A stem cell culture density optimization control method and system

Abstract

The application relates to a stem cell culture density optimization control method and system, and relates to the technical field of stem cell culture, and comprises the following steps: K culture boxes and P culture dishes are configured; when mesenchymal stem cells are cultured, the P culture dishes are randomly and equally distributed to the K culture boxes; P cell imaging sequences in a historical time window are monitored; P real-time cell proliferation rates and morphological characteristics are obtained through analysis; the P culture dishes are similar clustered based on the same; K similar culture dish sets and K cell proliferation rate averages are determined; the K similar culture dish sets are mapped and placed in the K culture boxes; and the K culture boxes are respectively optimized and adjusted in culture environment parameters based on the K cell proliferation rate averages, so that K adaptive environment parameters are obtained, and the K culture boxes in a future time window are controlled. The application solves the problems that in traditional stem cell culture, a unified culture environment is adopted, targeted culture is lacked, and the culture density control precision is low and the cell quality is unstable.

Description

Technical Field

[0001] This application relates to the field of stem cell culture, and in particular to a method and system for optimizing and controlling stem cell culture density. Background Technology

[0002] With the development of stem cell culture technology, precise control of cell culture density is crucial for ensuring cell quality and has become a key technical challenge in the field of stem cell culture. Currently, traditional stem cell culture methods use a uniform environment to manage cells in different culture dishes, which makes it difficult to adapt to differences in cell proliferation states caused by individual cell differences and surface coating uniformity. This can easily lead to a mismatch between the culture environment and cell requirements, which not only reduces the efficiency and quality stability of stem cell culture but also increases the uncertainty and cost of subsequent cell applications. Summary of the Invention

[0003] To address the aforementioned technical problems, this application provides a method and system for optimizing and controlling stem cell culture density, which improves upon the current situation in traditional stem cell culture where the lack of targeted environmental regulation leads to low precision in culture density control and unstable cell quality.

[0004] The embodiments of this application disclose the following technical solutions:

[0005] In a first aspect, embodiments of this application provide a method for optimizing and controlling stem cell culture density, the method comprising:

[0006] Configure K incubators and P culture dishes, wherein the culture environment of each incubator can be controlled independently, K is an integer greater than or equal to 2, P is an integer greater than or equal to 12, and P is an integer multiple of K;

[0007] During the mesenchymal stem cell culture process, the P culture dishes are randomly and equally distributed to the K incubators for stem cell culture. The imaging sequences of the P cells in the P culture dishes within the historical time window are monitored and obtained, and the proliferation rate and morphological characteristics of the P cells in real time are analyzed.

[0008] Based on the P real-time cell proliferation rates and P real-time cell morphology characteristics, similarity clustering is performed on the P culture dishes to determine K similar culture dish sets and K average cell proliferation rates;

[0009] The K similar culture dish sets are mapped and placed in the K incubators, and the culture environment parameters of the K incubators are optimized and adjusted based on the average cell proliferation rate of the K cells to obtain K suitable environment parameters. The environment of the K incubators is then managed within a future time window.

[0010] Secondly, embodiments of this application provide a stem cell culture density optimization control system, the system comprising:

[0011] The equipment configuration module is used to configure K incubators and P culture dishes, wherein the culture environment of each incubator can be controlled independently, K is an integer greater than or equal to 2, P is an integer greater than or equal to 12, and P is an integer multiple of K.

[0012] The monitoring and analysis module is used to randomly and equally distribute the P culture dishes to the K incubators for stem cell culture during the mesenchymal stem cell culture process, monitor and acquire the P cell imaging sequences of the P culture dishes within a historical time window, and analyze the P real-time cell proliferation rate and P real-time cell morphology characteristics.

[0013] The cluster mean module is used to perform similar clustering on the P culture dishes based on the P real-time cell proliferation rates and P real-time cell morphology characteristics, to determine a set of K similar culture dishes and the mean of K cell proliferation rates;

[0014] The environment optimization and control module is used to map and place the K similar culture dish sets into the K incubators, and optimize and adjust the culture environment parameters of the K incubators based on the average cell proliferation rate of the K cells to obtain K suitable environment parameters, and control the environment of the K incubators within a future time window.

[0015] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0016] This application proposes a method and system for optimizing and controlling stem cell culture density. By configuring incubators and culture dishes with independently controllable environments, combining cell imaging analysis of cell proliferation rate and morphological characteristics, similar cell clustering is performed. Then, based on the average cell proliferation rate of each group, the culture environment parameters are optimized and adjusted collaboratively to achieve precise control of stem cell culture density. First, K incubators and P culture dishes are configured, and the culture dishes are randomly distributed in the incubators. P cell imaging sequences are periodically captured using a fluorescence microscope, and real-time cell proliferation rate and morphological characteristics are analyzed. Based on these characteristics, the culture dishes are clustered to obtain K similar culture dish sets and their corresponding average cell proliferation rates. Then, culture dishes from the same set are placed in the same incubator. Based on the difference between the average cell proliferation rate and the current standard cell proliferation rate, environmental parameters such as temperature, humidity, carbon dioxide concentration, and oxygen concentration are optimized and adjusted to obtain suitable environmental parameters and manage the incubator environment within a future time window.

[0017] The technical solution of this application solves the problems of inconsistent cell states and low density control precision caused by uniform environmental control in traditional stem cell culture by integrating features acquisition from cell imaging analysis, similar clustering with dynamic weight adjustment, environmental parameter optimization based on proliferation deviation, and incubator environment regulation with group control. It realizes differentiated and precise regulation of the culture environment, and provides technical support for ensuring the quality and density optimization of stem cell culture. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A schematic flowchart illustrating a method for optimizing and controlling stem cell culture density, provided in an embodiment of this application;

[0020] Figure 2 This is a schematic diagram of a stem cell culture density optimization control system provided in an embodiment of this application.

[0021] The components represented by each number in the attached diagram are explained below:

[0022] Equipment configuration module 01, monitoring and analysis module 02, cluster mean module 03, and environmental optimization and control module 04. Detailed Implementation

[0023] This application provides a method and system for optimizing and controlling stem cell culture density, which addresses the technical problems in the prior art where factors such as individual cell differences and surface coating uniformity lead to different cell proliferation states under the same initial conditions, yet a uniform culture environment is used, resulting in a lack of targeted regulation, and consequently, low precision in culture density control and unstable cell quality.

[0024] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0025] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0026] In the description of this application, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0027] Example 1, as shown in the appendix Figure 1 As shown, this application provides a method for optimizing and controlling the density of stem cell culture, the method comprising the following steps:

[0028] S110: Configure K incubators and P culture dishes, wherein the culture environment of each incubator can be controlled independently, K is an integer greater than or equal to 2, P is an integer greater than or equal to 12, and P is an integer multiple of K;

[0029] In this embodiment of the application, in order to achieve precise and independent control of different culture environments during stem cell culture and to adapt to the differentiated culture needs brought about by cell differences, it is necessary to reasonably configure incubators and culture dishes that can individually adjust environmental parameters.

[0030] In the method provided in this application embodiment, "the incubator can independently control the culture environment", wherein the controllable culture environment parameters include at least temperature, humidity, carbon dioxide concentration and oxygen concentration.

[0031] Specifically, stem cell culture requires stringent stability and controllability of culture environment parameters; the incubator must have the function of individually controlling temperature, humidity, carbon dioxide concentration, and oxygen concentration.

[0032] Among these, carbon dioxide concentration is achieved through reaction with water to form carbonic acid, releasing or absorbing H₂. +Maintaining a stable pH in the culture medium (e.g., in most cases, a 5% carbon dioxide concentration needs to be maintained to ensure the pH of the culture medium is suitable) is one of the key environmental conditions for the normal proliferation of stem cells.

[0033] Furthermore, based on this functional requirement, the incubator and petri dishes are configured.

[0034] Specifically, the first step is to configure incubators. Based on the requirements of stem cell culture for multiple independent environments, the number of incubators, K, is determined to be an integer greater than or equal to 2. Equipment with high-precision environmental control functions is preferred, with temperature control accuracy of ±0.1℃, humidity ±2%RH, carbon dioxide concentration ±0.1%, and oxygen concentration ±0.2%, to ensure the accuracy and stability of environmental parameter adjustment.

[0035] For example, if two sets of environmental comparison experiments are to be carried out, K=2 incubators can be configured, labeled as incubator A and incubator B respectively, and connected to the laboratory intelligent control system. Basic environmental parameters such as 37℃ (suitable temperature for stem cells) and 5% carbon dioxide concentration (conventional culture conditions) are pre-stored. After power-on, the internal sensors are automatically calibrated so that the deviation of the initial environmental parameters is controlled within the allowable range.

[0036] Secondly, the culture dishes were prepared. Based on the experimental requirements for sample size and uniform distribution, the number of culture dishes P was determined to be greater than or equal to 12 and an integer multiple of K. At the same time, culture dishes with good surface coating uniformity and material suitable for stem cell adherence growth were selected. The coating roughness Ra was ensured to be ≤10nm by atomic force microscopy to reduce the impact of differences in culture dishes on cell proliferation.

[0037] For example, when K=2, configure P=12 culture dishes, randomly number them D1-D12, seal them in aseptic packaging, and place them in a clean bench for later use.

[0038] Furthermore, after configuring the incubator and petri dishes, overall adaptation and debugging are required.

[0039] Specifically, P culture dishes were grouped into K groups and placed into their respective incubators. The incubator simulation program was started and run for 24 hours in a preset environment of 37°C, 5% carbon dioxide concentration, and 95% humidity. At the same time, the microenvironmental parameters around each culture dish were monitored (using miniature sensors built into the culture dish) to ensure that the environmental differences between different culture dishes in the same incubator were ≤ temperature ±0.05°C and carbon dioxide concentration ±0.05%. This verified the incubator's ability to control the environmental uniformity of multiple culture dishes and laid a stable hardware foundation for subsequent stem cell seeding and culture.

[0040] Ultimately, by precisely configuring K incubators and P culture dishes with independently controllable environments, a hardware platform for stem cell culture was constructed, providing a stable and comparable experimental environment for subsequent steps such as "random isopleial distribution of culture dishes, cell imaging monitoring, cluster analysis, and environmental optimization control".

[0041] S120: During the mesenchymal stem cell culture process, the P culture dishes are randomly and equally distributed to the K incubators for stem cell culture. The imaging sequences of the P cells in the P culture dishes within the historical time window are monitored and obtained. The proliferation rate and morphological characteristics of the P cells in real time are analyzed.

[0042] In this embodiment of the application, in scenarios where the proliferation status of stem cells is difficult to control precisely due to individual cell differences and surface coating uniformity, cell characteristic analysis is required in order to accurately obtain cell proliferation and morphology data to support subsequent culture regulation.

[0043] Specifically, within a historical time window, cell images are first captured using a fluorescence microscope at preset time intervals to obtain P cell imaging sequences.

[0044] The preset time interval is set according to the characteristics of the mesenchymal stem cell proliferation cycle to ensure that the dynamic changes of cell proliferation can be fully captured; the fluorescence microscope must have high-resolution imaging capabilities to clearly present cell morphological details.

[0045] Furthermore, based on the historical culture records of mesenchymal stem cells, sample cell imaging sets and sample cell number label sets are collected. The sample cell imaging sets and sample cell number label sets are then input into a convolutional neural network for training until convergence, resulting in an automatic cell counter.

[0046] Simultaneously, the cell count of P cell imaging sequences is identified using this automatic cell counter, and the time dimension data is combined to calculate and output the real-time cell proliferation rate of P cells.

[0047] Furthermore, a sample cell image set and a sample cell morphology feature label set are collected and used to train a convolutional neural network until convergence, thus obtaining a cell morphology recognizer.

[0048] Simultaneously, the terminal cell images of P cell imaging sequences are extracted and input into a cell morphology recognizer to identify and output the average cell area and average cell roundness.

[0049] Furthermore, by using the statistical distance method, the position of the cell centroid is recorded, the distance between any two cell centroids is calculated, and the cell distribution uniformity is obtained by the ratio of the standard deviation to the mean distance. Finally, P real-time cell morphology features are integrated and output.

[0050] This step, by constructing an automatic cell counter and a cell morphology recognizer, combined with multi-dimensional data acquisition and analysis, accurately obtained cell proliferation rate and morphological characteristics, providing key data support for subsequent similarity clustering and environmental parameter optimization.

[0051] Step S120 in the method provided in this application embodiment includes:

[0052] Within the historical time window, cell images of the P culture dishes are periodically captured using a fluorescence microscope at preset time intervals to obtain P cell imaging sequences.

[0053] Based on the P cell imaging sequences, cell proliferation rate analysis is performed, and P real-time cell proliferation rates are output.

[0054] P terminal cell images are extracted from the P cell imaging sequences respectively, and cell morphology features are identified to output P real-time cell morphology features, including mean cell area, mean cell roundness, and cell distribution uniformity.

[0055] In this embodiment of the application, in order to accurately obtain cell proliferation and cell morphology data during the stem cell culture process, multi-dimensional cell characteristic analysis is required to support the subsequent optimization and control of stem cell culture density.

[0056] First, cell imaging sequences were acquired. Specifically, within the historical time window of mesenchymal stem cell culture, based on the characteristics of the stem cell proliferation cycle (e.g., the conventional proliferation cycle of mesenchymal stem cells is about 24-48 hours), preset time intervals (e.g., every 12 hours) were set, and cell images were captured on P culture dishes using a fluorescence microscope.

[0057] The fluorescence microscope must have high-resolution imaging capabilities (such as a resolution of 0.1 μm) and be equipped with autofocus and automatic exposure compensation functions to ensure that cell images taken at different culture dishes and at different time points are clear and of consistent quality.

[0058] Finally, by standardizing the operation of the fluorescence microscope and unifying the imaging parameters at each time point, and taking pictures of each of the P culture dishes, and after data processing and storage, we obtained the imaging sequences of P cells covering the entire historical time window, providing basic visual data for subsequent analysis.

[0059] For example, if the historical time window is set to 72 hours (covering the entire stage of stem cell proliferation) and the preset time interval is 12 hours, then each culture dish needs to be photographed at 0 hours, 12 hours, 24 hours, 36 hours, 48 ​​hours, 60 hours and 72 hours.

[0060] Furthermore, if P=12, during each image capture, the culture dish is placed on the stage of a fluorescence microscope, the focal length is adjusted to clearly show the cell community, and the same fluorescence excitation intensity and exposure time are used to capture images of 12 culture dishes in sequence, acquiring 7 images of each culture dish to form P (12) cell imaging sequences. Each sequence contains 7 cell images at different time points, thus preserving complete visual data on the dynamic proliferation of cells for subsequent analysis.

[0061] Furthermore, by training a convolutional neural network using a sample cell imaging set combined with a sample cell number label set and a sample cell morphology feature label set, an automatic cell counter and a cell morphology recognizer were constructed to obtain cell proliferation rate and morphological features, providing data support for subsequent similarity clustering of culture dishes and optimization of culture environment parameters.

[0062] The method provided in this application also includes:

[0063] Based on the historical culture records of mesenchymal stem cells, a sample cell imaging set, a sample cell number label set, and a sample cell morphology characteristic label set were collected.

[0064] Using the sample cell imaging set and sample cell number label set, a convolutional neural network is trained until convergence to obtain an automatic cell counter for cell number identification, and the cell proliferation rate is calculated based on the P cell number identification results.

[0065] Using the sample cell imaging set and the sample cell morphology feature label set, a convolutional neural network is trained until convergence to obtain a cell morphology recognizer for cell morphology feature recognition.

[0066] In this embodiment of the application, in order to achieve automated and accurate identification of cell number and morphological characteristics, it is necessary to train a dual convolutional neural network based on sample data such as sample cell imaging set, sample cell number label set, and sample cell morphological feature label set collected from the historical culture records of mesenchymal stem cells. This will result in an automatic cell counter that can accurately identify the number of cells and a cell morphology recognizer that can extract the average cell area, average roundness, and distribution uniformity, providing data support for subsequent cell proliferation rate analysis and similarity clustering.

[0067] Specifically, cell proliferation rate analysis was performed first. Based on historical culture records of mesenchymal stem cells, a large-scale sample cell imaging set and a sample cell number label set were collected.

[0068] The sample cell imaging set includes images of stem cells at different proliferation stages and under different culture conditions; the sample cell number label set is the number of cells in the corresponding images labeled by a high-precision cell counter.

[0069] For example, the sample cell imaging set may include stem cell images at 6 hours post-inoculation (lag phase), 36 hours post-inoculation (exponential phase), and 72 hours post-inoculation (stationary phase), and cover different culture conditions such as temperatures of 36°C, 37°C, and 38°C, and carbon dioxide concentrations of 4%, 5%, and 6%.

[0070] In addition, the sample cell number label set corresponds to the number of cells in the above images. For example, an image under 37°C culture conditions during a certain exponential phase was verified to have 208 cells by a high-precision cell counter and was ultimately labeled as having 208 cells.

[0071] Based on this, the sample cell imaging set is used as input and the sample cell number label set is used as supervision to train the convolutional neural network until convergence, thus obtaining an automatic cell counter.

[0072] Specifically, for the cell automatic counter, a network architecture was built that includes a convolutional layer with a 3×3 kernel, a max pooling layer with a stride of 2×2, and a fully connected layer. The sample cell imaging set was divided into a training set, a validation set, and a test set in a 7:2:1 ratio. The sample cell number label set was used as a supervision signal, and the network parameters were optimized through the backpropagation algorithm.

[0073] Meanwhile, the initial learning rate is set to 0.001. Every 50 iterations, the learning rate is reduced to 1 / 10 of the previous rate. If the mean squared error of the validation set decreases by less than 0.01 for 10 consecutive iterations, the model is considered to have converged and the automatic cell counter is completed.

[0074] After training, the average deviation between the predicted cell number and the actual label in the test set is ≤5%. For example, the number of cells labeled in the imaging under the 37℃ culture condition during a certain exponential phase is 209, and the prediction result of the automatic cell counter is stable in the range of 205-213, which meets the accuracy requirements.

[0075] Furthermore, using the sample cell imaging set as input and the sample cell morphology feature label set as supervision, a cell morphology recognizer is constructed using a homogeneous network architecture.

[0076] The sample cell morphology feature label set includes not only the mean cell area and mean roundness, but also the cell distribution uniformity based on the statistical distance method. That is, the position of the centroid of all cells in the image is recorded, the distance between any two cell centroids is calculated, the standard deviation of the distance σ and the mean distance μ are obtained, and "cell distribution uniformity = σ / μ" is used as the label value.

[0077] For example, in a sample cell imaging, the average cell area labeled by a high-precision cell counter was 75 μm. 2The mean roundness is 0.68. After recording the centroid positions of 50 cells within the image, the standard deviation σ of the distance between any two cell centroids is calculated to be 7.8 μm and the mean distance μ is 22.3 μm. Therefore, the cell distribution uniformity label value is 7.8 / 22.3≈0.35. These data together constitute the cell morphology feature label of the sample, which is used for training supervision of the cell morphology recognizer.

[0078] Building upon this, a convolutional neural network architecture identical to that of an automatic cell counter is constructed, comprising an input layer, convolutional layers, pooling layers, fully connected layers, and an output layer. The convolutional layers use 3×3 convolutional kernels to extract local features related to cell morphology, the pooling layers compress the feature dimension through max pooling, and the fully connected layers map high-dimensional features to predicted values ​​of cell morphology features.

[0079] During training, the prediction errors of the three main features are optimized simultaneously to ensure that the mean deviation of cell area is ≤ ±5%, the mean deviation of roundness is ≤ ±0.05, and the deviation of cell distribution uniformity is ≤ ±0.1.

[0080] Furthermore, by co-training the sample cell imaging set with the sample cell number label set and the sample cell morphology feature label set, and inputting them into the constructed automatic cell counter and cell morphology recognizer, the automatic cell counter accurately identifies the number of cells at each time point, and P real-time cell proliferation rates are obtained through curve fitting.

[0081] The cell proliferation rate was derived by using an automatic cell counter to identify the number of cells at each time point in P cell imaging sequences, obtaining a dataset showing the change in cell number over time, and then using the existing Logistic model to fit the dataset to a curve, and finally calculating the slope of the fitted curve.

[0082] For example, the number of cells in a culture dish at 0 hours, 12 hours, and 24 hours were identified as 50, 80, and 140, respectively. After fitting the curve to these data using a Logistic model, the slope of the curve at 24 hours was calculated to be 3 cells / hour, that is, the cell proliferation rate at this time point was 3 cells / hour.

[0083] Simultaneously, the cell morphology identifier extracts the mean area and mean roundness of P terminal cell images and the cell distribution uniformity calculated according to the statistical distance method (recording the centroid, calculating the centroid distance, and calculating the ratio of σ to μ), outputting P real-time cell morphology features, providing multi-dimensional and high-precision quantitative basis for cell state for subsequent similar clustering and environmental parameter optimization in stem cell culture.

[0084] For example, in the cell imaging sequence of a culture dish, the number of cells identified by an automatic cell counter was 52 at 0 hours, 68 at 12 hours, and 135 at 24 hours. The cell proliferation rate was calculated to be 0.05 cells / hour through curve fitting. The terminal cell images were extracted by a cell morphology identifier, and the average cell area was 76 μm² (instrument label 78 μm², deviation 2.6%), the average roundness was 0.69 (instrument label 0.70, deviation 1.4%), and the cell distribution uniformity was 0.36 (instrument calculation 0.35, deviation 2.9%). All indicators met the accuracy requirements, providing reliable data for subsequent classification of the culture dish into the corresponding similar set and optimization of the environmental parameters of its incubator.

[0085] S130: Based on the P real-time cell proliferation rates and P real-time cell morphology characteristics, perform similarity clustering on the P culture dishes to determine K similar culture dish sets and K average cell proliferation rates;

[0086] In this embodiment of the application, after obtaining P real-time cell proliferation rates and P real-time cell morphological characteristics, in order to achieve precise grouping and control of culture dishes, it is necessary to group culture dishes with similar states into one category through similar clustering, so as to provide a basis for subsequent differentiated environmental control.

[0087] Specifically, the first step is to clarify the multiple proliferation stages in the cell culture process and the corresponding standard cell proliferation rates.

[0088] The proliferation phase includes at least the lag phase (cells adhere to the wall and adapt, proliferation is slow, and the standard rate is low), the exponential phase (cells divide rapidly, proliferation is active, and the standard rate is high), and the stationary phase (cell density is close to saturation, proliferation slows down, and the standard rate declines).

[0089] Furthermore, the ratio of the current standard cell proliferation rate at the current proliferation stage to the sum of the standard cell proliferation rates at multiple stages is calculated. This ratio is used as the cell proliferation rate weight adjustment coefficient to correct the initial rate weight (set to 0.5) to obtain an appropriate rate weight (ranging from 0.3 to 0.7).

[0090] Meanwhile, the adaptation rate weight is subtracted from 1 to obtain the adaptation form weight, and the two are combined to form the similarity fusion weight.

[0091] Based on this, the P culture dishes are clustered using existing clustering algorithms, and the final output is a set of K similar culture dishes, with each set containing the same number of culture dishes (P / K).

[0092] Finally, the average of the P / K real-time cell proliferation rates in the K similar culture dish sets was obtained to obtain the K average cell proliferation rates. This average reflects the overall proliferation level of each culture dish set, providing a quantitative benchmark for subsequent adjustment of culture environment parameters for different sets.

[0093] This step dynamically adjusts the similarity fusion weights to make the clustering results more closely match the characteristics of the current cell proliferation stage, ensuring consistency in core features among similar culture dishes and laying the foundation for precise and differentiated optimization of the culture environment.

[0094] Step S130 in the method provided in this application embodiment includes:

[0095] Multiple proliferation phases during cell culture and multiple standard cell proliferation rates for the multiple proliferation phases are obtained, wherein the proliferation phase includes at least a lag phase, an exponential phase, and a stationary phase;

[0096] Similarity fusion weights are configured based on the current standard cell proliferation rate at the current proliferation stage and the proliferation rates of the multiple standard cells;

[0097] Based on the similarity fusion weight, the P culture dishes are clustered similarly according to the P real-time cell proliferation rates and P real-time cell morphology features, and K similar culture dish sets are output, wherein the number of culture dishes in each similar culture dish set is the same.

[0098] The average proliferation rates of P / K real-time cells in the K similar culture dish sets are calculated to obtain the average proliferation rate of K cells.

[0099] In this embodiment of the application, in order to achieve precise grouping and control of P culture dishes, it is necessary to perform similar clustering based on the real-time state characteristics of cells, so that culture dishes with similar states are grouped into the same set, providing a basis for subsequent differentiated environmental regulation.

[0100] First, it is necessary to define the multiple proliferation phases of cell culture and their corresponding standard cell proliferation rates. These proliferation phases include at least the lag phase, the exponential phase, and the stationary phase.

[0101] Specifically, the lag phase refers to the adaptation period immediately after inoculation, during which cell proliferation is slow. The standard cell proliferation rate is set at 0.02 cells / hour. For example, 0-12 hours after inoculation, the number of cells increases from 50 to 55. The exponential phase refers to the rapid division phase, during which cells proliferate actively. The standard cell proliferation rate is set at 0.08 cells / hour. For example, 24-72 hours after inoculation, the number of cells increases from 100 to 300.

[0102] Furthermore, the stationary phase refers to a period where cell density approaches saturation and proliferation slows down. The standard cell proliferation rate is set at 0.01 cells / hour; for example, 96-120 hours after inoculation, the cell number increases from 800 to 820. These standard cell proliferation rates are derived from historical culture data of mesenchymal stem cells to reflect the ideal cell proliferation state at each stage, providing a benchmark for subsequent weighting and cluster analysis.

[0103] Furthermore, similarity fusion weights are calculated and combined based on the standard cell proliferation rate at the current proliferation stage and the sum of the standard cell proliferation rates at all stages.

[0104] The method provided in this application embodiment includes the step of "configuring similarity fusion weights according to the current standard cell proliferation rate and the plurality of standard cell proliferation rates":

[0105] The ratio of the current standard cell proliferation rate to the sum of the proliferation rates of the multiple standard cells is set as the rate weighting adjustment coefficient;

[0106] The initial rate weight is corrected using the rate weight adjustment coefficient to obtain the adapted rate weight, wherein the initial rate weight is 0.5 and the adapted rate weight is greater than or equal to 0.3 and less than or equal to 0.7.

[0107] The adaptation rate weight is obtained by subtracting the adaptation form weight from 1, and the similarity fusion weight is obtained by combining the adaptation rate weight and the adaptation form weight.

[0108] In this embodiment, the configuration of similarity fusion weights is the core of achieving accurate similarity clustering. That is, based on the characteristics of the cell’s current proliferation stage, the weight ratio of proliferation rate and morphological features in the clustering analysis is dynamically allocated so that the clustering results are more in line with the culture regulation needs of different stages.

[0109] Specifically, the rate weighting adjustment coefficient is first calculated based on the obtained multiple proliferation stages and the corresponding multiple standard cell proliferation rates.

[0110] The rate weighting adjustment coefficient is the ratio of the current standard cell proliferation rate at the current proliferation stage to the sum of the standard cell proliferation rates at all proliferation stages. Its magnitude directly reflects the relative importance of the proliferation characteristics at the current stage in the overall culture process. The specific calculation formula can be expressed as "rate weighting adjustment coefficient = current standard cell proliferation rate / Σ standard cell proliferation rate at each stage".

[0111] Furthermore, after obtaining the rate weight adjustment coefficient, it is necessary to use it to correct the initial rate weight (fixed at 0.5) to obtain the adaptive rate weight. The specific calculation formula can be expressed as "adaptive rate weight = initial rate weight × rate weight adjustment coefficient".

[0112] In order to ensure the balance between cell proliferation rate and cell morphology characteristics in cluster analysis, the fitting rate weight is limited to the range of 0.3 to 0.7. If the correction result exceeds this range, the boundary value is automatically taken (0.3 when it is below 0.3 and 0.7 when it is above 0.7).

[0113] Based on this, the adaptation rate weight is obtained by subtracting the adaptation morphology weight from 1, so as to balance the influence ratio of proliferation rate and morphological features in similar clusters. The specific calculation formula can be expressed as "adaptation morphology weight = 1 - adaptation rate weight".

[0114] Furthermore, the obtained adaptation rate weights and adaptation morphology weights are combined to form similarity fusion weights, which together determine the influence of cell proliferation rate and cell morphology characteristics in clustering.

[0115] For example, assume the standard cell proliferation rates during the lag phase, exponential phase, and stationary phase are 0.02 cells / hour, 0.08 cells / hour, and 0.01 cells / hour, respectively, with a total rate of 0.11 cells / hour. When the current phase is the exponential phase, the rate weight adjustment coefficient is 0.08 / 0.11≈0.73, the adaptation rate weight is 0.5×0.73≈0.365 (in the range of 0.3-0.7), and the adaptation morphology weight is 1-0.365=0.635. At this time, the similarity fusion weight is (0.365, 0.635).

[0116] Furthermore, when the current stage is the lag period, the rate weight adjustment coefficient is 0.02 / 0.11≈0.18, the corrected adaptation rate weight is 0.5×0.18=0.09 (below 0.3, take 0.3), the adaptation morphology weight is 1-0.3=0.7, and the similarity fusion weight is (0.3, 0.7).

[0117] By dynamically configuring similarity fusion weights, the cell proliferation rate has a relatively higher weight in the clustering process during the exponential phase when cell proliferation is active, with a greater emphasis on grouping based on cell proliferation status. In contrast, during the lag and stationary phases when cell proliferation is slow, the weight of cell morphology characteristics is increased, with a greater focus on the consistency of cell morphology. This ensures that the clustering results can accurately match the culture characteristics of different stages, providing a reasonable grouping basis for subsequent differentiated environmental control of the incubator.

[0118] Based on this, and using the obtained similarity fusion weights, the P culture dishes are clustered for similarity by combining the P real-time cell proliferation rates and P real-time cell morphology features.

[0119] Specifically, the real-time cell proliferation rate and real-time cell morphology characteristics of each culture dish are used as multi-dimensional feature vectors. The similarity between each culture dish is calculated by using the cosine similarity calculation method and combining similarity fusion weights.

[0120] The similarity between each culture dish is calculated by first assigning weight values ​​determined by similarity fusion weights to the real-time cell proliferation rate and real-time cell morphology features, then combining the weighted features into new feature vectors, and finally calculating the cosine similarity between these new feature vectors.

[0121] For example, suppose that the real-time cell proliferation rate of a culture dish A is 0.08 cells / hour, and the real-time cell morphology characteristics are as follows: average cell area of ​​70 μm², average cell roundness of 0.7, and cell distribution uniformity of 0.3; and the real-time cell proliferation rate of a culture dish B is 0.075 cells / hour, and the real-time cell morphology characteristics are as follows: average cell area of ​​72 μm², average cell roundness of 0.68, and cell distribution uniformity of 0.32.

[0122] Based on this, if the adaptation rate weight is 0.4 and the adaptation morphology weight is 0.6 in the current similarity fusion weight, then the real-time cell proliferation rate of culture dishes A and B are assigned a weight of 0.4 respectively, and the three morphological features of mean cell area, mean cell roundness, and uniformity of cell distribution are assigned a weight of 0.6 respectively (the weight of 0.6 can be further evenly distributed among the three features, each to 0.2).

[0123] After weighting, the feature vector of petri dish A is [0.08×0.4, 70×0.2, 0.7×0.2, 0.3×0.2], and the feature vector of petri dish B is [0.075×0.4, 72×0.2, 0.68×0.2, 0.32×0.2]. The similarity between the two petri dishes can be obtained by calculating the cosine similarity between these two vectors.

[0124] Meanwhile, existing clustering algorithms (such as K-means) are used to group similar petri dishes into one class, and finally output K sets of similar petri dishes, with each set containing the same number of petri dishes, P / K.

[0125] For example, if P=12 and K=2, then each set of similar culture dishes should contain 6 culture dishes. After calculating the cosine similarity, the K-means algorithm is used to group the 6 culture dishes with the highest similarity among the 12 culture dishes into the first set, and the remaining 6 into the second set. The culture dishes in the two sets have high similarity in real-time cell proliferation rate and cell morphology characteristics, which meets the requirement that the number of culture dishes in each set is the same.

[0126] Furthermore, after completing the similarity clustering, the mean of the real-time cell proliferation rates of P / K cells in the K similar culture dish sets is calculated. That is, the mean of the K cell proliferation rates is obtained by the formula "mean cell proliferation rate = (Σ real-time cell proliferation rate of culture dishes in the set) / (number of culture dishes in the set P / K)".

[0127] For example, if a set of similar culture dishes contains 4 culture dishes (P / K=4), and their real-time cell proliferation rates are 0.07 cells / hour, 0.08 cells / hour, 0.075 cells / hour, and 0.085 cells / hour, respectively, then the average cell proliferation rate of the set is (0.07+0.08+0.075+0.085) / 4=0.0775 cells / hour.

[0128] Ultimately, the average cell proliferation rate obtained can reflect the overall cell proliferation level within each similar set, providing a key quantitative reference for subsequent optimization and adjustment of environmental parameters based on the overall state of the incubator.

[0129] S140: Map the K similar culture dish sets onto the K incubators, and optimize the culture environment parameters of the K incubators based on the average cell proliferation rate of the K cells to obtain K suitable environment parameters, and manage the environment of the K incubators within a future time window.

[0130] In this embodiment of the application, in order to achieve precise and differentiated control of the stem cell culture environment, it is necessary to first complete the mapping between a set of similar culture dishes and the incubator, and then adjust the adaptation environment parameters according to the cell proliferation characteristics in order to dynamically adapt to the needs of different proliferation stages of stem cells and construct a culture environment that conforms to the cell growth pattern.

[0131] Specifically, firstly, based on the obtained set of K similar culture dishes, the positions of P culture dishes in the K incubators are adjusted, and culture dishes of the same set are placed in the same incubator. The incubators are mapped and marked according to the average cell proliferation rate of K, so that the set of similar culture dishes is associated with the corresponding incubator.

[0132] Meanwhile, for the average proliferation rate of K cells, combined with the corresponding proliferation stage, the difference between the average proliferation rate and the current standard cell proliferation rate is calculated to obtain the proliferation rate difference of K cells (the difference can be positive or negative, and a negative number indicates that the cells in the corresponding set are proliferating relatively too fast).

[0133] Furthermore, based on the obtained K cell proliferation rate differences, and referring to the preset environmental parameter control logic, the culture environment parameters of the K incubators are optimized and adjusted respectively, thereby obtaining K suitable environmental parameters.

[0134] Finally, based on these adaptive environmental parameters, environmental control will be implemented for K incubators within a future time window to dynamically adapt the culture environment to the stem cell proliferation requirements, ensuring the accuracy and effectiveness of stem cell culture density optimization control, and providing environmental support for subsequent stable and efficient stem cell culture.

[0135] Step S140 in the method provided in this application embodiment includes:

[0136] Based on the K sets of similar culture dishes, the positions of P culture dishes in the K incubators are adjusted so that culture dishes belonging to the same set of similar culture dishes are placed in the same incubator, and the K incubators are mapped and marked according to the average cell proliferation rate of the K cells.

[0137] Using the current standard cell proliferation rate at the current proliferation stage as a benchmark, the deviation of the mean proliferation rate of the K cells is calculated to obtain the difference in proliferation rate of the K cells.

[0138] Based on the differences in the proliferation rates of the K cells, the culture environment parameters of the K incubators were optimized and adjusted to obtain K suitable environment parameters.

[0139] In this embodiment of the application, after completing the clustering of K similar culture dish sets, in order to achieve precise grouping and control of the stem cell culture environment, it is necessary to establish a physical association between the culture dish sets and the incubator, so as to ensure that the culture dishes of the same type are in a consistent environmental space that is suitable for their proliferation characteristics, thus laying a solid spatial foundation for subsequent adjustment of differentiated environmental parameters.

[0140] First, by traversing K incubators and P petri dishes in the corresponding space, and based on the similarity set labels to which the petri dishes belong, the petri dishes of the same similar set are placed together in the same incubator by means of a robotic arm.

[0141] For example, if K=3 and P=12, clustering yields three sets of similar petri dishes (named set A, set B, and set C, respectively), each set containing four petri dishes. By traversing incubator 1, incubator 2, incubator 3, and the 12 petri dishes scattered in each incubator, it is identified that petri dishes 1-4 belong to set A, petri dishes 5-8 belong to set B, and petri dishes 9-12 belong to set C.

[0142] Based on this, using a robotic arm to grasp, culture dishes 1-4 were moved from their original distribution locations to incubator 1 and arranged neatly, culture dishes 5-8 were moved to incubator 2, and culture dishes 9-12 were moved to incubator 3. This allowed each incubator to store culture dishes of the same similar set, establishing a physical correspondence between "set-incubator" and creating grouping conditions for subsequent environmental control based on incubators.

[0143] Furthermore, after adjusting the position of the culture dish, the incubator was mapped and marked according to the average proliferation rate of K cells.

[0144] Among them, the mapping and labeling rules combine the characteristics of the proliferation stage. For example, the set of cells with a higher average cell proliferation rate than the current stage standard cell proliferation rate is labeled as "high proliferation group incubator", and the set with an average rate close to the standard is labeled as "conventional proliferation group incubator". This allows operators to intuitively identify the overall level of cell proliferation in each incubator, providing a physical and labeling basis for subsequent environmental control.

[0145] Furthermore, when calculating the difference in cell proliferation rates, the specific proliferation stage of the cells is determined based on characteristics such as cell density and growth status during cell culture, and the corresponding standard cell proliferation rate is retrieved at the same time.

[0146] Furthermore, for each set of similar culture dishes, the cell proliferation rate difference is calculated using the formula "cell proliferation rate difference = average cell proliferation rate of the set - standard cell proliferation rate of the current proliferation stage".

[0147] The cell proliferation rate difference result can be positive, negative, or zero. If the cell proliferation rate difference is positive, it means that the cell proliferation rate in the similar culture dish set is higher than the current standard cell proliferation rate, and may be in a more active growth state. If the difference is negative, it means that the cell proliferation rate is lower than the standard, indicating that cell growth may be slower. If the difference is zero, it means that the cell proliferation level is in line with the standard.

[0148] Similarly, the cell proliferation rate difference corresponding to group K is calculated one by one using the same method to quantify the degree of deviation of each similar culture dish set from the ideal cell proliferation state, providing directional guidance for adjusting the culture environment parameters.

[0149] Based on this, the incubator's environmental parameters were optimized and adjusted to ensure that the environment in each incubator was adapted to the proliferation requirements of cells in similar culture dishes, so that the cell proliferation state would approach the ideal standard.

[0150] Specifically, based on the sign of the difference in the proliferation rates of K cells, the environmental parameters of the incubator, such as temperature, humidity, carbon dioxide concentration, and oxygen concentration, are adjusted accordingly.

[0151] The adjustment range needs to be flexibly set according to the size of the difference in cell proliferation rate. The larger the absolute value of the difference in cell proliferation rate, the greater the range of environmental parameter adjustment should be, so as to more effectively correct cell proliferation deviation. At the same time, it is necessary to ensure that the adjusted environmental parameters are within a reasonable range for cell survival and normal metabolism, so as to avoid damage to cells due to excessive adjustment.

[0152] If the difference in cell proliferation rate is positive, it indicates that the cell proliferation rate is too fast. In this case, the temperature can be appropriately reduced, the carbon dioxide concentration can be lowered, etc., to slow down the metabolic activity and division rate of cells, and to avoid excessive cell proliferation leading to excessive density, insufficient nutrition, or abnormal morphology.

[0153] Conversely, if the difference in cell proliferation rate is negative, it indicates that the cell proliferation rate is too slow. In this case, the temperature can be increased, the carbon dioxide concentration can be increased, etc., to enhance the metabolic activity of the cells, promote cell division and proliferation, and bring the cell proliferation rate closer to the standard level of the current stage.

[0154] For example, if the current proliferation phase is in the exponential phase, the standard cell proliferation rate is 0.08 cells / hour. If the average cell proliferation rate of a set of similar culture dishes in a certain incubator is 0.09 cells / hour, and the difference in cell proliferation rate is +0.01 (positive value), it indicates that the cells are proliferating too quickly. The incubator temperature can be lowered from 37°C to 36.5°C, and the carbon dioxide concentration can be adjusted from 5% to 4% to slow down cell metabolism and division, and maintain the stability and optimization of stem cell culture density.

[0155] In addition, the average cell proliferation rate of the same set of culture dishes in the other incubator was 0.07 cells / hour, and the difference in cell proliferation rate was -0.01 (negative value), indicating that cell proliferation was too slow. The temperature could be increased to 37.5℃ and the carbon dioxide concentration to 6% to adjust the cell proliferation state to be closer to the standard cell proliferation rate at the current stage through differential environmental adjustment. After adjustment, the suitable environmental parameters for the two incubators were formed.

[0156] Finally, based on the obtained adaptive environmental parameters, targeted environmental settings were made for the K incubators within the future time window to ensure that parameters such as temperature, humidity, carbon dioxide concentration, and oxygen concentration in each incubator accurately matched the cell proliferation requirements of the corresponding set of similar culture dishes.

[0157] This step precisely matches cell proliferation status with culture environment parameters, and then manages the environment of K incubators within a future time window to continuously maintain the appropriate environment in each incubator. This ensures that stem cells are in growth conditions conducive to density optimization throughout the entire culture cycle, thereby achieving precise control over stem cell culture density.

[0158] The embodiments of this application, through the specific implementation methods described above, achieve the following technical effects:

[0159] This application proposes a method for optimizing and controlling the density of stem cell culture. First, K incubators with individually controllable environmental parameters and P culture dishes are configured. The P culture dishes are randomly and equally distributed among the K incubators for mesenchymal stem cell culture. Within a historical time window, cell imaging sequences are periodically captured using a fluorescence microscope. Using a trained convolutional neural network model, P real-time cell proliferation rates and P real-time cell morphological characteristics, including average cell area, average roundness, and distribution uniformity, are analyzed. Based on the cell proliferation rate and morphological characteristics, and combined with the standard cell proliferation rate at the cell's current proliferation stage, similarity fusion weights are configured. A clustering algorithm is used to divide the P culture dishes into K similar culture dish sets, and the average cell proliferation rate of each set is calculated. Culture dishes from the same set are placed in the same incubator and labeled. The difference in cell proliferation rate between sets is calculated based on the current standard cell proliferation rate. Based on this, the environmental parameters of the incubator, such as temperature, humidity, carbon dioxide concentration, and oxygen concentration, are optimized and adjusted to obtain K suitable environmental parameters. Environmental control of the incubator is then implemented within future time windows.

[0160] The method provided in this application, through the technical solution of "equipment configuration - feature monitoring - similar clustering - environment optimization", solves the problems of inconsistent cell states and low density control precision caused by uniform environmental regulation in traditional stem cell culture. It realizes differentiated culture environment regulation based on real-time cell state, improves the accuracy and stability of stem cell culture density control, and provides reliable technical support for efficient stem cell culture and quality assurance.

[0161] Example 2, as shown in the appendix Figure 2 As shown, based on the inventive concept of the stem cell culture density optimization control method provided in Embodiment 1, this application also provides a stem cell culture density optimization control system, specifically including:

[0162] The equipment configuration module 01 is used to configure K incubators and P culture dishes, wherein the culture environment of each incubator can be controlled independently, K is an integer greater than or equal to 2, P is an integer greater than or equal to 12, and P is an integer multiple of K.

[0163] The monitoring and analysis module 02 is used to randomly and equally distribute the P culture dishes to the K incubators for stem cell culture during the mesenchymal stem cell culture process, monitor and acquire the P cell imaging sequences of the P culture dishes within a historical time window, and analyze the P real-time cell proliferation rate and P real-time cell morphology characteristics.

[0164] Clustering mean module 03 is used to perform similar clustering on the P culture dishes based on the P real-time cell proliferation rates and P real-time cell morphology characteristics, to determine a set of K similar culture dishes and the mean of K cell proliferation rates;

[0165] The environment optimization and control module 04 is used to map and place the K similar culture dish sets into the K incubators, and optimize and adjust the culture environment parameters of the K incubators based on the average cell proliferation rate of the K cells to obtain K suitable environment parameters, and control the environment of the K incubators within a future time window.

[0166] In one embodiment, the monitoring and analysis module 02 is further configured to:

[0167] Within the historical time window, cell images of the P culture dishes are periodically captured using a fluorescence microscope at preset time intervals to obtain P cell imaging sequences.

[0168] Based on the P cell imaging sequences, cell proliferation rate analysis is performed, and P real-time cell proliferation rates are output.

[0169] P terminal cell images are extracted from the P cell imaging sequences respectively, and cell morphology features are identified to output P real-time cell morphology features, including mean cell area, mean cell roundness, and cell distribution uniformity.

[0170] In one embodiment, the cluster mean module 03 is further configured to:

[0171] Multiple proliferation phases during cell culture and multiple standard cell proliferation rates for the multiple proliferation phases are obtained, wherein the proliferation phase includes at least a lag phase, an exponential phase, and a stationary phase;

[0172] Similarity fusion weights are configured based on the current standard cell proliferation rate at the current proliferation stage and the proliferation rates of the multiple standard cells;

[0173] Based on the similarity fusion weight, the P culture dishes are clustered similarly according to the P real-time cell proliferation rates and P real-time cell morphology features, and K similar culture dish sets are output, wherein the number of culture dishes in each similar culture dish set is the same.

[0174] The average proliferation rates of P / K real-time cells in the K similar culture dish sets are calculated to obtain the average proliferation rate of K cells.

[0175] In one embodiment, the environmental optimization and control module 04 is further configured to:

[0176] Based on the K sets of similar culture dishes, the positions of P culture dishes in the K incubators are adjusted so that culture dishes belonging to the same set of similar culture dishes are placed in the same incubator, and the K incubators are mapped and marked according to the average cell proliferation rate of the K cells.

[0177] Using the current standard cell proliferation rate at the current proliferation stage as a benchmark, the deviation of the mean proliferation rate of the K cells is calculated to obtain the difference in proliferation rate of the K cells.

[0178] Based on the differences in the proliferation rates of the K cells, the culture environment parameters of the K incubators were optimized and adjusted to obtain K suitable environment parameters.

[0179] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0180] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0181] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. A method for optimizing and controlling stem cell culture density, characterized in that the method... include: Configure K incubators and P culture dishes, wherein the culture environment of each incubator can be controlled independently, K is an integer greater than or equal to 2, P is an integer greater than or equal to 12, and P is an integer multiple of K; During the mesenchymal stem cell culture process, the P culture dishes are randomly and equally distributed to the K incubators for stem cell culture. The imaging sequences of the P cells in the P culture dishes within the historical time window are monitored and obtained, and the real-time cell proliferation rate and morphological characteristics of the P cells are analyzed. Based on the P real-time cell proliferation rates and P real-time cell morphology characteristics, similarity clustering is performed on the P culture dishes to determine K similar culture dish sets and K average cell proliferation rates; The K similar culture dish sets are mapped and placed in the K incubators, and the culture environment parameters of the K incubators are optimized and adjusted based on the average cell proliferation rate of the K cells to obtain K suitable environment parameters. The environment of the K incubators is then managed within a future time window. Specifically, based on the P real-time cell proliferation rates and P real-time cell morphology characteristics, the P culture dishes are clustered to determine K similar culture dish sets, and K mean cell proliferation rates are also defined, including: Multiple proliferation phases during cell culture and multiple standard cell proliferation rates for the multiple proliferation phases are obtained, wherein the proliferation phase includes at least a lag phase, an exponential phase, and a stationary phase; Based on the current standard cell proliferation rate at the current proliferation stage and the plurality of standard cell proliferation rates, similarity fusion weights are configured, including: The ratio of the current standard cell proliferation rate to the sum of the proliferation rates of the multiple standard cells is set as the rate weighting adjustment coefficient; The initial rate weight is corrected using the rate weight adjustment coefficient to obtain the adapted rate weight, wherein the initial rate weight is 0.5 and the adapted rate weight is greater than or equal to 0.3 and less than or equal to 0.

7. The adaptation rate weight is obtained by subtracting the adaptation rate weight from 1, and the similarity fusion weight is obtained by combining the adaptation rate weight and the adaptation form weight. Based on the similarity fusion weight, the P culture dishes are clustered similarly according to the P real-time cell proliferation rates and P real-time cell morphology features, and K similar culture dish sets are output, wherein the number of culture dishes in each similar culture dish set is the same. The average proliferation rates of P / K real-time cells in the K similar culture dish sets are calculated to obtain the average proliferation rate of K cells.

2. The method for optimizing and controlling stem cell culture density according to claim 1, characterized in that, The incubator can control the culture environment independently, wherein the controllable culture environment parameters include at least temperature, humidity, carbon dioxide concentration and oxygen concentration.

3. The method for optimizing and controlling stem cell culture density according to claim 1, characterized in that, Monitoring and acquiring P cell imaging sequences from P culture dishes within a historical time window, analyzing P real-time cell proliferation rates and P real-time cell morphological characteristics, including: Within the historical time window, cell images of the P culture dishes are periodically captured using a fluorescence microscope at preset time intervals to obtain P cell imaging sequences. Based on the P cell imaging sequences, cell proliferation rate analysis is performed, and P real-time cell proliferation rates are output. P terminal cell images are extracted from the P cell imaging sequences respectively, and cell morphology features are identified to output P real-time cell morphology features, including mean cell area, mean cell roundness, and cell distribution uniformity.

4. The method for optimizing and controlling stem cell culture density according to claim 3, characterized in that, The method further includes: Based on the historical culture records of mesenchymal stem cells, a sample cell imaging set, a sample cell number label set, and a sample cell morphology characteristic label set were collected. Using the sample cell imaging set and sample cell number label set, a convolutional neural network is trained until convergence to obtain an automatic cell counter for cell number identification, and the cell proliferation rate is calculated based on the P cell number identification results. Using the sample cell imaging set and the sample cell morphology feature label set, a convolutional neural network is trained until convergence to obtain a cell morphology recognizer for cell morphology feature recognition.

5. The method for optimizing and controlling stem cell culture density according to claim 1, characterized in that, Mapping the K sets of similar culture dishes onto the K incubators includes: Based on the K sets of similar culture dishes, the positions of P culture dishes in the K incubators are adjusted so that culture dishes belonging to the same set of similar culture dishes are placed in the same incubator, and the K incubators are mapped and marked according to the average cell proliferation rate of the K cells.

6. The method for optimizing and controlling stem cell culture density according to claim 1, characterized in that, Based on the average proliferation rate of the K cells, the culture environment parameters of the K incubators were optimized and adjusted to obtain K suitable environment parameters, including: Using the current standard cell proliferation rate at the current proliferation stage as a benchmark, the deviation of the mean proliferation rate of the K cells is calculated to obtain the difference in proliferation rate of the K cells. Based on the differences in the proliferation rates of the K cells, the culture environment parameters of the K incubators were optimized and adjusted to obtain K suitable environment parameters.

7. A stem cell culture density optimization control system, characterized in that, The system is used to execute the stem cell culture density optimization control method according to any one of claims 1-6, the system comprising: The equipment configuration module is used to configure K incubators and P culture dishes, wherein the culture environment of each incubator can be controlled independently, K is an integer greater than or equal to 2, P is an integer greater than or equal to 12, and P is an integer multiple of K. The monitoring and analysis module is used to randomly and equally distribute the P culture dishes to the K incubators for stem cell culture during the mesenchymal stem cell culture process, monitor and acquire the P cell imaging sequences of the P culture dishes within a historical time window, and analyze the P real-time cell proliferation rate and P real-time cell morphology characteristics. The cluster mean module is used to perform similar clustering on the P culture dishes based on the P real-time cell proliferation rates and P real-time cell morphology characteristics, to determine a set of K similar culture dishes and the mean of K cell proliferation rates; The environment optimization and control module is used to map and place the K similar culture dish sets into the K incubators, and optimize and adjust the culture environment parameters of the K incubators based on the average cell proliferation rate of the K cells to obtain K suitable environment parameters, and control the environment of the K incubators within a future time window; Specifically, based on the P real-time cell proliferation rates and P real-time cell morphology characteristics, the P culture dishes are clustered to determine K similar culture dish sets, and K mean cell proliferation rates are also defined, including: Multiple proliferation phases during cell culture and multiple standard cell proliferation rates for the multiple proliferation phases are obtained, wherein the proliferation phase includes at least a lag phase, an exponential phase, and a stationary phase; Based on the current standard cell proliferation rate at the current proliferation stage and the plurality of standard cell proliferation rates, similarity fusion weights are configured, including: The ratio of the current standard cell proliferation rate to the sum of the proliferation rates of the multiple standard cells is set as the rate weighting adjustment coefficient; The initial rate weight is corrected using the rate weight adjustment coefficient to obtain the adapted rate weight, wherein the initial rate weight is 0.5 and the adapted rate weight is greater than or equal to 0.3 and less than or equal to 0.

7. The adaptation rate weight is obtained by subtracting the adaptation rate weight from 1, and the similarity fusion weight is obtained by combining the adaptation rate weight and the adaptation form weight. Based on the similarity fusion weight, the P culture dishes are clustered similarly according to the P real-time cell proliferation rates and P real-time cell morphology features, and K similar culture dish sets are output, wherein the number of culture dishes in each similar culture dish set is the same. The average proliferation rates of P / K real-time cells in the K similar culture dish sets are calculated to obtain the average proliferation rate of K cells.

Images