Intelligent monitoring and quality evaluation system for stem cell culture environment

Abstract

This application relates to the field of stem cell culture technology and discloses an intelligent monitoring and quality assessment system for stem cell culture environment, including: a multi-parameter sensing module for real-time acquisition of multi-dimensional parameters of the stem cell culture environment, wherein the multi-dimensional parameters include temperature (T), CO2, etc. ₂ The system collects data on concentration (C), dissolved oxygen (DO), pH (P), and culture medium flow rate (V), converting the collected data into digital signals. The predictive analysis module receives real-time data from the multi-parameter sensing module and generates output results based on a preset dynamic prediction algorithm. By predicting changes in environmental parameters in advance through the dynamic prediction algorithm, and combining this with real-time optimization through residual correction, the fluctuations in key indicators such as temperature and CO₂ concentration can be minimized. For example, reducing temperature fluctuations during embryonic stem cell culture reduces the problem of decreased cell activity caused by sudden environmental changes, allowing stem cells to grow in a more stable microenvironment.

Description

Technical Field

[0001] This invention relates to the field of stem cell culture technology, specifically to an intelligent monitoring and quality assessment system for stem cell culture environment. Background Technology

[0002] As a type of cell with self-renewal and multi-directional differentiation potential, stem cells have irreplaceable value in regenerative medicine, disease model construction, and drug development.

[0003] Monitoring the stem cell culture environment primarily relies on traditional CO2 incubators and other equipment to regulate basic parameters. Built-in sensors monitor parameters such as temperature and CO2 concentration at specific points, and a simple PID control algorithm maintains these parameters within set ranges. Quality assessment largely depends on offline methods, such as flow cytometry to detect cell viability, qPCR analysis of pluripotency gene expression, or microscopic observation of cell morphology to indirectly determine culture quality. Some laboratories manually record environmental data and test results, adjusting culture conditions to optimize results.

[0004] However, traditional monitoring systems have a lag in regulating environmental parameters. They can only be passively adjusted after the parameters deviate from the set values. Furthermore, they are unable to cope with the coupled effects between parameters such as temperature, CO2 concentration, and dissolved oxygen, resulting in large fluctuations in the culture environment (such as temperature fluctuations often exceeding ±0.2℃ and CO2 concentration deviations exceeding ±0.1%), which can easily lead to a decrease in stem cell activity or abnormal differentiation direction. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an intelligent monitoring and quality assessment system for stem cell culture environments, which solves the problem of decreased cell activity caused by sudden environmental changes.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an intelligent monitoring and quality assessment system for stem cell culture environment, comprising:

[0007] Multi-parameter sensing module: used to collect multi-dimensional parameters of the stem cell culture environment in real time, including temperature (T), CO2 concentration (C), dissolved oxygen (DO), pH value (P) and culture medium flow rate (V), and convert the collected data into digital signals;

[0008] Predictive analysis module: Receives real-time data collected by the multi-parameter sensing module and generates output results based on a preset dynamic prediction algorithm;

[0009] Intelligent decision-making module: Calculates the environmental regulation amount based on the output results generated by the predictive analysis module and the preset threshold range of the stem cell culture environment;

[0010] The execution adjustment module receives the environmental adjustment parameters output by the intelligent decision-making module and drives the corresponding actuators to correct the environmental parameters.

[0011] Quality assessment module: Based on the real-time data collected by the multi-parameter sensing module and combined with the preset cell growth state model, the module outputs the assessment results of stem cell culture quality; the intelligent decision-making module also adjusts the environmental regulation amount or issues an early warning signal according to the quality assessment results.

[0012] Algorithm optimization module: Optimizes the dynamic prediction algorithm (first algorithm) in real time through feedback error analysis.

[0013] Preferably, the dynamic prediction algorithm is a recursive prediction algorithm based on lag compensation, used to directly reduce the lag in feedback control, and the formula is:

[0014]

[0015] in:

[0016] Predict the environmental parameter values ​​at the current time k+d (d is the lag time, in seconds).

[0017] : The actual environmental parameter collection values ​​at the i-th historical moment;

[0018] : The adjustment amount of the actuator at historical time j;

[0019] , Model coefficients, representing the weights of historical parameters and adjustment factors, respectively;

[0020] The lag compensation term based on the current rate of change of parameters. ,in This is the compensation coefficient.

[0021] Preferably, the preset threshold range for the stem cell culture environment in the intelligent decision-making module is:

[0022] Temperature (T): 36.5±0.2℃ (mesenchymal stem cells) or 37.0±0.1℃ (embryonic stem cells);

[0023] CO2 concentration (C): 5.0 ± 0.1%;

[0024] Dissolved oxygen (DO): 5%-8% (hypoxia culture) or 18%-21% (noroxic culture);

[0025] pH value (P): 7.2±0.1;

[0026] Culture medium flow rate (V): 0.5-2.0 mL / min;

[0027] The threshold range can be manually adjusted through a human-computer interaction interface or automatically matched with a preset template based on cell type.

[0028] Preferably, the environmental adjustment amount calculated by the intelligent decision-making module is the product of the parameter deviation and the adjustment coefficient, as shown in the formula:

[0029]

[0030] in:

[0031] : Environmental regulation at the current moment;

[0032] : The regulation coefficient matrix (5×5 dimensions) includes the regulation weights of temperature (T), CO2 concentration (C), dissolved oxygen (DO), pH value (P), and culture medium flow rate (V);

[0033] The future parameter values ​​output by the predictive analysis module:

[0034] : The median of the threshold values ​​for stem cell culture environment.

[0035] Preferably, the execution mechanism of the execution adjustment module includes:

[0036] Temperature control mechanism: It adopts a combination of semiconductor heating element and Peltier cooler, supporting dynamic temperature compensation from -0.5℃ to +1℃;

[0037] CO2 regulating mechanism: CO2 cylinder passage controlled by a high-precision solenoid valve, gas flow range 0-20%, regulating accuracy ±0.05%;

[0038] Dissolved oxygen regulation mechanism: combined system of oxygen permeation membrane and nitrogen purging, response time ≤8s;

[0039] pH adjustment mechanism: A dual-pump linked micro-injection device for acid and alkali solutions, with a single injection volume accuracy of ±0.1μL;

[0040] Flow rate regulation mechanism: a peristaltic pump driven by a stepper motor, with a flow rate control accuracy of ±0.01mL / min.

[0041] Preferably, in the dynamic prediction algorithm, the lag time d is determined through offline modeling using historical data, specifically:

[0042]

[0043] Where N is the historical data sample size (N≥1000 groups).

[0044] Preferably, the algorithm optimization module includes correcting the predicted value using the following formula:

[0045]

[0046] The prediction residual at time k is the difference between the actual measured value and the predicted value before time d of the first algorithm.

[0047] The corrected parameter predictions for the future time k+d:

[0048] : Residual correction coefficient, the value of which ranges from 0.3 to 0.7 and is adaptively adjusted according to the type of each environmental parameter;

[0049] , This is the preset deviation threshold.

[0050] This invention provides an intelligent monitoring and quality assessment system for stem cell culture environments. It has the following beneficial effects:

[0051] 1. This invention uses a dynamic prediction algorithm to predict changes in environmental parameters in advance, and then combines it with real-time optimization of residual correction to reduce the fluctuation range of key indicators such as temperature and CO2 concentration. For example, when culturing embryonic stem cells, the temperature fluctuation is reduced, which reduces the problem of decreased cell activity caused by sudden changes in the environment, allowing stem cells to grow in a more stable microenvironment.

[0052] 2. This invention feeds back quality data such as cell viability and differentiation potential to the intelligent decision-making module in real time. Once an abnormal cell growth state is detected, the system will automatically increase the adjustment weight of key parameters or trigger an early warning, forming a closed loop of "environmental monitoring - quality assessment - strategy adjustment". Attached Figure Description

[0053] Figure 1 This is a system block diagram of the present invention. Detailed Implementation

[0054] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. 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.

[0055] Example:

[0056] Please see the appendix Figure 1 This invention provides an intelligent monitoring and quality assessment system for stem cell culture environment, including a multi-parameter sensing module, a predictive analysis module, an intelligent decision-making module, an execution adjustment module, a quality assessment module, and an algorithm optimization module.

[0057] Specifically, it includes:

[0058] I. Multi-parameter sensing module, used to collect key parameters of the stem cell culture environment in real time, using a distributed sensor array, including:

[0059] Temperature sensor: High-precision Pt1000 platinum resistance sensor is selected, with a measurement range of 35-40℃, accuracy of ±0.05℃, and sampling frequency of 10Hz;

[0060] CO2 sensor: Employs a non-dispersive infrared (NDIR) sensor with a measurement range of 0-10%, accuracy of ±0.01%, and response time ≤5s;

[0061] Dissolved oxygen (DO) sensor: fiber optic fluorescence dissolved oxygen probe, measurement range 0-30%, resolution 0.1%;

[0062] pH sensor: Glass electrode pH probe, measurement range 6.5-8.0, accuracy ±0.01 pH;

[0063] Culture medium flow rate sensor: electromagnetic flow meter, measuring range 0-5 mL / min, accuracy ±0.01 mL / min.

[0064] The analog signals collected by the sensor are converted into digital signals by a 16-bit AD converter and transmitted to the system main control unit (such as an STM32H743 microprocessor) via an RS485 bus to realize real-time data storage and preprocessing (such as filtering and noise reduction).

[0065] Furthermore, the sensor array of the multi-parameter sensing module is distributed inside the incubator. The temperature and CO2 sensors are fixed 10cm above the culture dish, the dissolved oxygen and pH sensors are inserted into the culture medium (2cm deep), and the flow rate sensor is connected in series in the culture medium circulation pipeline to ensure that the collected data reflects the true state of the cell microenvironment.

[0066] Noise reduction was achieved using a moving average filtering method with a sliding window size of 5 sampling points (based on a 10Hz sampling frequency and an equivalent filtering time of 0.5s) to remove transient pulse interference.

[0067] Second, the predictive analysis module, based on real-time data from the multi-parameter sensing module, generates predicted values ​​of environmental parameters at time d in the future through a dynamic prediction algorithm, thus solving the problem of feedback control lag.

[0068] The dynamic prediction algorithm specifically includes the following steps:

[0069] S1. Read the real-time parameter value at the current time k. Parameter values ​​at the i-th historical moment (i=1,2,…,n, n=5) and the adjustment amount at historical j time points (j = 0, 1, ..., m, m = 3);

[0070] S2, Calculate the lag compensation term The compensation coefficient Set according to parameter type (temperature) .6, concentration ;

[0071] S3. Substitute the values ​​into the prediction formula to generate the predicted parameter values ​​for the future time k+d:

[0072]

[0073] Among them, model coefficients (historical parameter weights) and (Adjustment weights) are determined through offline training (e.g., temperature parameters a1=0.3, a2=0.2, ..., b0=0.1).

[0074] Determination of lag time d: calculated offline using historical data.

[0075] Where N=5000 sets of historical data, the temperature regulation lag time was calculated. Concentration adjustment lag time .

[0076] III. The intelligent decision-making module calculates the adjustment amount based on the prediction results and preset thresholds, specifically including:

[0077] When culturing mesenchymal stem cells, the temperature threshold was set to 36.5±0.2℃, and the dissolved oxygen was set to 5%-8% (hypoxic culture).

[0078] When culturing embryonic stem cells, the temperature threshold is set to 37.0±0.1℃, and the dissolved oxygen is set to 18%-21% (noroxic culture).

[0079] The threshold can be manually adjusted (e.g., the pH threshold can be adjusted to 7.3±0.1 via the interface).

[0080]

[0081] Adjustment coefficient matrix for A 3D diagonal matrix, where the diagonal elements are temperatures. , concentration Dissolved oxygen pH Flow rate ;

[0082] Threshold median Take the midpoint of the threshold range (e.g., temperature) , concentration ).

[0083] IV. The execution adjustment module dynamically corrects environmental parameters through the actuator.

[0084] Temperature regulation mechanism: A combination of a semiconductor heating element (power 50W) and a Peltier cooler (cooling capacity 30W) receives the regulation amount u(k) and controls the power through a PWM signal, supporting dynamic compensation from -0.5℃ to +1℃ (e.g., when the predicted temperature will drop to 36.3℃, the heating element is activated to compensate 0.2℃).

[0085] CO2 regulation mechanism: A high-precision solenoid valve (response time ≤100ms) controls the CO2 cylinder passage and corrects the concentration by adjusting the gas flow rate (0-20% range). For example, if the concentration is predicted to rise to 5.1%, the flow rate is reduced by 0.5% / s.

[0086] Dissolved oxygen regulation mechanism: The oxygen permeation membrane (area 10cm²) is linked with the nitrogen purge pump. When the predicted DO is lower than 5%, the nitrogen purge rate is reduced to increase the oxygen concentration, with a response time ≤8s.

[0087] pH adjustment mechanism: Dual Syringe pumps (accuracy ±0.1μL) load acid (0.1mol / L HCl) and alkali (0.1mol / L NaOH) respectively. When the pH is predicted to drop to 7.1, 0.5μL of alkali is injected.

[0088] Flow rate regulation mechanism: a peristaltic pump driven by a stepper motor (step angle 1.8°), which corrects the flow rate (0.5-2.0 mL / min) by adjusting the rotation speed, with a control accuracy of ±0.01 mL / min.

[0089] V. Quality Assessment Module: This module assesses stem cell quality in real time based on environmental parameters and provides feedback for regulation. Specifically, it includes:

[0090] Cell growth state model: Fusion of two sub-models:

[0091] Cell viability sub-model: based on temperature fluctuation amplitude ( Y) and pH stability index ( ), Han outputs a vitality score from 0-100 (score) Divided into high quality);

[0092] Vitality score The calculation formula is:

[0093]

[0094] in:

[0095] The maximum temperature fluctuation in the past hour (unit: °C);

[0096] Standard deviation of pH value over the past hour;

[0097] Preset weighting coefficients (default) );

[0098] Scoring Instructions For "high quality", For "good", It is considered "bad".

[0099] Differentiation potential sub-model: based on dissolved oxygen rate of change (DO rate) and CO Stability, predicting adipogenic / osteogenic differentiation potential.

[0100] Application of evaluation results: When the vitality score is <60, the intelligent decision-making module automatically issues an audible and visual warning and adjusts the adjustment coefficient matrix. (For example, increase the temperature regulation weight to 1.0).

[0101] Cell viability score (0-100), differentiation potential level (high / medium / low), and overall quality status (normal / warning / abnormal).

[0102] The cell growth status model of the quality assessment module takes real-time data (temperature, CO2 concentration, etc.) from the multi-parameter sensing module as input, and transmits the quality assessment results to the intelligent decision-making module in real time. When the assessment result is lower than the preset threshold (e.g., 60 points), the environmental regulation amount is reduced or an early warning signal is issued.

[0103] The algorithm optimization module improves long-term control accuracy by optimizing the prediction algorithm through residual correction. This includes the following steps:

[0104] S1. Residual Calculation: Real-time Calculation (The deviation between the actual measured value and the predicted value before time d);

[0105] S2, Prediction Correction: Through the formula Correcting future forecasts, including residual correction factors. Set according to parameter type (temperature) concentration

[0106] S3, Dynamic Optimization: When (temperature Temporarily increase The value is increased to 1.2 times the original value, triggering the prediction algorithm. The coefficients are recalibrated (updated offline every 24 hours).

[0107] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A stem cell culture environment intelligent monitoring and quality assessment system, characterized in that, include: Multi-parameter sensing module: used to collect multi-dimensional parameters of the stem cell culture environment in real time, including temperature (T), CO2 concentration (C), dissolved oxygen (DO), pH value (P) and culture medium flow rate (V), and convert the collected data into digital signals; Predictive analysis module: Receives real-time data collected by the multi-parameter sensing module and generates output results based on a preset dynamic prediction algorithm; Intelligent decision-making module: Calculates the environmental regulation amount based on the output results generated by the predictive analysis module and the preset threshold range of the stem cell culture environment; The execution adjustment module receives the environmental adjustment parameters output by the intelligent decision-making module and drives the corresponding actuators to correct the environmental parameters. Quality assessment module: Based on the real-time data collected by the multi-parameter sensing module and combined with the preset cell growth state model, the module outputs the assessment results of stem cell culture quality; the intelligent decision-making module also adjusts the environmental regulation amount or issues an early warning signal according to the quality assessment results. Algorithm optimization module: Optimizes the dynamic prediction algorithm in real time through feedback error analysis.

2. The stem cell culture environment intelligent monitoring and quality evaluation system according to claim 1, wherein, The dynamic prediction algorithm is a recursive prediction algorithm based on lag compensation, used to directly reduce the lag in feedback control. The formula is: in: Predict the environmental parameter values ​​at the current time k+d (d is the lag time, in seconds). : The actual environmental parameter collection values ​​at the i-th historical moment; : The adjustment amount of the actuator at historical time j; , Model coefficients, representing the weights of historical parameters and adjustment factors, respectively; The lag compensation term based on the current rate of change of parameters. ,in This is the compensation coefficient.

3. The intelligent monitoring and quality assessment system for stem cell culture environment according to claim 1, characterized in that, The preset threshold range for the stem cell culture environment in the intelligent decision-making module is as follows: Temperature (T): 36.5±0.2℃ (mesenchymal stem cells) or 37.0±0.1℃ (embryonic stem cells); CO2 concentration (C): 5.0 ± 0.1%; Dissolved oxygen (DO): 5%-8% (hypoxia culture) or 18%-21% (noroxic culture); pH value (P): 7.2±0.1; Culture medium flow rate (V): 0.5-2.0 mL / min; The threshold range can be manually adjusted through a human-computer interaction interface or automatically matched with a preset template based on cell type.

4. The intelligent monitoring and quality assessment system for stem cell culture environment according to claim 1, characterized in that, The environmental adjustment amount calculated by the intelligent decision-making module is the product of the parameter deviation and the adjustment coefficient, as shown in the formula: in: : Environmental regulation at the current moment; : The regulation coefficient matrix (5×5 dimension) includes the regulation weights of temperature (T), CO2 concentration (C), dissolved oxygen (DO), pH value (P), and culture medium flow rate (V); The future parameter values ​​output by the predictive analysis module: : The median of the threshold values ​​for stem cell culture environment.

5. The intelligent monitoring and quality assessment system for stem cell culture environment according to claim 1, characterized in that, The execution mechanism of the execution adjustment module includes: Temperature control mechanism: It adopts a combination of semiconductor heating element and Peltier cooler, supporting dynamic temperature compensation from -0.5℃ to +1℃; CO2 regulating mechanism: CO2 cylinder passage controlled by a high-precision solenoid valve, gas flow range 0-20%, regulating accuracy ±0.05%; Dissolved oxygen regulation mechanism: combined system of oxygen permeation membrane and nitrogen purging, response time ≤8s; pH adjustment mechanism: A dual-pump linked micro-injection device for acid and alkali solutions, with a single injection volume accuracy of ±0.1μL; Flow rate regulation mechanism: a peristaltic pump driven by a stepper motor, with a flow rate control accuracy of ±0.01mL / min.

6. The intelligent monitoring and quality assessment system for stem cell culture environment according to claim 2, characterized in that, In the dynamic prediction algorithm, the lag time d is determined through offline modeling using historical data, specifically: Where N is the historical data sample size (N≥1000 groups).

7. The intelligent monitoring and quality assessment system for stem cell culture environment according to claim 2, characterized in that, The algorithm optimization module includes correcting the predicted value using the following formula: The prediction residual at time k is the difference between the actual measured value and the predicted value before time d of the first algorithm. The corrected parameter predictions for the future time k+d: : Residual correction coefficient, the value of which ranges from 0.3 to 0.7 and is adaptively adjusted according to the type of each environmental parameter; , This is the preset deviation threshold.

Images