Cell culture equipment for cell biology experiments based on artificial intelligence

Abstract

The present application relates to the technical field of bioreactor, in particular to a cell culture equipment for cell biology experiment based on artificial intelligence, which comprises a culture shell, a plurality of hollow fiber bundles are arranged in the culture shell, and a rectifier assembly is arranged at both ends of the culture shell; the hollow fiber bundles are in a spiral configuration, and the two ends of the hollow fiber bundles are centrally symmetrically distributed along the axis of the culture shell; two groups of pump liquid devices are arranged on the outside of the culture shell, and each group of pump liquid devices is communicated with part of the hollow fiber bundles through two rectifier assemblies; under the drive of the two groups of pump liquid devices, the culture medium is divided into two flow directions from both ends of the culture shell and enters the corresponding communicated hollow fiber bundles, and the culture medium fluid of the two flow directions flows in a staggered manner along the direction of approaching each other in the culture shell; the present application guides the culture medium fluid in the hollow fiber bundle, adopts reverse staggered perfusion nutrition delivery at both ends, and solves the phenomenon of axial decrease of nutrition in the traditional unidirectional perfusion.

Description

Technical Field

[0001] This invention relates to the field of bioreactor technology, and more specifically to a cell culture device for cell biology experiments based on artificial intelligence. Background Technology

[0002] Cell culture is a method that simulates in vivo environments such as temperature, pH, and nutritional conditions to enable cells to survive, grow, multiply, and maintain their main structures and functions. This technology is one of the most important fundamental tools in cell and molecular biology research and is widely used in fields such as metabolic research, aging research, drug screening, toxicity evaluation, and large-scale production of vaccines and therapeutic proteins.

[0003] In existing cell culture technologies, the hollow fiber cell system employs a cylindrical shell encapsulating multiple parallel hollow fiber membranes, forming a component structure similar to an optical fiber bundle. Utilizing the hollow fiber membrane as a selective barrier, cells are seeded and cultured within the cell growth space (outer cavity) formed by the outer wall of the fiber and the inner wall of the shell, while the cell culture medium continuously flows through the inner cavity (inner cavity). This allows for bidirectional transmembrane exchange of nutrients, gases, and metabolites through the membrane wall of the hollow fiber. This mimics the material transport pattern of the capillary network in vivo, supporting long-term, high-density cell culture and achieving stable microenvironment control. Compared to traditional static culture, it offers higher product enrichment efficiency and batch consistency.

[0004] In hollow fiber cell culture systems, because the culture medium flows unidirectionally through the fiber lumen, nutrients are gradually consumed along the flow path after being taken up by cells on the outer wall of the tube. This results in a lower nutrient concentration at the fiber outlet than at the inlet, leading to a spatial gradient effect that reduces the homogeneity of the culture environment. Although existing techniques have proposed periodically alternating the pumping direction to compensate for the reverse distribution of nutrients, the fluid shear stress generated by the culture medium flow and the time-limited nature of nutrient transmembrane exchange make it difficult to completely avoid the nutrient concentration gradient along the fiber axis and the resulting spatial heterogeneity in cell metabolism and proliferation. Therefore, it is necessary to propose an artificial intelligence-based cell culture device for cell biology experiments to solve the above-mentioned technical problems. Summary of the Invention

[0005] To address the aforementioned problems, this invention provides a cell culture device for cell biology experiments based on artificial intelligence. By guiding the culture medium fluid within the hollow fiber bundle, it employs a reverse-interleaved perfusion nutrient delivery system at both ends, specifically solving the problem of nutrient reduction along the fiber axis in traditional unidirectional perfusion.

[0006] To achieve the above objectives, the technical solution of the present invention is as follows: A cell culture device for cell biology experiments based on artificial intelligence, comprising a cylindrical culture shell, with an inlet tube and a collection tube respectively connected to the top and bottom of the culture shell, and a plurality of hollow fiber bundles disposed inside the culture shell, characterized in that rectifier components are provided at both ends inside the culture shell, and the rectifier components are used to distribute fluid to each hollow fiber bundle and integrate the inlet and outlet flow paths of the fluid.

[0007] The hollow fiber bundle has a helical configuration and its two ends are centrally symmetrically distributed along the axis of the culture shell. The two ends of the hollow fiber bundle are respectively connected to the corresponding rectifier components.

[0008] Two sets of pump components are provided on the outside of the culture shell. Each pump component is connected to a feeding tube for replenishing fresh culture medium and a waste discharge tube for discharging cell metabolic waste. Each set of pump components is connected to a portion of the hollow fiber bundle through two rectifier components.

[0009] Driven by two sets of pump components, the culture medium flows into the corresponding connected hollow fiber bundles from both ends of the culture shell in two directions. The two flow directions of culture medium fluids flow alternately in the culture shell in a direction that is close to each other.

[0010] The technical principle of the above scheme is as follows: This scheme adopts a cylindrical culture shell, with rectifier components at both ends inside the culture shell for distributing and integrating the inflow and outflow of fluids; the culture shell contains several hollow fiber bundles with a spiral configuration, with both ends of the hollow fiber bundles symmetrically distributed along the center of the shell axis and connected to the rectifier components at both ends respectively; two sets of pumps with opposite pumping directions are set on the outside of the shell, which are connected to a portion of the hollow fiber bundles through the rectifier components. Driven by the two sets of pumps, the culture medium flows into the corresponding hollow fiber bundles from both ends of the culture shell in two directions. The two fluids flow alternately in directions that are close to each other inside the shell, forming countercurrent and secondary flow disturbances, thereby improving the spatial uniformity of nutrient delivery.

[0011] The above approach has the following beneficial effects:

[0012] 1. This scheme guides the culture medium fluid within the hollow fiber bundle and uses reverse staggered perfusion at both ends to address the phenomenon of decreasing nutrition along the fiber axis in traditional unidirectional perfusion, thereby ensuring a balanced nutrient supply to cells in different regions within the shell and improving culture uniformity.

[0013] 2. The opposing and secondary flow disturbances in the culture medium fluid generated by the countercurrent flow in this scheme can suppress cell accumulation or uneven scouring caused by the continuous unidirectional action of the fluid tangential force, and reduce the adverse effects of shear stress on cell spatial distribution.

[0014] 3. This scheme improves the efficiency of material exchange through the synergistic effect of the spiral-shaped hollow fiber bundles and the reverse perfusion fluid movement, which is conducive to achieving high-density, long-term stable cell culture.

[0015] Furthermore, each rectifier assembly includes a hollow, disc-shaped rectifier housing, and a partition is fixedly connected inside the rectifier housing. The partition divides the internal space of the rectifier housing into an upper cavity and a lower cavity that are symmetrically arranged.

[0016] Beneficial effects: The baffle divides the rectifier shell into an upper and lower symmetrical chamber, realizing physical isolation and directional distribution of fluid inlet and outlet flow paths; this design simplifies the flow channel structure, reduces the probability of fluid crosstalk, ensures the independent and controllable flow of culture medium in each hollow fiber bundle, and improves the reliability and manufacturability of the perfusion system.

[0017] Furthermore, the two ends of the hollow fiber bundle are respectively connected to the upper cavity at one end of the culture shell and the lower cavity at the other end of the culture shell.

[0018] Beneficial effects: The hollow fiber bundle connects the upper cavity at one end of the culture shell and the lower cavity at the other end, forming a cross-connected path; this allows the two sets of pump components to independently drive two perfusion cycles in opposite directions, reducing the probability of flow path overlap and interference, and providing basic topological support for subsequent realization of counter-flow and temperature gradient control.

[0019] Furthermore, the output ends of the two sets of pump components are connected to different lower chambers, and the receiving ends of the two sets of pump components are connected to the corresponding upper chambers.

[0020] Beneficial effects: The output ends of the two sets of pump components are connected to different lower chambers, and the receiving ends are connected to different upper chambers respectively, which clarifies the correspondence between the pump components and the chambers and ensures that each set of pump components independently drives a directional flow path from the lower chamber through the hollow fiber bundle to the upper chamber; it ensures that the culture medium fluid flows in opposite directions at the same time and that the two fluid paths do not mix, thus improving the accuracy of system control.

[0021] Furthermore, the culture shell is also provided with several support disks, each of which has several support holes corresponding to hollow fiber bundles. The hollow fiber bundles are all inserted and attached to the corresponding support holes.

[0022] Beneficial effects: The design of the support plate, combined with the support holes on its surface, forms multi-point radial constraints on the fiber bundle, which can effectively suppress fiber displacement and deformation caused by fluid pulsation or mechanical vibration, reduce the risk of fiber bundle breakage due to bending friction, and provide a mechanical contact interface for subsequent strain detection.

[0023] Furthermore, strain sensors are installed at each support hole to collect strain signals generated by the hollow fiber bundles due to the flow of culture medium and cell growth in real time.

[0024] Beneficial effects: The strain sensor can collect strain signals generated by the hollow fiber bundle in real time due to the flow of culture medium and cell growth; it can convert the slight expansion of the hollow fiber bundle into a measurable electrical signal, providing a non-invasive sensing method for monitoring local pressure changes, blockage status and cell metabolic activities within the fiber bundle, and laying the data foundation for intelligent feedback control.

[0025] Furthermore, a cell culture device for cell biology experiments based on artificial intelligence also includes a processing system, in which several strain sensors and pump components are connected to the processing system via signals.

[0026] All feed pipes are connected to feed solenoid valves, and all waste discharge pipes are connected to waste discharge solenoid valves. Both feed solenoid valves and waste discharge solenoid valves are connected to the processing system signal.

[0027] Beneficial effects: The processing system can receive strain sensor signals and control the actuators to achieve automated monitoring and intervention in the cultivation process, reduce manual operation, reduce the risk of contamination, and has real-time response capability.

[0028] Furthermore, the processing system includes a strain data analysis module, a uniformity assessment module, and a feedback control module. The strain data analysis module has a built-in nutrient exchange prediction model.

[0029] The strain data analysis module is used to receive real-time strain data collected by several strain sensors and to perform time-series analysis and feature extraction on each strain data.

[0030] The nutrient exchange prediction model is pre-trained based on historical strain data and corresponding blockage degree labels, and is used to output the prediction results of the local blockage degree of each hollow fiber bundle.

[0031] The uniformity assessment module is used to assess the spatial uniformity of nutrient delivery and the spatial consistency of cell metabolic activity within the culture shell based on the predicted results of the local blockage degree of each hollow fiber bundle.

[0032] The feedback control module automatically generates control instructions when the uniformity assessment module determines that the culture uniformity is lower than the preset threshold. The control instructions include adjusting the pumping rate, pumping direction and timing of the two sets of pump components or the feed flow rate of the feed pipe, so as to achieve adaptive feedback adjustment of the culture medium perfusion parameters.

[0033] Beneficial effects: This design enables adaptive adjustment of irrigation parameters, effectively mitigating batch-to-batch variability caused by axial nutrient gradient.

[0034] Furthermore, the culture shell is equipped with several temperature acquisition devices, which are used to collect the real-time temperature difference between different areas inside the culture shell. All of these temperature acquisition devices are connected to the processing system signal.

[0035] Beneficial effects: The temperature acquisition device can obtain real-time temperature differences in different areas and transmit them to the processing system; it can detect abnormal temperature gradients in the culture environment, providing spatially resolved input data for subsequent temperature regulation, making up for the shortcomings of a single temperature monitoring point, and improving the precision of temperature control.

[0036] Furthermore, the processing system also includes a temperature control module, which has a built-in temperature field prediction model.

[0037] The temperature field prediction model is constructed by training on historical data of temperature gradient inside the culture shell recorded and collected during cell culture and the corresponding optimal control parameters.

[0038] The temperature control module is used to input real-time temperature difference data into the trained temperature field prediction model. The temperature field prediction model calculates and outputs adjustment instructions or temperature control suggestions for the temperature of the culture medium in the two sets of feeding tubes. By utilizing the temperature difference between the two ends of the culture shell and the opposite direction of the culture medium flowing into the hollow fiber bundle, and combined with the characteristic that the temperature decreases along the flow path of the culture medium in the spiral flow channel, a predictable gradient temperature control field is formed in the culture shell, so as to realize intelligent temperature compensation and regulation of the target area.

[0039] Beneficial effects: By controlling the temperature difference between the two ends of the influent culture medium, and taking into account the characteristic that the temperature decreases along the flow path of the culture medium in the spiral channel, a predictable gradient temperature field is formed in the shell, achieving precise temperature compensation for the target area; this design extends temperature control from uniform isothermal to a designable gradient temperature field, meeting the specific temperature distribution requirements of different cells or experiments.

[0040] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0041] Figure 1 This is a schematic diagram of the overall structure of an embodiment of the cell culture device for cell biology experiments based on artificial intelligence according to the present invention;

[0042] Figure 2 for Figure 1 Axonometric sectional view showing the interior of the culture shell;

[0043] Figure 3 for Figure 2 Axonometric sectional view illustrating the interior of the rectifier assembly;

[0044] Figure 4 This is an isometric view illustrating the connection between the pump component and the rectifier assembly in an embodiment of the cell culture device for cell biology experiments based on artificial intelligence of the present invention;

[0045] Figure 5 This is an isometric view illustrating the connection between the hollow fiber bundle and the rectifier assembly in an embodiment of the cell culture device for cell biology experiments based on artificial intelligence of the present invention;

[0046] Figure 6 for Figure 5 A partial isometric sectional view illustrating the support disc structure;

[0047] Figure 7 This is an isometric sectional view showing the arrangement of temperature acquisition components in an embodiment of the cell culture device for cell biology experiments based on artificial intelligence of the present invention.

[0048] Figure 8 This is an isometric view of the arrangement of isolation culture plates in an embodiment of the cell culture device for cell biology experiments based on artificial intelligence of the present invention.

[0049] The reference numerals in the accompanying drawings of the instruction manual include: 1. Culture shell; 2. Sample inlet tube; 3. Collection tube; 4. Rectifier assembly; 41. Rectifier shell; 42. Separator; 43. Upper chamber; 44. Lower chamber; 5. Hollow fiber bundle; 6. Pump component; 7. Feed tube; 8. Waste discharge tube; 9. Support plate; 10. Support hole; 11. Strain sensor; 12. Temperature acquisition component; 13. Isolation culture plate. Detailed Implementation

[0050] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0052] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0053] The following detailed description illustrates the specific implementation method:

[0054] Example 1:

[0055] This embodiment provides a cell culture device for cell biology experiments based on artificial intelligence, specifically as follows: Figure 1 As shown, it includes a cylindrical culture shell 1, with a sample inlet tube 2 and a collection tube 3 connected to the top and bottom of the culture shell 1, respectively; combined with Figure 2 and Figure 3 As shown, the culture shell 1 has rectifier components 4 at both ends inside. Each rectifier component 4 includes a hollow disc-shaped rectifier shell 41. Each rectifier shell 41 has an integrally formed partition 42, which divides the internal space of the rectifier shell 41 into an upper cavity 43 and a lower cavity 44 that are symmetrically arranged. The culture shell 1 also has several hollow fiber bundles 5 (the hollow fiber bundles 5 are preferably hollow fiber membranes made of natural hydrophilic polymers in existing hollow fiber cell culture systems. In this embodiment, the pore size of the inner surface dense layer of the hollow fiber membrane is 0.1-0.45 μm. The pore size of the inner surface dense layer can be customized according to the target cell type or the molecular weight to be retained. The pore size of the outer support layer is 1-5 μm). This hollow fiber bundle 5 structure ensures bidirectional free transmembrane diffusion of nutrients (such as glucose, amino acids, and growth factors) while effectively blocking the reverse permeation of cells and macromolecular metabolites (such as lactic acid and ammonia), maintaining the stability of the cell growth environment in the outer cavity. The bidirectional permeability coefficient of hollow fiber bundle 5 has been standardized and tested. The water flux is ≥200 L / m²·h·bar and the solute molecular weight cutoff (MWCO) is 50~300 kDa. It can be flexibly selected according to the culture objectives.

[0056] Each hollow fiber bundle 5 is arranged in a spiral configuration within the culture shell 1 (see...). Figure 3The hollow fiber bundles 5 shown in the diagram are centrally symmetrical about the axis of the culture shell 1 at both ends. The two ends of each hollow fiber bundle 5 are connected to the upper cavity 43 at one end of the culture shell 1 and the lower cavity 44 at the other end, respectively. Compared with the traditional parallel straight tube configuration of hollow fiber tubes, the spiral flow channel used in the hollow fiber bundles 5 of this invention has the following advantages: First, the spiral flow channel of the hollow fiber bundles 5 introduces centrifugal force and secondary flow (Dean vortex), periodically disturbing the boundary layer and inhibiting excessive local aggregation of cells on the outer wall of the hollow fiber bundles 5. Second, the curved tube structure of the hollow fiber bundles 5 causes the fluid direction to change continuously, reducing the tendency of cells to migrate and accumulate along the fiber axis, and reducing the risk of local blockage and nutrient depletion caused by cell aggregation. Third, the spiral geometry of the hollow fiber bundles 5 extends the effective length of the flow channel, improving the nutrient exchange path under the same shell size.

[0057] Two sets of pumping components 6 (preferably diaphragm pumps in an alternating tangential flow ATF system) are provided on the outer side of the culture shell 1. The pumping directions of the two sets of pumping components 6 are opposite, and the output ends of the pumping components 6 are respectively connected to the lower chamber 44 on different sides, and the receiving ends of the pumping components 6 are respectively connected to the upper chamber 43 on different sides (specifically as follows). Figure 4 (As shown).

[0058] Specifically, based on the aforementioned structural arrangement, the following flow paths will be formed (with... Figure 4 For example, define the left and right directions, and combine them with the flow diagram. Figure 4 and Figure 5 As shown): The upper cavity 43 on the left side of the culture shell 1 is connected to the lower cavity 44 on the right side through corresponding hollow fiber bundles 5, while the lower cavity 44 on the left side is connected to another symmetrical hollow fiber bundle 5 of the upper cavity 43 on the right side; with the cooperation of two sets of pumping components 6, one set of pumping components 6 realizes: pumping liquid into the lower cavity 44 on the left side, and the liquid is transported through the hollow fiber bundle 5 to the upper cavity 43 on the right side for collection and outflow, while the other set of pumping components 6 realizes: pumping liquid into the lower cavity 44 on the right side, and the liquid is transported through the hollow fiber bundle 5 to the upper cavity 43 on the left side for collection and outflow; thus forming two opposite, relatively independent, and centrally symmetrical circulating perfusion paths inside the culture shell 1.

[0059] Each pump component 6 is connected to a feed pipe 7 and a waste pipe 8. The feed pipe 7 is connected to the output end of the pump component 6 (i.e., the end port where the diaphragm pump generates positive pressure), while the waste pipe 8 is connected to the receiving end of the pump component 6 (i.e., the end port where the diaphragm pump generates negative pressure suction). This design references the mature approach of using a feed pump and an output pump for culture medium replacement in existing alternating tangential flow (ATF) perfusion culture systems. The pump component 6 (diaphragm pump), driven by compressed air, operates during the alternating gas inflow and outflow process. Positive and negative pressures are generated, driving the culture medium to reciprocate between the hollow fiber bundle 5 and the corresponding upper chamber 43 and lower chamber 44. During the positive pressure stroke, the pump 6 pumps fresh culture medium into the circulation loop through the feed pipe 7 via the output end. At the same time, during the negative pressure stroke, the consumed culture medium rich in metabolic waste is extracted from the loop through the waste discharge pipe 8 via the receiving end. Based on the alternating pressure switching of the diaphragm pump, the two pipelines work together to achieve synchronous staggered alternation of feed and waste discharge, maintaining the dynamic constant liquid volume in the circulation loop.

[0060] In this process, the culture medium flows into the spiral hollow fiber bundle 5 from both ends of the culture shell 1 in an alternating direction. The two opposing fluids generate controllable counterflow and secondary flow disturbance in the spiral channel, which can counteract the phenomenon of uneven cell accumulation or scouring on the outer wall of the fiber caused by the continuous unidirectional action of the fluid tangential force in the conventional unidirectional perfusion system. This alleviates the interference of the shear stress gradient generated by the flow in the hollow fiber bundle 5 on the spatial distribution of cells, thereby improving the spatial uniformity of cell density, proliferation activity and metabolite expression in the culture shell 1.

[0061] The spiral hollow fiber bundle 5 structure proposed in this embodiment is compatible with both adherent and suspension cell culture modes: For adherent cells: Before inoculation in the outer cavity (between the outer wall of the hollow fiber bundle 5 and the inner wall of the culture shell 1), extracellular matrix proteins (such as collagen, fibronectin, and laminin) can be introduced through the injection tube 2 for pre-coating, promoting cell adhesion. The counter-current perfusion reduces the scouring effect of shear stress on adherent cells, improving cell adhesion rate and long-term culture survival rate. For suspension cell culture: The hollow fiber bundle 5 does not require pre-coating and can be directly inoculated into the outer cavity. The micro-convective disturbance generated by the counter-current flow maintains a uniform distribution of suspension cells, preventing cell sedimentation and aggregation, and increasing cell density and product expression levels.

[0062] Furthermore, by taking advantage of the characteristic that the temperature decreases along the flow path when the culture medium flows in opposite directions within the spiral hollow fiber bundle 5, a predictable gradient temperature field can be formed within the culture shell 1, enabling precise temperature compensation for specific target areas. At the same time, this opposite perfusion structure can also eliminate the decreasing gradient of nutrients along the fiber axis in traditional unidirectional perfusion, allowing cells in different regions within the culture shell 1 to receive balanced nutrient supply while being in a stable thermal environment. Nutrient delivery and temperature regulation complement and synergistically enhance the overall stability and controllability of the cell microenvironment during long-term high-density culture.

[0063] Example 2:

[0064] As attached Figure 6 As shown, the difference from Embodiment 1 is that the culture shell 1 is also provided with several support disks 9. The support disks 9 are engaged and fixedly connected to the inner side wall of the culture shell 1. Each support disk 9 is provided with several support holes 10 corresponding to hollow fiber bundles 5 (the number of support holes 10 corresponds to the number of hollow fiber bundles 5 and is uniformly circumferentially arrayed on the support disk 9 along the center of the support disk 9, and the inner diameter of the support holes 10 is larger than the tube diameter of the hollow fiber bundles 5). The hollow fiber bundles 5 are all correspondingly inserted into the corresponding support holes 10 and are attached to the inner wall of the support holes 10.

[0065] By inserting and attaching each hollow fiber bundle 5 into the corresponding support hole 10, a multi-point radial constraint structure is formed inside the culture shell 1. This effectively limits the relative displacement and flexural deformation of the hollow fiber bundle 5 caused by fluid pulsation or mechanical vibration during long-term pumping of culture medium, reducing the probability of fiber bundle morphology changes disrupting the uniformity of nutrient delivery space. It also reduces the risk of tube wall rupture or fatigue damage caused by excessive bending or mutual friction of the fiber bundles, improving the mechanical stability of the culture system. In addition, when the inner cavity of the hollow fiber bundle 5 is filled with culture medium and under pumping pressure, the tube wall of the hollow fiber bundle 5 will undergo slight radial expansion due to internal pressure (i.e., in response to the rate of nutrient absorption by cells and whether the hollow fiber bundle 5 is blocked, etc.), causing the initial gap between its outer wall and the inner wall of the support hole 10 to change. This generates contact pressure parameters related to the internal pressure and fiber rigidity, which can provide mechanical sensing basis for subsequent monitoring of local pressure changes and blockage status within the fiber bundle.

[0066] Correspondingly, strain sensors 11 are provided at each of the support holes 10. The acquisition end of each strain sensor 11 is attached to the outer wall of the hollow fiber bundle 5 located at the support hole 10. The strain sensor 11 is used to collect the strain signal generated by the hollow fiber bundle 5 due to the flow of culture medium and cell growth in real time. The cell culture equipment for cell biology experiments based on artificial intelligence proposed in this embodiment also includes a processing system. Several strain sensors 11 and pump components 6 are all connected to the processing system. Each feed pipe 7 is connected to a feed solenoid valve, and each waste discharge pipe 8 is connected to a waste discharge solenoid valve. Both the feed solenoid valve and the waste discharge solenoid valve are connected to the processing system.

[0067] The processing system includes a strain data analysis module, a uniformity assessment module, and a feedback control module. The strain data analysis module has a built-in nutrient exchange prediction model.

[0068] The strain data analysis module receives strain signals collected by strain sensors 11 deployed at each support hole 10 in real time at a fixed sampling frequency (1Hz). It performs digital filtering, baseline drift removal, and normalization preprocessing on the raw strain data to eliminate environmental noise and temperature drift interference. Subsequently, it performs time series analysis on the preprocessed strain sequence to extract multidimensional feature vectors including strain mean, peak value, fluctuation frequency, and trend slope.

[0069] The nutrient exchange prediction model is a deep learning-based classification and regression model, preferably employing a Long Short-Term Memory (LSTM) network or a one-dimensional convolutional neural network (1D-CNN) architecture. The model uses strain feature sequences recorded in historical culture batches and the degree of local blockage of hollow fiber bundles 5 (denoted by transmembrane pressure change rate or residual flow percentage) obtained offline at corresponding times as training samples. Supervised learning is employed until the model's prediction accuracy on the validation set reaches a preset threshold (95%). During operation, the model receives feature vectors output from the strain data analysis module in real time and outputs the predicted degree of local blockage for each hollow fiber bundle 5, which can be in the form of a blockage probability (0–1) or a blockage level (e.g., no blockage, mild, moderate, severe).

[0070] The uniformity assessment module constructs spatial blockage distribution maps along the axial and circumferential directions inside the culture shell 1 based on the predicted blockage degree of all hollow fiber bundles 5. It quantitatively assesses the spatial uniformity of nutrient delivery by calculating the coefficient of variation (CV) and spatial autocorrelation index (Moran's I) of the blockage degree. Simultaneously, it maps the blockage distribution to a spatial consistency index of cell metabolic activity using a historical experience model. When the spatial uniformity index is below a preset threshold (CV value greater than 0.3) and the duration exceeds a set time window (30 minutes), the culture uniformity is deemed unsatisfactory.

[0071] Based on the prediction and data acquisition from the uniformity assessment module, the processing system may also include a spatial sampling optimization module to improve the representativeness of sampling points. This module reconstructs a continuous distribution map of nutrient concentration and temperature within the culture shell 1 using Kriging interpolation, based on the spatial distribution of strain sensor 11 and temperature acquisition device 12. Combined with historical data, it automatically identifies key areas with the largest nutrient gradients and significant differences in cell activity, guiding the sampling strategies of injection tube 2 and acquisition tube 3. This ensures that the sampling results reflect the true state of the overall culture environment and reduces the misleading influence of local outliers on experimental conclusions.

[0072] Upon receiving a trigger signal from the uniformity assessment module, the feedback control module automatically generates and outputs control commands based on the current blockage distribution pattern and severity, invoking preset fuzzy logic rules or lightweight optimization algorithms. These control commands include: adjusting the pumping rates of the two sets of pump components 6 (increasing the inflow rate on the blockage side by 5%–15%), switching the pumping direction sequence (shortening the direction alternation cycle or changing simultaneous bidirectional pumping to alternating pumping), and adjusting the feed flow rate of the feed pipe 7 (proportionally increasing the fresh culture medium inflow rate). Each command is sent to the corresponding pump component 6, feed solenoid valve, or waste discharge solenoid valve via the processing system's I / O interface, achieving dynamic maintenance of spatial consistency in nutrient delivery and cell metabolism within the culture shell 1 without manual intervention.

[0073] Example 3:

[0074] As attached Figure 7 As shown, the difference from Embodiment 2 is that the culture shell 1 is provided with a number of temperature acquisition devices 12. Specifically, a number of temperature acquisition devices 12 (preferably thermocouples or RTD temperature sensors) are arranged at intervals along the axial and circumferential directions in the culture shell 1 to collect temperature data at different spatial locations in the culture shell 1 in real time and calculate the real-time temperature difference between any two points. All of the temperature acquisition devices 12 are connected to the processing system signal.

[0075] The processing system also includes a temperature control module, which has a built-in temperature field prediction model.

[0076] The temperature field prediction model employs a data-driven machine learning model, preferably a radial basis function neural network (RBF-NN) or a Gaussian process regression (GPR) model. The training process for the temperature field prediction model is as follows: In historical batches of cell culture, temperature gradient response data for each region within the culture shell 1 are recorded using temperature acquisition device 12 under different environmental conditions (including room temperature fluctuations and changes in circulation pump speed) and different feeding temperature settings. Simultaneously, the optimal control parameters required to achieve the target temperature field (overall isothermal or a preset gradient temperature field) are recorded (including the heating / cooling temperature settings of the culture medium in the two sets of feeding tubes 7, the flow rate of the pump 6, etc.). The final function of the temperature field prediction model is to use the collected temperature gradient data as input features and the corresponding optimal control parameters as output labels.

[0077] During real-time operation, the temperature control module inputs the real-time temperature difference data fed back by each temperature acquisition unit 12 into the trained temperature field prediction model. Based on the deviation between the current temperature gradient distribution and the preset target temperature field, the temperature field prediction model calculates and outputs adjustment instructions for the culture medium temperature in the two sets of feeding tubes 7 (specifically, the target temperature setpoint of the independent heating / cooling device upstream of the two sets of feeding tubes 7) or directly outputs temperature control suggestions (for operators to confirm and execute). When the CV value exceeds the preset threshold (e.g., 0.05), the temperature field prediction model automatically adjusts the temperature setpoint of the culture medium in the two sets of feeding tubes 7 and the flow rate of the pump component 6 to achieve temperature uniformity control in the culture area within ±0.2℃, meeting the requirements of high-sensitivity cell experiments for thermal environment consistency.

[0078] To adapt to the optimal growth temperature of different cell types (e.g., mammalian cells 37℃, insect cells 27℃, fish cells 15-25℃), this embodiment defines two sets of independent heating / cooling modules upstream of the feeding tubes (7), which can independently adjust the inlet temperature of the culture medium, with an adjustment range of 15℃~40℃ and an accuracy of ±0.1℃. Through the synergy of reverse perfusion and spiral flow channel, a stable and programmable gradient temperature field can be formed in the culture shell (1), with a maximum temperature difference of up to 5℃ / cm, meeting the needs of heat stimulation experiments or temperature-sensitive cell co-culture.

[0079] The above-mentioned control method utilizes the culture medium with a preset temperature difference at both ends of the culture shell 1 to flow simultaneously into the spiral hollow fiber bundle 5 in opposite directions. Combined with the temperature decrease along the flow path of the culture medium due to heat exchange with the tube wall, a predictable and adjustable gradient temperature control field is formed inside the culture shell 1. That is, by adjusting the temperature difference and pumping rate of the culture medium flowing into the two ends in real time, precise temperature compensation for a target area inside the culture shell 1 can be achieved (e.g., raising the temperature of the low-temperature area or lowering the temperature of the high-temperature area). This ensures the long-term uniformity of the cell culture environment temperature inside the culture shell 1 or the gradient distribution that can be set as needed, and improves the intelligence level and spatial accuracy of temperature control.

[0080] Example 4:

[0081] As attached Figure 8 As shown, the difference from Embodiment 2 is that several isolation culture plates are fixedly connected inside the culture shell 1; several partition plates (not shown in the figure) are fixedly connected inside the rectifier shell, and several unit feed pipes (not shown in the figure) are connected to the rectifier shell. The unit feed pipes are connected to the hollow fiber bundles in the area enclosed between the corresponding partition plates; the isolation culture plates divide the originally connected outer cavity space into multiple physically independent cell culture areas, and the several partition plates divide the rectifier chamber into multiple independent fluid distribution sub-units. The fluid distribution sub-units are independently connected to the several hollow fiber bundles corresponding to the culture area through the corresponding unit feed pipes, thereby constructing a complete and non-interfering culture medium circulation loop for each culture area from pump liquid components, feeding and waste discharge to hollow fiber bundle perfusion.

[0082] By utilizing spatial isolation and fluid loop decoupling, different cell types or experimental conditions can be inoculated in different culture zones, and the perfusion rate, nutrient supply, temperature, and metabolic waste removal of each zone can be independently controlled. This enables parallel, synchronous, and cross-contamination-free co-culture of multiple cell types within the same device, improving equipment utilization efficiency and experimental throughput. It supports cell-cell interaction studies (such as non-contact co-culture), reduces competition or signal interference caused by mixed culture, and allows each zone to be independently configured with optimal culture conditions to meet personalized and differentiated cell culture needs. This further expands the flexibility of the device in applications such as drug screening, tissue engineering, and multi-cell model construction.

[0083] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A cell culture device for cell biology experiments based on artificial intelligence, comprising a cylindrical culture shell (1), wherein the top and bottom of the culture shell (1) are respectively connected to an injection tube (2) and a collection tube (3), and the culture shell (1) contains a plurality of hollow fiber bundles (5), characterized in that, The two ends of the culture shell (1) are provided with rectifier components (4). The rectifier components (4) are used to distribute fluid to each hollow fiber bundle (5) and integrate the inlet and outlet flow paths of the fluid. The hollow fiber bundle (5) has a spiral configuration and the two ends of the hollow fiber bundle (5) are centrally symmetrically distributed along the axis of the culture shell (1). The two ends of the hollow fiber bundle (5) are respectively connected to the corresponding rectifier assembly (4). Two sets of pump components (6) are provided on the outside of the culture shell (1). Each pump component (6) is connected to a feeding tube (7) for replenishing fresh culture medium and a waste discharge tube (8) for discharging cell metabolic waste. Each set of pump components (6) is connected to a portion of the hollow fiber bundle (5) through two rectifier components (4). Driven by two sets of pump components (6), the culture medium flows into the corresponding connected hollow fiber bundles (5) from both ends of the culture shell (1) in two directions. The two flow directions of the culture medium fluid flow alternately in the culture shell (1) in the direction of mutual proximity.

2. The cell culture device for cell biology experiments based on artificial intelligence according to claim 1, characterized in that, The rectifier assembly (4) includes a hollow disc-shaped rectifier housing (41), and a partition (42) is fixedly connected inside the rectifier housing (41). The partition (42) divides the space inside the rectifier housing (41) into an upper cavity (43) and a lower cavity (44) that are symmetrical.

3. The cell culture device for cell biology experiments based on artificial intelligence according to claim 2, characterized in that, The two ends of the hollow fiber bundle (5) are connected to the upper cavity (43) at one end of the culture shell (1) and the lower cavity (44) at the other end of the culture shell (1).

4. The cell culture device for cell biology experiments based on artificial intelligence according to claim 3, characterized in that, The output ends of the two sets of pump components (6) are connected to different lower chambers (44), and the receiving ends of the two sets of pump components (6) are connected to the corresponding upper chambers (43).

5. The cell culture device for cell biology experiments based on artificial intelligence according to claim 4, characterized in that, The culture shell (1) is also provided with several support plates (9), and several support holes (10) corresponding to hollow fiber bundles (5) are opened on the support plates (9). The hollow fiber bundles (5) are all inserted and attached to the corresponding support holes (10).

6. The cell culture device for cell biology experiments based on artificial intelligence according to claim 5, characterized in that, Strain sensors (11) are provided at each support hole (10). The strain sensors (11) are used to collect strain signals generated by the hollow fiber bundle (5) due to the flow of culture medium and cell growth in real time.

7. The cell culture device for cell biology experiments based on artificial intelligence according to claim 6, characterized in that, It also includes a processing system, and several strain sensors (11) and pump components (6) are connected to the processing system for signal transmission; The feeding pipe (7) is connected to a feeding solenoid valve, and the waste discharge pipe (8) is connected to a waste discharge solenoid valve. Both the feeding solenoid valve and the waste discharge solenoid valve are connected to the processing system signal.

8. The cell culture device for cell biology experiments based on artificial intelligence according to claim 7, characterized in that, The processing system includes a strain data analysis module, a uniformity assessment module, and a feedback control module. The strain data analysis module has a built-in nutrient exchange prediction model. The strain data analysis module is used to receive real-time strain data collected by several strain sensors (11) and perform time-series analysis and feature extraction on each strain data. The nutrient exchange prediction model is pre-trained based on historical strain data and corresponding blockage degree labels, and is used to output the prediction results of the local blockage degree of each hollow fiber bundle (5); The uniformity assessment module is used to assess the spatial uniformity of nutrient delivery and the spatial consistency of cell metabolic activity within the culture shell (1) based on the predicted results of the local blockage degree of each hollow fiber bundle (5). When the uniformity assessment module determines that the culture uniformity is lower than the preset threshold, the feedback control module automatically generates control instructions. The control instructions include adjusting the pumping rate, pumping direction and timing of the two sets of pump liquid components (6) or the feed flow rate of the feed pipe (7) to achieve adaptive feedback adjustment of the culture medium perfusion parameters.

9. The cell culture device for cell biology experiments based on artificial intelligence according to claim 8, characterized in that, The culture shell (1) is equipped with several temperature acquisition devices (12). The temperature acquisition devices (12) are used to collect the real-time temperature difference between different areas inside the culture shell (1). All the temperature acquisition devices (12) are connected to the processing system signal.

10. The cell culture device for cell biology experiments based on artificial intelligence according to claim 9, characterized in that, The processing system also includes a temperature control module, which has a built-in temperature field prediction model. The temperature field prediction model is constructed by training on historical data of temperature gradient inside the culture shell (1) recorded and collected during cell culture and the corresponding optimal control parameters. The temperature control module is used to input real-time temperature difference data into the trained temperature field prediction model. The temperature field prediction model calculates and outputs adjustment instructions or temperature control suggestions for the culture medium temperature in the two sets of feeding tubes (7). By utilizing the culture medium with temperature difference at both ends of the culture shell (1) to flow into the hollow fiber bundle (5) in opposite directions, and combined with the characteristic that the temperature decreases along the flow path of the culture medium in the spiral flow channel, a predictable gradient temperature control field is formed in the culture shell (1), so as to realize intelligent temperature compensation and regulation of the target area.

Images