds the preset threshold, the system will generate fault warning information to help the operator promptly discover and deal with problems in the culture process, ensuring the stability and success rate of the cell culture process. This method improves the controllability and reliability of the cell culture process, helps optimize the culture conditions, and improves the quality and efficiency of cell culture.

[0141] like Figure 3 As shown, the present invention also discloses a kidney and liver cell culture system 6 based on multimodal information, wherein the kidney and liver cell culture system comprises a memory 41 and a processor 52, wherein the memory 41 stores a kidney and liver cell culture method program, and when the kidney and liver cell culture method program is executed by the processor 52, any one of the steps of the kidney and liver cell culture method is implemented.

[0142] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for culturing kidney and liver cells based on multimodal information, characterized in that: The following steps are involved: Obtaining several preset culture schemes for kidney and liver cells to be cultured, importing each preset culture scheme into a kidney and liver cell simulation culture system for simulation culture, and screening out a final culture scheme and a final multimodal simulation culture dataset based on the multimodal simulation culture datasets of each preset culture scheme; Controlling the culture equipment to actually culture the kidney and liver cells to be cultured according to the final culture plan; In the actual culture process, actual multimodal culture data of kidney and liver cells in the culture device are obtained at a preset culture time node, wherein the actual multimodal culture data includes actual cell morphology data, actual cell function data, and actual cell proliferation data; Extracting the final multimodal simulation culture data corresponding to the preset culture time node from the final multimodal simulation culture data set; analyzing the culture state of kidney and liver cells at the preset culture time node according to the final multimodal simulation culture data and the multimodal culture actual data; If the culture state of the kidney and liver cells is normal at the preset culture time point, no regulation or control is performed on the culture device; If the culture state of the kidney and liver cells is abnormal at the preset culture time point, regulating and processing the culture equipment; The steps of the method for constructing the kidney and liver cell simulation culture system are as follows: Acquire target multi-source biological data through a big data network, wherein the target multi-source biological data includes tissue samples of kidney and liver cells, gene sequences, protein expression profiles, and metabolites; Extracting features from the target multi-source biological data to obtain key feature information of kidney and liver cells in a physiological state, wherein the key feature information includes structural composition, functional units, intercellular communication mechanisms, material and energy metabolism rules, and growth and differentiation logic; According to the key feature information, based on biological principles, the structural composition is constructed as a basic framework, the functional units are set as modules, the signal transduction pathway is established according to the intercellular communication mechanism, the material exchange and conversion process is set according to the material and energy metabolism rules, and the cell state transition rules are set according to the growth and differentiation logic to initialize the cell culture simulation model; Obtaining a functional specification of the culture equipment, and setting a working range of physical and chemical environmental parameters in the cell culture simulation model according to the functional specification, wherein the working range of the physical and chemical environmental parameters includes the type and concentration gradient of nutrients, temperature range, pH fluctuation, and gas exchange conditions; Obtaining the initial state of the kidney and liver cells to be cultured, and setting the type, quantity, distribution pattern and initial functional state of the cells in the cell culture simulation model according to the initial state of the kidney and liver cells to be cultured; After the setting is completed, the model parameters of the cell culture simulation model are saved, and the cell culture simulation model is connected and integrated with external input and output devices to form a kidney and liver cell simulation culture system.

2. The method for culturing kidney and liver cells based on multimodal information according to claim 1, characterized in that: Each preset culture scheme is introduced into the kidney and liver cell simulation culture system for simulation culture, and the final culture scheme and the final multimodal simulation culture data set are screened out according to the multimodal simulation culture data set of each preset culture scheme, specifically: Obtaining several preset culture schemes for kidney and liver cells to be cultured, and introducing each preset culture scheme into the kidney and liver cell simulation culture system; Based on each preset culture scheme, the kidney and liver cell simulation culture system is controlled to perform simulation culture work, and the simulated culture cell morphology data, simulated culture cell function data and simulated culture cell proliferation data of the kidney and liver cells in the kidney and liver cell simulation culture system after culture by each preset culture scheme at several preset culture time nodes are obtained to generate a multimodal simulation culture data set corresponding to each preset culture scheme; Obtaining the culture task requirements of the target kidney and liver cells, obtaining the standard cell morphology data, standard cell function data, and standard cell proliferation data of the kidney and liver cells at each preset culture time node according to the culture task requirements, and generating a multimodal standard culture data set corresponding to the culture target kidney and liver cells; Introducing a cosine similarity algorithm, and calculating the cosine similarity value between the multimodal simulation training data set corresponding to each preset training scheme and the multimodal standard training data set based on the cosine similarity algorithm; Comparing the cosine similarity value between the multimodal simulation training data set corresponding to each preset training scheme and the multimodal standard training data set with a preset threshold; If there is at least one cosine similarity value greater than a preset threshold, the cosine similarity values ​​greater than the preset threshold are sorted by size to extract the maximum cosine similarity value; The preset training program corresponding to the maximum cosine similarity value is marked as the final training program, and the multimodal simulation training data set corresponding to the final training program is marked as the final multimodal simulation training data set; the final training program and its final multimodal simulation training data set are output.

3. The method for culturing kidney and liver cells based on multimodal information according to claim 2, characterized in that: The following steps are also included: If all cosine similarity values ​​are not greater than the preset threshold, all preset training plans are marked as invalid training plans, and a warning message that the preset training plans need to be re-planned is generated; The re-planned preset culture scheme is obtained, and the re-planned preset culture scheme is continued to be simulated based on the kidney and liver cell simulation culture system until the cosine similarity value between the multimodal simulation culture data set corresponding to at least one re-planned preset culture scheme and the multimodal standard culture data set is greater than a preset threshold.

4. The method for culturing kidney and liver cells based on multimodal information according to claim 1, characterized in that: The culture state of the kidney and liver cells at the preset culture time node is analyzed according to the final multimodal simulation culture data and the multimodal culture actual data, specifically: Acquiring actual multimodal culture data of kidney and liver cells in a culture device at a preset culture time node, and obtaining actual multimodal culture data of kidney and liver cells at the preset culture time node; Extracting the final multimodal simulation culture data of kidney and liver cells at the preset culture time node from the final multimodal simulation culture data set; Calculate the cosine similarity value between the actual multimodal culture data of kidney and liver cells at the preset culture time node and the final multimodal simulation culture data; If the cosine similarity value between the actual multimodal culture data of the kidney and liver cells at the preset culture time node and the final multimodal simulation culture data is greater than the preset similarity threshold, the culture state of the kidney and liver cells at the preset culture time node is marked as a normal state; If the cosine similarity value between the actual multimodal culture data of the kidney and liver cells at the preset culture time node and the final multimodal simulation culture data is not greater than the preset similarity threshold, the culture state of the kidney and liver cells at the preset culture time node will be marked as an abnormal state.

5. The method for culturing kidney and liver cells based on multimodal information according to claim 4, characterized in that: If the culture state of the kidney and liver cells is abnormal at the preset culture time point, the culture equipment is regulated and processed, specifically: If the culture state of the kidney and liver cells is abnormal at the preset culture time node, then obtaining various actual working parameters of the culture device at the current preset culture time node based on various preset sensors in the culture device; and obtaining various preset working parameters of the cultivation equipment at the current preset cultivation time node in the final cultivation plan; Calculate the difference between various actual working parameters of the cultivation equipment at the current preset cultivation time node and the corresponding preset working parameters; Compare the difference between each actual working parameter and the corresponding preset working parameter with the preset difference threshold value respectively; If the difference between a certain actual working parameter and the corresponding preset working parameter is greater than the preset difference threshold, the actual working parameter is marked as an abnormal working parameter, and the offset amplitude of the abnormal working parameter is determined according to the difference between the actual working parameter and the corresponding preset working parameter, so as to obtain the abnormal working parameter and its offset amplitude of the cultivation equipment at the current preset cultivation time node; If the difference between a certain actual working parameter and the corresponding preset working parameter is not greater than the preset difference threshold, the actual working parameter is marked as a normal working parameter.

6. The method for culturing kidney and liver cells based on multimodal information according to claim 5, characterized in that: If the culture state of the kidney and liver cells is abnormal at the preset culture time point, the culture device is regulated, and the following steps are also included: Formulate in advance the control measures corresponding to the various deviation amplitude ranges of various working parameters of the culture equipment during operation; Constructing a knowledge index library, importing the control measures corresponding to various deviation amplitude ranges of various working parameters of the cultivation equipment during the working process into the knowledge index library; and regularly updating the knowledge index library; Obtaining abnormal working parameters and offset amplitudes of the cultivation equipment at the current preset cultivation time node; Importing the abnormal working parameters and the offset amplitudes of the cultivation equipment at the current preset cultivation time node into the knowledge index library for matching, and obtaining corresponding control measures; The acquired corresponding control measures are transmitted to the control terminal of the culture device, so that the control terminal performs control processing on the abnormal working parameters of the culture device based on the corresponding control measures.

7. A kidney and liver cell culture system based on multimodal information, characterized in that: The kidney and liver cell culture system includes a memory and a processor, wherein a kidney and liver cell culture method program is stored in the memory, and when the kidney and liver cell culture method program is executed by the processor, the kidney and liver cell culture method according to any one of claims 1 to 6 is implemented.

Images