Methods and systems for measuring cell culture plate temperature, training and deployment of models.

Abstract

This disclosure relates to the field of cell culture plate temperature measurement, and provides a method, model training and deployment method, and system for cell culture plate temperature measurement. The training method includes: acquiring temperature time-series data corresponding to the four outer walls of the cell culture plate and actual chamber temperature data at a specified location on the cell culture plate; normalizing the temperature time-series data to obtain normalized temperature time-series data; inputting the normalized temperature time-series data into the cell culture plate temperature measurement model for iterative training to obtain corresponding chamber temperature sequence prediction data; performing inverse normalization on the chamber temperature sequence prediction data to obtain chamber temperature prediction data; and adjusting the model parameters based on the chamber temperature prediction data and its corresponding actual chamber temperature data until the model error reaches a preset accuracy requirement. This disclosure utilizes a cell culture plate temperature measurement model to achieve non-invasive and accurate measurement of the chamber temperature of a cell culture plate.

Description

Technical Field

[0001] This disclosure relates to the field of cell culture plate temperature measurement technology, and in particular to a cell culture plate temperature measurement method, a model training and deployment method and system. Background Technology

[0002] Cells are extremely sensitive to temperature; excessively high or low temperatures can significantly affect their activity, thus impacting experimental results. Currently, when observing cells under a microscope, small culture containers are typically used to hold the cells and place them on the microscope platform. These small containers are compact and transparent, facilitating cell observation. However, their sealing is often less effective than larger containers, which can cause fluctuations in the ambient temperature. Therefore, temperature control is crucial for ensuring cell viability and accurate experimental results when observing cells under a microscope.

[0003] Existing incubator temperature control methods typically employ one or more Proportional-Integral-Derivative (PID) algorithms to control the power of one or more heaters within the incubator, thereby regulating the incubator temperature. However, this method primarily obtains the ambient temperature within the incubator, rather than the actual temperature inside the cell culture plate. Since there is usually a difference between the ambient temperature inside the incubator and the temperature inside the cell culture plate, accurately measuring and controlling the temperature inside the cell culture plate is crucial for effectively ensuring cell viability and experimental accuracy.

[0004] Existing technologies also provide several methods for measuring the temperature inside cell culture plates, which can be broadly categorized into the following four types.

[0005] Method 1: Place a temperature sensor inside the incubator to indirectly measure the temperature of the well plate's inner chamber by monitoring and controlling the temperature inside the incubator. The accuracy of this method depends on the incubator's sealing and insulation performance, but high sealing and good insulation performance usually increase the size and weight of the incubator, making it unsuitable for certain scenarios.

[0006] Method 2: Using non-contact temperature measurement equipment such as thermal imagers, the temperature of the well plate surface is measured using infrared technology. While this method can provide accurate temperature data of the well plate surface, the equipment is expensive and there is a potential risk of thermal radiation damage to the cells.

[0007] Method 3: A dedicated reference plate is placed inside the incubator. This reference plate is equipped with a temperature sensor, and its temperature is used to indirectly represent the temperature of the actual cell culture plate. While this method can improve the accuracy of cell culture plate temperature measurement, the installation and maintenance of the reference plate are relatively cumbersome, and it increases the complexity and size of the equipment.

[0008] Method 4: In-situ temperature measurement by directly embedding temperature sensor electrodes inside the cell culture plate. While this method can directly obtain temperature information within the cell culture plate and ensure measurement accuracy, the sensor electrodes used are usually disposable products, resulting in high costs. This makes it unsuitable for large-scale experiments and can also affect the transparency of the cell culture plate, thus adversely impacting optical observation of the cells. Summary of the Invention

[0009] This disclosure aims to address at least one of the problems existing in the prior art by providing a method for measuring temperature in cell culture plates, a method and system for training and deploying models.

[0010] One aspect of this disclosure provides a training method for a cell culture plate temperature measurement model, the training method comprising:

[0011] Acquire time-series temperature data corresponding to the four outer walls of the cell culture plate and actual chamber temperature data at a specified location on the cell culture plate; wherein, the specified location on the cell culture plate includes the location of different regions of the cell culture plate;

[0012] The temperature time series data is normalized to obtain normalized temperature time series data;

[0013] The normalized temperature time series data is input into the cell culture plate temperature measurement model for iterative training to obtain the chamber temperature sequence prediction data corresponding to the normalized temperature time series data.

[0014] The chamber temperature sequence prediction data is inversely normalized to obtain the chamber temperature prediction data.

[0015] Based on the predicted chamber temperature data and the corresponding actual chamber temperature data, the parameters of the cell culture plate temperature measurement model are adjusted until the error of the cell culture plate temperature measurement model reaches the preset accuracy requirement.

[0016] Optionally, the cell culture plate temperature measurement model includes a sequence input layer, a long short-term memory network layer, and a fully connected layer connected in sequence;

[0017] The sequence input layer is used to receive the normalized temperature time series data and convert the normalized temperature time series data into corresponding tensor data.

[0018] The long short-term memory network layer is used to receive the tensor data, process and predict the dependencies in the tensor data, and output the corresponding sequence data;

[0019] The fully connected layer is used to perform tensor compression and processing on the sequence data to obtain the corresponding chamber temperature sequence prediction data.

[0020] Optionally, the normalization process for the temperature time series data to obtain normalized temperature time series data includes:

[0021] Calculate the mean and standard deviation of the temperature time series data;

[0022] The temperature time series data is normalized based on the average value and the standard deviation to obtain the normalized temperature time series data.

[0023] Another aspect of this disclosure provides a method for deploying a cell culture plate temperature measurement model, the deployment method comprising:

[0024] Obtain the cell culture plate temperature measurement model trained using the training method described above.

[0025] The trained cell culture plate temperature measurement model is quantized by converting it from a floating-point model to a fixed-point integer model to obtain the quantized cell culture plate temperature measurement model.

[0026] The quantified cell culture plate temperature measurement model was optimized by adjusting its memory usage and computing resource allocation to obtain an optimized cell culture plate temperature measurement model.

[0027] Benchmark tests were conducted on the optimized cell culture plate temperature measurement model to evaluate its performance.

[0028] The optimized cell culture plate temperature measurement model was converted into target code.

[0029] The target code is deployed to the chip.

[0030] Another aspect of this disclosure provides a method for measuring the temperature of a cell culture plate, the method comprising:

[0031] Obtain the actual temperature data corresponding to the four outer walls of the cell culture plate;

[0032] The actual temperature data is normalized to obtain actual temperature time series data;

[0033] The actual temperature time series data is input into the trained cell culture plate temperature measurement model to obtain the corresponding model output results; wherein, the trained cell culture plate temperature measurement model is trained using the training method for cell culture plate temperature measurement models described above;

[0034] The model output is inversely normalized to obtain the chamber temperature of the cell culture plate.

[0035] Optionally, the cell culture plate temperature measurement method further includes:

[0036] Based on the temperature measured in the chamber, the environment of the incubator containing the cell culture plate is controlled; and / or,

[0037] The measured temperature in the chamber is displayed visually.

[0038] Another aspect of this disclosure provides a training system for a cell culture plate temperature measurement model, the training system comprising:

[0039] The acquisition module is used to acquire the temperature time series data corresponding to the four outer walls of the cell culture plate and the actual temperature data of the chamber at a specified location of the cell culture plate; wherein, the specified location of the cell culture plate includes the location of different regions of the cell culture plate;

[0040] The normalization module is used to normalize the temperature time series data to obtain normalized temperature time series data.

[0041] An iterative module is used to input the normalized temperature time series data into the cell culture plate temperature measurement model for iterative training to obtain the chamber temperature sequence prediction data corresponding to the normalized temperature time series data.

[0042] The inverse normalization module is used to perform inverse normalization processing on the chamber temperature sequence prediction data to obtain chamber temperature prediction data;

[0043] The adjustment module is used to adjust the parameters of the cell culture plate temperature measurement model based on the predicted chamber temperature data and the corresponding actual chamber temperature data, until the error of the cell culture plate temperature measurement model reaches the preset accuracy requirement.

[0044] Another aspect of this disclosure provides a deployment system for a cell culture plate temperature measurement model, the deployment system comprising:

[0045] The model acquisition module is used to acquire the cell culture plate temperature measurement model trained using the training method described above.

[0046] The deployment module is used to quantize the trained cell culture plate temperature measurement model, converting it from a floating-point model to a fixed-point integer model to obtain a quantized cell culture plate temperature measurement model; optimize the quantized cell culture plate temperature measurement model by adjusting its memory usage and computing resource allocation to obtain an optimized cell culture plate temperature measurement model; benchmark the optimized cell culture plate temperature measurement model to evaluate its performance; convert the optimized cell culture plate temperature measurement model into target code; and deploy the target code to the chip.

[0047] Another aspect of this disclosure provides a cell culture plate temperature measurement system, the cell culture plate temperature measurement system comprising:

[0048] The temperature acquisition module is used to acquire the actual temperature data corresponding to the four outer walls of the cell culture plate.

[0049] The normalization processing module is used to normalize the actual temperature data to obtain actual temperature time series data.

[0050] The prediction module is used to input the actual temperature time series data into the trained cell culture plate temperature measurement model to obtain the corresponding model prediction results; wherein, the trained cell culture plate temperature measurement model is trained using the training method for cell culture plate temperature measurement models described above;

[0051] The inverse normalization module is used to inverse normalize the model prediction results to obtain the chamber measurement temperature of the cell culture plate.

[0052] Optionally, the cell culture plate temperature measurement system further includes:

[0053] An environmental control module is used to control the environment of the incubator containing the cell culture plate based on the measured temperature of the chamber; and / or,

[0054] A visualization module is used to visualize the measured temperature of the chamber.

[0055] Compared with the prior art, this disclosure addresses the issue of high cleanliness requirements in cell culture by designing a non-invasive measurement method. It utilizes a cell culture plate temperature measurement model to model the heat transfer process from the inner chamber of the cell culture plate to the surrounding area, and employs deep learning technology to process the temperature time series data of the cell culture plate, which can effectively achieve accurate measurement of the chamber temperature of the cell culture plate. Attached Figure Description

[0056] One or more embodiments are illustrated by way of example with the corresponding pictures in the accompanying drawings. These illustrations do not constitute a limitation on the embodiments. Elements with the same reference numerals in the drawings are denoted as similar elements. Unless otherwise stated, the figures in the drawings are not to be limited by scale.

[0057] Figure 1 A flowchart illustrating a training method for a cell culture plate temperature measurement model provided in one embodiment of this disclosure;

[0058] Figure 2 A schematic diagram of temperature acquisition and region division of a cell culture plate provided for another embodiment of this disclosure;

[0059] Figure 3 A flowchart illustrating the process of obtaining chamber temperature prediction data using a cell culture plate temperature measurement model, as provided in another embodiment of this disclosure;

[0060] Figure 4 A flowchart illustrating a method for deploying a cell culture plate temperature measurement model according to another embodiment of this disclosure;

[0061] Figure 5 A schematic flowchart of a cell culture plate temperature measurement method provided for another embodiment of this disclosure;

[0062] Figure 6 A schematic diagram of the structure of a training system for a cell culture plate temperature measurement model provided in another embodiment of this disclosure;

[0063] Figure 7 A schematic diagram of the structure of a deployment system for a cell culture plate temperature measurement model provided in another embodiment of this disclosure;

[0064] Figure 8 A schematic diagram of a cell culture plate temperature measurement system provided for another embodiment of this disclosure;

[0065] Figure 9 This is a schematic diagram of the workflow of a cell culture plate temperature measurement system provided for another embodiment of this disclosure. Detailed Implementation

[0066] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the various embodiments of this disclosure will be described in detail below with reference to the accompanying drawings. However, those skilled in the art will understand that many technical details have been provided in the various embodiments of this disclosure to facilitate a better understanding of the disclosure. However, the technical solutions claimed in this disclosure can be implemented even without these technical details and with various variations and modifications based on the following embodiments. The division of the various embodiments below is for ease of description and should not constitute any limitation on the specific implementation of this disclosure. The various embodiments can be combined with and referenced by each other without contradiction.

[0067] One embodiment of this disclosure relates to a training method for a cell culture plate temperature measurement model, the process of which is as follows: Figure 1 As shown, it includes steps S110 to S150.

[0068] Step S110: Obtain the temperature time series data corresponding to the four outer walls of the cell culture plate and the actual temperature data of the chamber at a specified location on the cell culture plate; wherein, the specified location on the cell culture plate includes the location of different regions of the cell culture plate.

[0069] Specifically, this embodiment does not limit the specific method by which the temperature time series data and the actual chamber temperature data are obtained in step S110. The cell culture plate here can be any type of culture plate commonly used in the field of cell culture. For example, the cell culture plate can be a well plate commonly used in the prior art. The following embodiments of this disclosure will use a well plate as an example for illustration.

[0070] For example, combined Figure 2 When the cell culture plate is a well plate, for a well plate 1 with multiple chambers, a person skilled in the art can place the well plate 1 in a clamp 16 and place the clamp 16 containing the well plate 1 under an open incubator with a suitable microscope stage for observation. The outer walls of the well plate 1 are in contact with temperature sensors 2, 3, 4, and 5 on the clamp 16, respectively, to acquire time-series temperature data corresponding to the outer walls of the well plate 1 through temperature sensors 2 to 5. This achieves non-invasive temperature measurement, avoids contamination of the cell culture process, and improves the cleanliness of the cell culture.

[0071] In particular, to make the temperature time series data corresponding to the four outer walls of the cell well plate 1 acquired by temperature sensors 2 to 5 more representative, such as Figure 2As shown, temperature sensors 2 to 5 can be made to contact the middle position of each of the four edges of the cell well plate 1, so that the temperature time series data at the middle position of each of the four edges of the cell well plate 1 can be used as the temperature time series data corresponding to the four outer walls of the cell well plate 1.

[0072] To obtain actual data on the chamber temperature at a specific location in a cell culture plate, especially a well plate, such as... Figure 2 As shown, those skilled in the art can divide the cell plate 1 into five regions: region 11, region 12, region 13, region 14, and region 15. These five regions completely cover all chambers in the cell plate 1 to address the possibility of uneven temperature control among the chambers during culture, enabling temperature monitoring of these five regions. Region 15 is located at the center of the cell plate 1, and regions 11 to 14 are located at the upper left, lower left, lower right, and upper right of region 15, respectively. Next, the most representative chamber from each of the five regions is selected as a sampling point, and the actual temperature data of each sampling point is used as the actual chamber temperature data corresponding to each region in the cell plate. For example, the chamber in the upper left corner of region 11 can be used as sampling point 6, the chamber in the lower left corner of region 12 can be used as sampling point 7, the chamber in the lower right corner of region 13 can be used as sampling point 8, the chamber in the upper right corner of region 14 can be used as sampling point 9, and the chamber in region 15 located slightly to the left of the middle, i.e., in the second row and third column, can be used as sampling point 10.

[0073] To obtain a sufficient dataset for model training, temperature sensors 2 to 5 and sampling points 6 to 10 can be used multiple times, based on possible culture conditions, to obtain the temperature time series data corresponding to the four outer walls of the cell well plate 1 and the actual temperature data of the chamber at a specified location on the cell well plate.

[0074] For example, those skilled in the art can collect 100 datasets containing temperature time series data and actual chamber temperature data through the heating and holding processes.

[0075] Each heating process yields a corresponding heating dataset. In each heating process, the incubator containing cell plate 1 is first heated to raise the temperature from room temperature to the target heating temperature and stabilize it. During the two hours of heating from room temperature, data is collected every second, specifically readings from temperature sensors 2 to 5 and the temperatures at sampling points 6 to 10. This results in a heating dataset consisting of the time-series temperature data of cell plate 1 and the actual temperature data of the chamber. Heating is then stopped for two hours, and the incubator is cooled to room temperature. The target heating temperature ranges from 33℃ to 40℃, with each 0.1℃ increment within this range corresponding to a target heating temperature. This results in 70 heating processes and 70 heating datasets obtained from 70 target heating temperatures. For example, a target heating temperature of 33℃ corresponds to one heating process, 33.1℃ to another, 33.2℃ to yet another, and so on. Since the 33℃-40℃ range can be divided into 71 temperature points with an interval of 0.1℃ (i.e., 33℃, 33.1℃, 33.2℃, 33.3℃, ​​33.4℃, 33.5℃, 33.6℃, ​​33.7℃, 33.8℃, 33.9℃, 34℃, ..., 39℃, 39.1℃, 39.2℃, 39.3℃, 39.4℃, 39.5℃, 39.6℃, 39.7℃, 39.8℃, 39.9℃, 40℃), we can choose to ignore the 33℃ or 40℃ temperature point and use the remaining 70 temperature points as the corresponding 70 target heating temperatures.

[0076] Each incubation process yields a corresponding incubation dataset. In each incubation process, cell plate 1 is first removed from a standard incubator at the target incubation temperature and placed in a culture box for 2 hours. During these 2 hours, data is collected every second, specifically the readings from temperature sensors 2 to 5 and the temperatures at sampling points 6 to 10. This results in an incubation dataset composed of the time-series temperature data of cell plate 1 and the actual temperature data of the chamber. The culture box is preheated to the target incubation temperature for 30 minutes before cell plate 1 is placed in. Subsequently, heating of the culture box is stopped for 2 hours, and the temperature is cooled to room temperature. The target incubation temperature range is 36℃-37℃, with each 0.1℃ interval corresponding to a target incubation temperature. Ten incubation processes are obtained from ten target incubation temperatures. Each of these ten incubation processes is repeated three times, resulting in 30 incubation processes and 30 incubation datasets. For example, a target insulation temperature of 36℃ corresponds to one insulation process, a target insulation temperature of 36.1℃ corresponds to one insulation process, a target insulation temperature of 36.2℃ corresponds to one insulation process, and so on. Since the 36℃-37℃ range can be divided into 11 temperature points when the interval is 0.1℃, namely 36℃, 36.1℃, 36.2℃, 36.3℃, 36.4℃, 36.5℃, 36.6℃, 36.7℃, 36.8℃, 36.9℃, and 37℃, it is possible to ignore the temperature point of 36℃ or 37℃ and use the remaining 10 temperature points as the corresponding 10 target insulation temperatures.

[0077] Step S120: Normalize the temperature time series data to obtain normalized temperature time series data.

[0078] Specifically, normalization can improve the training effect of the model, speed up the convergence speed, and enhance the generalization ability of the model.

[0079] For example, step S120 includes: calculating the mean and standard deviation of the temperature time series data; and normalizing the temperature time series data based on the mean and standard deviation to obtain normalized temperature time series data.

[0080] Specifically, suppose a temperature value X is found in a temperature time series data. μ This represents the average value of the temperature time series data. σ Let X be the standard deviation of the temperature time series data, then X corresponds to the normalized temperature value. It can be represented as .

[0081] Step S130: Input the normalized temperature time series data into the cell culture plate temperature measurement model for iterative training to obtain the chamber temperature sequence prediction data corresponding to the normalized temperature time series data.

[0082] Specifically, step S130 can utilize existing toolkits in MATLAB software to iteratively train the cell culture plate temperature measurement model. During the iterative training process, if the actual number of iterations reaches the preset number of iterations, the cell culture plate temperature measurement model can be considered to have been successfully trained.

[0083] For example, the cell culture plate temperature measurement model includes a sequence input layer, a Long Short-Term Memory (LSTM) network layer, and a fully connected layer connected in sequence.

[0084] The sequence input layer receives normalized temperature time-series data and transforms it into corresponding tensor data. The Long Short-Term Memory (LSTM) network layer receives the tensor data, processes and predicts dependencies within it, and outputs the corresponding sequence data. The fully connected layer performs tensor compression and processing on the sequence data to obtain the corresponding chamber temperature sequence prediction data.

[0085] Specifically, combined Figure 3 Step S130 may specifically include steps S1 to S4.

[0086] Step S1 is used to... Figure 2 The normalized temperature time series data of the four well walls of cell culture plate 1 is input into the sequence input layer of the cell culture plate temperature measurement model. Step S2 is used to receive the normalized temperature time series data through the sequence input layer of the cell culture plate temperature measurement model, convert it into corresponding tensor data, and then pass it to the subsequent layer, namely the Long Short-Term Memory (LSTM) network layer. The LSTM network layer is a special type of recurrent neural network layer used to process and predict dependencies in time series data or sequence data. For example, the LSTM network layer can be configured with 128 LSTM units. Through its memory cells and gating mechanism, the LSTM units can capture and retain information on long-term dependencies, which is suitable for modeling the process of heat transfer from the inner chamber of a cell culture plate to its surroundings over time. Step S3 is used to receive tensor data through the LSTM network layer, process and predict dependencies in the tensor data, and output the corresponding sequence data. Step S4 integrates the features extracted from the previous layer (i.e., the Long Short-Term Memory network layer) using the fully connected layer and generates the final output. This allows for tensor compression and processing of the sequence data to obtain the corresponding chamber temperature sequence prediction data. Each neuron in the fully connected layer is connected to all neurons in the previous layer, and the input features are processed through weighted summation and a nonlinear activation function.

[0087] Step S140: Perform inverse normalization on the chamber temperature sequence prediction data to obtain the chamber temperature prediction data.

[0088] For example, combined Figure 3 Step S140, as step S5, is used to perform inverse normalization processing on the chamber temperature sequence prediction data output from the fully connected layer in the cell culture plate temperature measurement model to obtain... Figure 2 The temperatures of the five wells in cell plate 1 shown are the predicted chamber temperatures corresponding to cell plate 1.

[0089] Here, the inverse normalization process is the restoration process corresponding to the normalization process in step S120, which restores the chamber temperature sequence prediction data output by the cell culture plate temperature measurement model to chamber temperature prediction data with actual physical meaning through inverse normalization process.

[0090] Step S150: Based on the predicted chamber temperature data and the corresponding actual chamber temperature data, adjust the parameters of the cell culture plate temperature measurement model until the error of the cell culture plate temperature measurement model reaches the preset accuracy requirement.

[0091] Specifically, step S150 can first calculate the root mean square error (RMSE) of the chamber temperature prediction data and the corresponding actual chamber temperature data. Based on the calculated RMSE, the parameters of the trained cell culture plate temperature measurement model are adjusted until the RMSE of the chamber temperature prediction data and the corresponding actual chamber temperature data obtained by the cell culture plate temperature measurement model reaches the preset accuracy requirement, thus obtaining the final cell culture plate temperature measurement model.

[0092] The cell culture plate temperature measurement model training method provided in this disclosure, compared with the prior art, uses a cell culture plate temperature measurement model to model the heat transfer process from the inner cavity of the cell culture plate to the surroundings. This allows the trained cell culture plate temperature measurement model to be applied to non-invasive and accurate measurement of the cavity temperature of the cell culture plate, thereby effectively avoiding contamination of the cell culture plate cavity during the temperature measurement process, improving the accuracy of temperature measurement, and reducing costs.

[0093] Another embodiment of this disclosure relates to a deployment method for a cell culture plate temperature measurement model, comprising: obtaining a cell culture plate temperature measurement model trained using the training method described in the above embodiments; quantizing the trained cell culture plate temperature measurement model by converting it from a floating-point model to a fixed-point integer model to obtain a quantized cell culture plate temperature measurement model; optimizing the quantized cell culture plate temperature measurement model by adjusting its memory usage and computing resource allocation to obtain an optimized cell culture plate temperature measurement model; benchmarking the optimized cell culture plate temperature measurement model to evaluate its performance; converting the optimized cell culture plate temperature measurement model into target code; and deploying the target code to a chip.

[0094] Specifically, such as Figure 4 As shown, the deployment method of the cell culture plate temperature measurement model can be represented as steps S6 to S11.

[0095] Step S6 involves acquiring the trained cell culture plate temperature measurement model and converting it to ONNX (Open Neural Network Exchange) format to improve the efficiency and flexibility of model deployment. ONNX is an open file format for representing machine learning models, designed to provide a unified standard for model conversion between different deep learning frameworks. Through the ONNX format, developers can seamlessly migrate models between different frameworks, thereby improving development efficiency and flexibility.

[0096] Step S7 is used for model quantization, converting the cell culture plate temperature measurement model from a floating-point model (e.g., a 32-bit floating-point model) to a fixed-point integer model (e.g., an 8-bit integer model), resulting in a quantized cell culture plate temperature measurement model. Quantization significantly reduces the data size of the cell culture plate temperature measurement model, thereby improving the efficiency of the central processing unit (CPU) / microcontroller unit (MCU) and inference speed.

[0097] Step S8 is used for model optimization, adjusting the memory usage and computational resource allocation of the quantized cell culture plate temperature measurement model to achieve optimal model performance.

[0098] Step S9 is used for benchmarking to evaluate the performance of the optimized cell culture plate temperature measurement model, including inference time, memory usage, and accuracy. By benchmarking the optimized cell culture plate temperature model, it can be ensured that the target code generated based on this model performs consistent with the performance of the originally trained cell culture plate temperature measurement model on the deployed chip, and the time for the model to perform one computation can be obtained.

[0099] Step S10 is used to generate code. For example, step S10 can use the tool STM32Cube.AI to convert the optimized cell culture plate temperature measurement model into C code suitable for embedded processors such as STM32 microcontrollers, so that artificial intelligence (AI) inference can be run on the embedded processor using this C code. Of course, step S10 can also convert the optimized cell culture plate temperature measurement model into corresponding target code suitable for other devices such as other types of processors, so that the optimized cell culture plate temperature measurement model can be applied to other devices.

[0100] Step S11 involves deploying the target code corresponding to the cell culture plate temperature measurement model to the chip, enabling real-time monitoring and analysis of the cell culture plate temperature in the corresponding culture scenario. For example, step S11 can incorporate the target code into the culture box code, allowing real-time monitoring and analysis of the cell culture plate temperature within the culture box during the culture process.

[0101] In particular, when the chip is an STM32 series microcontroller, the deployment method of the above cell culture plate temperature measurement model is preferably implemented using the official STM32 tools to improve deployment efficiency and reduce deployment difficulty.

[0102] The method for deploying a cell culture plate temperature measurement model provided in this disclosure, compared with the prior art, enables independent and rapid non-invasive monitoring of the chamber temperature of the cell culture plate using a chip with the deployed cell culture plate temperature measurement model, while improving accuracy and reducing cost.

[0103] Another embodiment of this disclosure relates to a method for measuring the temperature of a cell culture plate, comprising: acquiring actual temperature data corresponding to the four outer walls of the cell culture plate; normalizing the actual temperature data to obtain actual temperature time series data; inputting the actual temperature time series data into a trained cell culture plate temperature measurement model to obtain the corresponding model output result; wherein, the trained cell culture plate temperature measurement model is trained using the training method for cell culture plate temperature measurement models described in the above embodiment; and performing inverse normalization on the model output result to obtain the chamber measurement temperature of the cell culture plate.

[0104] Specifically, such as Figure 5 As shown, the cell culture plate temperature measurement method can be represented as steps S12 to S23.

[0105] Step S12 is used to initialize and create a neural network. The network structure and parameters of the trained cell culture plate temperature measurement model are loaded into memory. The required memory is allocated to each part of the network structure, including weights, biases, activation buffers, etc., and the parameters and states of the network structure are configured to ensure that the network structure is correctly connected and ready.

[0106] Step S13 is used to acquire the actual temperature data corresponding to the four outer walls of the cell culture plate through temperature acquisition. Specifically, as shown... Figure 2 As shown, step S13 can use temperature sensors 2 to 5 set on the outer walls of the cell well plate 1 to collect the actual temperature data corresponding to the outer walls of the cell well plate 1, so as to achieve non-invasive temperature measurement, avoid contamination of the cell culture process, and improve the cleanliness of cell culture.

[0107] Step S14 is used for data normalization processing, which normalizes the actual temperature data.

[0108] Step S15 is used to complete the temperature sequence input, inputting the actual temperature time series data corresponding to the normalized actual temperature data into the ready cell culture plate temperature measurement model.

[0109] Steps S16 to S19 are all implemented by the sequence output layer in the cell culture plate temperature measurement model. Specifically, step S16 is used to receive actual temperature time series data as input and output a 1-dimensional 64-bit signed tensor containing the shape of the input tensor. This step is used to obtain the dimensional information of the actual temperature time series data.

[0110] Step S17 receives a data tensor containing dimensional information of the actual temperature time series data and an index tensor that has been initialized through an index collection operation. It collects entries along the specified axes of the data tensor by index and concatenates them into the output tensor. This step is used to extract data from the data tensor by the specified index.

[0111] Step S18 connects a list of tensors into a single tensor using a tensor join operation.

[0112] Step S19 receives a tensor as input via a tensor tiling operation and outputs a new tensor generated by repeating the operation a specified number of times along each dimension of the input tensor to the long short-term memory network layer in the cell culture plate temperature measurement model. This step is used to construct a new tensor generated by tiling a given tensor.

[0113] Step S20 involves computation using a long short-term memory network in the cell culture plate temperature measurement model. The long short-term memory network receives actual temperature time-series data, captures and retains long-term dependencies, and outputs processed sequence data.

[0114] Step S21 is used for result processing. The results output by the long short-term memory network are tensor compressed and processed through the fully connected layer in the cell culture plate temperature measurement model to obtain the calculation results of the cell culture plate temperature measurement model.

[0115] Step S22 is used to output the calculation results of the cell culture plate temperature measurement model.

[0116] Step S23 is used to inversely normalize the calculation results of the cell culture plate temperature measurement model to obtain the final result, which is the chamber temperature measurement of the cell culture plate. This is combined with... Figure 2 When step S13 uses temperature sensors 2 to 5 located on the outer walls of the cell well plate 1 to collect the actual temperature data corresponding to the outer walls of the cell well plate 1, the chamber measurement temperature obtained in step S23 includes the temperatures corresponding to the five regions 11, 12, 13, 14, and 15 in the cell well plate 1, thereby realizing non-invasive measurement of the cell well plate chamber temperature and avoiding contamination of the cell well plate culture process during the temperature measurement process.

[0117] For example, the cell culture plate temperature measurement method further includes: controlling the environment of the incubator in which the cell culture plate is located based on the measured temperature of the chamber.

[0118] For example, if the measured temperature of a chamber is lower than the preset chamber temperature, the incubator containing the cell culture plate can be heated to increase the temperature of each chamber in the cell culture plate. Conversely, if the measured temperature of a chamber is higher than the preset chamber temperature, the incubator containing the cell culture plate can be cooled to decrease the temperature of each chamber in the cell culture plate.

[0119] By controlling the environment of the incubator where the cell culture plate is located based on the measured temperature in the chamber, the chamber temperature of the cell culture plate can be precisely controlled during the culture process, ensuring that the chamber temperature meets the culture requirements.

[0120] For example, the cell culture plate temperature measurement method further includes: visualizing the measured temperature of the chamber.

[0121] Specifically, visualization can be achieved using a corresponding display screen or visual operation screen, so that users can view the measured temperature of the chamber more intuitively.

[0122] The cell culture plate temperature measurement method provided in this disclosure addresses the high cleanliness requirements of cell culture by designing a non-invasive method. It utilizes a cell culture plate temperature measurement model to model the heat transfer process from the cell culture plate's internal chamber to its surroundings, and employs deep learning technology to process the temperature time series data of the cell culture plate, effectively achieving accurate measurement of the cell culture plate's chamber temperature.

[0123] Another embodiment of this disclosure relates to a training system for a cell culture plate temperature measurement model, such as... Figure 6 As shown, it includes an acquisition module 610, a normalization module 620, an iteration module 630, an anti-normalization module 640, and an adjustment module 650.

[0124] The acquisition module 610 is used to acquire the temperature time series data corresponding to the four outer walls of the cell culture plate and the actual temperature data of the chamber at a specified location of the cell culture plate; wherein, the specified location of the cell culture plate includes the location of different regions of the cell culture plate.

[0125] The normalization module 620 is used to normalize temperature time series data to obtain normalized temperature time series data.

[0126] The iteration module 630 is used to input normalized temperature time series data into the cell culture plate temperature measurement model for iterative training, and obtain the chamber temperature sequence prediction data corresponding to the normalized temperature time series data.

[0127] The inverse normalization module 640 is used to perform inverse normalization processing on the chamber temperature sequence prediction data to obtain the chamber temperature prediction data.

[0128] The adjustment module 650 is used to adjust the parameters of the cell culture plate temperature measurement model based on the predicted chamber temperature data and the corresponding actual chamber temperature data, until the error of the cell culture plate temperature measurement model reaches the preset accuracy requirement.

[0129] For a detailed implementation of the training system for the cell culture plate temperature measurement model provided in this disclosure, please refer to the training method for the cell culture plate temperature measurement model provided in this disclosure, which will not be repeated here.

[0130] The cell culture plate temperature measurement model training system provided in this disclosure, compared with the prior art, uses a cell culture plate temperature measurement model to model the heat transfer process from the inner cavity of the cell culture plate to the surroundings. This allows the trained cell culture plate temperature measurement model to be applied to non-invasive and accurate measurement of the cavity temperature of the cell culture plate, thereby effectively avoiding contamination of the cell culture plate cavity during the temperature measurement process, improving the accuracy of temperature measurement, and reducing costs.

[0131] Another embodiment of this disclosure relates to a deployment system for a cell culture plate temperature measurement model, such as Figure 7 As shown, it includes a model acquisition module 710 and a deployment module 720.

[0132] The model acquisition module 710 is used to acquire the cell culture plate temperature measurement model trained using the training method of the cell culture plate temperature measurement model described in the above embodiments.

[0133] The deployment module 720 is used to quantize the trained cell culture plate temperature measurement model, converting it from a floating-point model to a fixed-point integer model to obtain a quantized cell culture plate temperature measurement model; optimize the quantized cell culture plate temperature measurement model by adjusting its memory usage and computing resource allocation to obtain an optimized cell culture plate temperature measurement model; benchmark the optimized cell culture plate temperature measurement model to evaluate its performance; convert the optimized cell culture plate temperature measurement model into target code; and deploy the target code to the chip.

[0134] For a detailed implementation method of the deployment system of the cell culture plate temperature measurement model provided in this disclosure, please refer to the deployment method of the cell culture plate temperature measurement model provided in this disclosure, which will not be repeated here.

[0135] The cell culture plate temperature measurement model deployment system provided in this disclosure, compared with the prior art, enables independent and rapid non-invasive monitoring of the chamber temperature of the cell culture plate using a chip with the cell culture plate temperature measurement model deployed thereon, while improving accuracy and reducing cost.

[0136] Another embodiment of this disclosure relates to a cell culture plate temperature measurement system, such as Figure 8 As shown, it includes a temperature acquisition module 810, a normalization processing module 820, a prediction module 830, and an inverse normalization processing module 840.

[0137] The temperature acquisition module 810 is used to acquire the actual temperature data corresponding to the four outer walls of the cell culture plate.

[0138] The normalization processing module 820 is used to normalize the actual temperature data to obtain the actual temperature time series data.

[0139] The prediction module 830 is used to input the actual temperature time series data into the trained cell culture plate temperature measurement model to obtain the corresponding model prediction results; wherein, the trained cell culture plate temperature measurement model is trained using the training method of the cell culture plate temperature measurement model described in the above embodiment.

[0140] The inverse normalization module 840 is used to inverse normalize the model prediction results to obtain the chamber measurement temperature of the cell culture plate.

[0141] For example, the cell culture plate temperature measurement system also includes an environmental control module. The environmental control module is used to control the environment of the incubator containing the cell culture plate based on the measured temperature in the chamber.

[0142] As an example, the cell culture plate temperature measurement system also includes a visualization module. The visualization module is used to visually display the measured temperature in the chamber.

[0143] The specific implementation method of the cell culture plate temperature measurement system provided in this disclosure can be found in the cell culture plate temperature measurement method provided in this disclosure, and will not be repeated here.

[0144] The cell culture plate temperature measurement system provided in this disclosure addresses the high cleanliness requirements of cell culture by designing a non-invasive method for measuring cell culture plate temperature. It utilizes a cell culture plate temperature measurement model to model the heat transfer process from the inner chamber of the cell culture plate to its surroundings, and employs deep learning technology to process the temperature time series data of the cell culture plate, effectively achieving accurate measurement of the chamber temperature of the cell culture plate.

[0145] For example, such as Figure 9As shown, a cell culture plate temperature measurement system includes a temperature acquisition module, a main control board, and a visual operation screen. The temperature acquisition module uses temperature sensors installed on the outer walls of the cell culture plate (e.g., a well plate) in the incubator to acquire four temperature data points, representing the actual temperatures of the four outer walls of the cell culture plate. This data is then transmitted to the main control board via RS485. The main control board is an STM32F407 microcontroller, employing a multi-tasking system. One task is used for neural network computation, specifically normalizing the actual temperature data to obtain a time-series data. This time-series data is then input into a pre-trained cell culture plate temperature measurement model incorporating a long short-term memory network to obtain the model's prediction results. These prediction results are then de-normalized to obtain five temperature regions: the center and four corners of the well plate, which are used as the chamber measurement temperatures of the cell culture plate. Besides neural network computation tasks, other task systems on the main control board can be used for environmental control of the incubator based on the incubator control system. For example, the temperature inside the incubator can be controlled based on the temperature measured in the chambers of cell culture plates, such as well plates. After obtaining the temperatures of the five zones, the main control board can use a Universal Asynchronous Receiver-Transmitter (UART) to transmit the temperatures of the five zones to a visual operation screen for display.

[0146] right Figure 9 The cell culture plate temperature measurement system shown was tested. Eight sets of data correspond to eight heating processes in a cell culture plate (e.g., a well plate), starting from room temperature and heating to the target temperature for 2 hours, followed by a 2-hour heating stop. The other eight sets of data correspond to eight incubation processes in a cell culture plate (e.g., a well plate), where the cell culture plate (e.g., a well plate) was removed from an incubator at 37°C and placed in a culture box for 2 hours, followed by a 2-hour heating stop. The root mean square error (RMSE) between the measured chamber temperature of the cell culture plate (e.g., a well plate) obtained through the cell culture plate temperature measurement system and the actual chamber temperature obtained by inserting a probe into the corresponding chamber of the cell culture plate (e.g., a well plate) is calculated. The corresponding results are shown in Table 1 below.

[0147] Table 1. Root Mean Square Error (RMSE) for different test conditions

[0148]

[0149] As can be seen from Table 1, the root mean square errors are all within the range of 0.22 to 0.38, indicating that the measured temperature of the cell culture plate, such as the cell well plate, obtained by the cell culture plate temperature measurement system is not much different from the actual temperature of the corresponding chamber in the cell culture plate, such as the cell well plate, and can meet the accuracy requirements.

[0150] Those skilled in the art will understand that the above embodiments are specific implementations of this disclosure, and in practical applications, various changes can be made to them in form and detail without departing from the spirit and scope of this disclosure.

Claims

1. A training method for a cell culture plate temperature measurement model, characterized in that, The training method includes: Acquire time-series temperature data corresponding to the four outer walls of the cell culture plate and actual temperature data of the chambers at designated locations on the cell culture plate; wherein, the designated locations on the cell culture plate include the locations of different regions of the cell culture plate; the different regions of the cell culture plate cover all chambers of the cell culture plate; The temperature time series data is normalized to obtain normalized temperature time series data; The normalized temperature time series data is input into the cell culture plate temperature measurement model for iterative training to obtain the chamber temperature sequence prediction data corresponding to the normalized temperature time series data. The chamber temperature sequence prediction data is inversely normalized to obtain the chamber temperature prediction data. Based on the predicted chamber temperature data and the corresponding actual chamber temperature data, the parameters of the cell culture plate temperature measurement model are adjusted until the error of the cell culture plate temperature measurement model reaches the preset accuracy requirement. The cell culture plate temperature measurement model models the heat transfer process from the inner cavity of the cell culture plate to the surrounding area. The trained cell culture plate temperature measurement model is used to perform non-invasive measurement of the cavity temperature of the cell culture plate. The cell culture plate temperature measurement model includes a sequence input layer, a long short-term memory network layer, and a fully connected layer connected in sequence. The sequence input layer is used to receive the normalized temperature time series data and convert the normalized temperature time series data into corresponding tensor data. The long short-term memory network layer is used to receive the tensor data, process and predict the dependencies in the tensor data, and output the corresponding sequence data; The fully connected layer is used to perform tensor compression and processing on the sequence data to obtain the corresponding chamber temperature sequence prediction data; The Long Short-Term Memory (LSTM) network layer is equipped with LSTM units. These LSTM units, through their memory cells and gating mechanisms, capture and retain information that depends on long-term conditions, and model the process of heat transfer from the cell culture plate chamber to the surrounding area over time.

2. The training method according to claim 1, characterized in that, The normalization process for the temperature time series data to obtain normalized temperature time series data includes: Calculate the mean and standard deviation of the temperature time series data; The temperature time series data is normalized based on the average value and the standard deviation to obtain the normalized temperature time series data.

3. A method for deploying a cell culture plate temperature measurement model, characterized in that, The deployment method includes: Obtain a cell culture plate temperature measurement model trained using the training method described in claim 1 or 2; The trained cell culture plate temperature measurement model is quantized by converting it from a floating-point model to a fixed-point integer model to obtain the quantized cell culture plate temperature measurement model. The quantified cell culture plate temperature measurement model was optimized by adjusting its memory usage and computing resource allocation to obtain an optimized cell culture plate temperature measurement model. Benchmark tests were conducted on the optimized cell culture plate temperature measurement model to evaluate its performance. The optimized cell culture plate temperature measurement model was converted into target code. The target code is deployed to the chip.

4. A method for measuring the temperature of a cell culture plate, characterized in that, The method for measuring the temperature of the cell culture plate includes: Obtain the actual temperature data corresponding to the four outer walls of the cell culture plate; The actual temperature data is normalized to obtain actual temperature time series data; The actual temperature time series data is input into the trained cell culture plate temperature measurement model to obtain the corresponding model output results; wherein, the trained cell culture plate temperature measurement model is trained using the training method of the cell culture plate temperature measurement model described in claim 1 or 2; The model output is inversely normalized to obtain the chamber temperature of the cell culture plate.

5. The method for measuring the temperature of a cell culture plate according to claim 4, characterized in that, The cell culture plate temperature measurement method further includes: Based on the temperature measured in the chamber, the environment of the incubator containing the cell culture plate is controlled; and / or, The measured temperature in the chamber is displayed visually.

6. A training system for a cell culture plate temperature measurement model, characterized in that, The training system includes: The acquisition module is used to acquire temperature time-series data corresponding to the four outer walls of the cell culture plate and actual temperature data of the chambers at specified locations on the cell culture plate; wherein, the specified locations on the cell culture plate include the locations of different regions of the cell culture plate; the different regions of the cell culture plate cover all chambers on the cell culture plate; The normalization module is used to normalize the temperature time series data to obtain normalized temperature time series data. An iterative module is used to input the normalized temperature time series data into the cell culture plate temperature measurement model for iterative training to obtain the chamber temperature sequence prediction data corresponding to the normalized temperature time series data. The inverse normalization module is used to perform inverse normalization processing on the chamber temperature sequence prediction data to obtain chamber temperature prediction data; The adjustment module is used to adjust the parameters of the cell culture plate temperature measurement model based on the predicted chamber temperature data and the corresponding actual chamber temperature data, until the error of the cell culture plate temperature measurement model reaches the preset accuracy requirement. The cell culture plate temperature measurement model models the heat transfer process from the inner cavity of the cell culture plate to the surrounding area. The trained cell culture plate temperature measurement model is used to perform non-invasive measurement of the cavity temperature of the cell culture plate. The cell culture plate temperature measurement model includes a sequence input layer, a long short-term memory network layer, and a fully connected layer connected in sequence. The sequence input layer is used to receive the normalized temperature time series data and convert the normalized temperature time series data into corresponding tensor data. The long short-term memory network layer is used to receive the tensor data, process and predict the dependencies in the tensor data, and output the corresponding sequence data; The fully connected layer is used to perform tensor compression and processing on the sequence data to obtain the corresponding chamber temperature sequence prediction data; The Long Short-Term Memory (LSTM) network layer is equipped with LSTM units. These LSTM units, through their memory cells and gating mechanisms, capture and retain information that depends on long-term conditions, and model the process of heat transfer from the cell culture plate chamber to the surrounding area over time.

7. A deployment system for a cell culture plate temperature measurement model, characterized in that, The deployment system includes: The model acquisition module is used to acquire the cell culture plate temperature measurement model trained using the training method of the cell culture plate temperature measurement model according to claim 1 or 2. The deployment module is used to quantize the trained cell culture plate temperature measurement model, converting it from a floating-point model to a fixed-point integer model to obtain a quantized cell culture plate temperature measurement model; optimize the quantized cell culture plate temperature measurement model by adjusting its memory usage and computing resource allocation to obtain an optimized cell culture plate temperature measurement model; benchmark the optimized cell culture plate temperature measurement model to evaluate its performance; convert the optimized cell culture plate temperature measurement model into target code; and deploy the target code to the chip.

8. A cell culture plate temperature measurement system, characterized in that, The cell culture plate temperature measurement system includes: The temperature acquisition module is used to acquire the actual temperature data corresponding to the four outer walls of the cell culture plate. The normalization processing module is used to normalize the actual temperature data to obtain actual temperature time series data. The prediction module is used to input the actual temperature time series data into the trained cell culture plate temperature measurement model to obtain the corresponding model prediction results; wherein, the trained cell culture plate temperature measurement model is trained using the training method of the cell culture plate temperature measurement model according to claim 1 or 2; The inverse normalization module is used to inverse normalize the model prediction results to obtain the chamber measurement temperature of the cell culture plate.

9. The cell culture plate temperature measurement system according to claim 8, characterized in that, The cell culture plate temperature measurement system also includes: An environmental control module is used to control the environment of the incubator containing the cell culture plate based on the measured temperature of the chamber; and / or, A visualization module is used to visualize the measured temperature of the chamber.

Images