Peripheral blood immune cell efficient collection and low temperature storage method

Abstract

The present application relates to the technical field of biological cell collection and preservation, in particular to a peripheral blood immune cell efficient collection and low-temperature storage method, comprising: collecting peripheral blood samples and obtaining donor health status information, generating a sample cell composition atlas containing lymphocyte subpopulation absolute count percentage, monocyte purity and granulocyte contamination ratio through full-automatic blood cell analysis, determining the target immune cell optimal collection amount and adaptive collection process through a preset data analysis model, enriching target cells through gradient density centrifugation and specific magnetic bead sorting, comparing and judging supplementary collection by combining cell quality inspection data and expected collection amount, and then calculating individualized cryopreservation liquid formula, cooling program and storage strategy through a low-temperature storage optimization algorithm. The method can realize precise individualized adaptation of the collection process and storage parameters, and improve the cell enrichment purity and the cell state stability after low-temperature storage.

Description

Technical Field

[0001] This invention relates to the field of biological cell collection and preservation technology, and in particular to a method for efficient collection and low-temperature storage of peripheral blood immune cells. Background Technology

[0002] In routine peripheral blood immune cell collection and cryopreservation procedures, peripheral blood samples undergo only basic blood cell counts. The collection volume is determined manually based on experience, and cell enrichment is achieved using a fixed gradient density centrifugation and specific magnetic bead sorting process. Cell viability testing is only used as a basic verification step before cryopreservation. The cryopreservation process employs a universal cryopreservation solution formula, standardized cooling procedures, and a uniform long-term storage strategy, without adjusting storage parameters based on the actual total cell collection volume and quality control data. Current technologies do not link donor health status information with cellular composition data such as the absolute percentage of lymphocyte subsets, monocyte purity, and granulocyte contamination ratio. Cell collection volume and process protocols lack model-based calculation support, and cryopreservation parameters lack individualized adaptation logic. Uncontrollable deviations exist in cell enrichment purity, collection volume accuracy, and cell state after cryopreservation. There are no quantitative comparison standards for initiating supplementary collection, and storage parameters cannot be matched with actual cell properties.

[0003] It is necessary to construct a data analysis model that links donor health status information with sample cell composition atlases. The model should be used to calculate the expected optimal collection amount of target immune cells and the corresponding optimal cell collection process. At the same time, based on the final total number of cells collected and the cell quality inspection data set, a low-temperature storage optimization algorithm should be used to calculate the individualized low-temperature preservation solution formula, cooling procedure and long-term storage strategy. This will make up for the lack of individualized and model-based adaptation in the collection and storage process of existing technologies. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and propose a method for efficient collection and low-temperature storage of peripheral blood immune cells.

[0005] To achieve the above objectives, the present invention employs the following technical solution: a method for efficient collection and cryogenic storage of peripheral blood immune cells, comprising:

[0006] Collect peripheral blood samples and obtain donor health status information associated with the peripheral blood samples;

[0007] The peripheral blood sample is subjected to fully automated blood cell analysis to generate a sample cell composition atlas, which includes the absolute percentage of lymphocyte subsets, the purity of monocytes, and the contamination ratio of granulocytes.

[0008] Based on the donor's health status information and the sample cell composition map, the expected optimal collection amount of target immune cells and the corresponding most suitable cell collection process are determined by calculation through a preset data analysis model.

[0009] According to the optimal cell collection process, the peripheral blood sample is subjected to gradient density centrifugation and specific magnetic bead sorting to generate an enriched target immune cell suspension.

[0010] The viability and activity of the enriched target immune cell suspension are detected to generate a cell quality control data set. Based on the comparison between the cell quality control data set and the expected optimal collection amount, it is determined whether to initiate a supplementary collection process.

[0011] Based on the final total number of cells collected and the cell quality inspection data set, a low-temperature storage optimization algorithm is invoked to calculate an individualized low-temperature preservation solution formula, cooling procedure, and long-term storage strategy for the enriched target immune cell suspension.

[0012] As a further aspect of the present invention, the peripheral blood sample is subjected to fully automated blood cell analysis to generate a sample cell composition atlas, including:

[0013] The donor's health status information includes the donor's age, recent medication records, and total white blood cell count in routine blood tests.

[0014] A first subsample for testing is separated from the peripheral blood sample. The first subsample for testing is subjected to multicolor flow cytometry analysis of cell surface markers to identify the absolute number of T cell subsets, B cells, and NK cells, and to calculate the percentage of each subset relative to nucleated cells.

[0015] A second subsample for testing is extracted from the peripheral blood sample. An automated cell morphology analysis based on cell light scattering characteristics is performed on the second subsample to distinguish the morphological differences between the monocytes and lymphocytes, and to calculate the purity of the monocytes.

[0016] Combining the results of multicolor flow cytometry analysis with the results of automated cell morphology analysis, the contamination ratio of the granulocytes is assessed by a preset contamination judgment logic, which makes judgments based on cell size, granularity, and the expression intensity of specific markers.

[0017] The absolute percentage count of the lymphocyte subsets, the purity of the monocytes, and the contamination ratio of the granulocytes are integrated to form a structured atlas of the sample cell composition.

[0018] As a further aspect of the present invention, based on the donor's health status information and the sample cell composition map, a preset data analysis model is used to calculate and determine the expected optimal collection amount of target immune cells and the corresponding most suitable cell collection process, including:

[0019] The donor's age, recent medication records, and total white blood cell count are input into the donor status assessment submodule of the data analysis model to calculate the donor cell collection suitability score.

[0020] The absolute percentage of lymphocyte subsets, the purity of monocytes, and the contamination ratio of granulocytes in the sample cell composition atlas are input into the cell quality prediction submodule of the data analysis model to predict the potential viability and proliferation capacity of the collected cell products.

[0021] By integrating the donor cell collection suitability score with the predicted results of the potential viability and proliferation capacity of the collected cell products, the expected optimal collection amount of the target immune cells is calculated using a target cell yield optimization algorithm, while ensuring cell quality.

[0022] Based on the expected optimal collection volume, the type of target immune cells, and the contamination ratio of granulocytes, the most suitable cell collection process scheme, including gradient density centrifugation parameters, magnetic bead sorting antibody combination, and sorting buffer formulation, is matched from a preset process library.

[0023] As a further aspect of the present invention, the peripheral blood sample is subjected to gradient density centrifugation and specific magnetic bead sorting to generate an enriched suspension of target immune cells, comprising:

[0024] According to the gradient density centrifugation parameters specified in the optimal cell collection process, including centrifugation force and centrifugation time, the peripheral blood sample is subjected to stratified centrifugation to separate a cell layer rich in mononuclear cells.

[0025] The cell layer rich in mononuclear cells was collected and washed and resuspended using the sorting buffer specified in the optimal cell collection process to prepare a pre-sorting cell suspension.

[0026] According to the magnetic bead sorting antibody combination specified in the optimal cell collection process, the specific magnetic beads are incubated with the pre-sorted cell suspension to form a magnetic bead-target cell complex.

[0027] The incubated mixed suspension was placed on a magnetic sorting rack to separate the target cell population bound to magnetic beads, and the unbound cells were removed.

[0028] The isolated target cell population was dissociated and washed to remove magnetic beads, ultimately yielding the enriched target immune cell suspension suspended in the preservation solution.

[0029] As a further aspect of the present invention, the enriched target immune cell suspension is subjected to viability and activity detection to generate a cell quality control data set, including:

[0030] Samples were taken from the enriched target immune cell suspension and subjected to trypan blue rejection assay. The number of live cells and the total number of cells were calculated using a cell counter to obtain the cell viability.

[0031] Flow cytometry analysis was performed on sampled cells to detect the expression ratio of early apoptosis markers and calculate the percentage of early apoptotic cells.

[0032] Staining and analyzing target functional markers on sampled cells to detect the proportion of cell subpopulations with specific functions in living cells;

[0033] The cell survival rate, the percentage of early apoptotic cells, and the proportion of cell subpopulations with specific functions are combined to form the cell quality inspection data set.

[0034] As a further aspect of the present invention, based on the comparison between the cell quality inspection data set and the expected optimal collection amount, determining whether to initiate a supplementary collection process includes:

[0035] The cell viability rate is read from the cell quality inspection data set. If the cell viability rate is lower than the preset viability rate threshold, the current collection is determined to be a failure and a new round of complete sample collection process needs to be performed.

[0036] If the cell survival rate is not lower than the preset survival rate threshold, the total number of live cells actually collected is compared with the expected optimal collection amount.

[0037] If the actual total number of live cells collected is lower than the expected optimal collection amount, but higher than the minimum effective treatment dose, the cell number difference is calculated, and it is determined whether the donor meets the physiological conditions for supplementary collection.

[0038] If the donor meets the physiological conditions for supplementary collection, the volume of peripheral blood to be supplemented is recalculated based on the difference in cell number and the cell composition profile of the sample, and the supplementary collection process is initiated.

[0039] If the total number of live cells actually collected reaches or exceeds the expected optimal collection amount, or if the donor does not meet the conditions for supplementary collection, the collection process ends and enters the low-temperature storage preparation stage.

[0040] As a further aspect of the present invention, the step of invoking a low-temperature storage optimization algorithm to calculate an individualized low-temperature preservation solution formulation, cooling procedure, and long-term storage strategy for the enriched target immune cell suspension includes:

[0041] The final total number of cells collected, along with the cell viability and percentage of early apoptotic cells from the cell quality inspection data set, are input into the low-temperature storage optimization algorithm.

[0042] The low-temperature storage optimization algorithm selects and determines the individualized low-temperature preservation solution formula from multiple candidate formulas based on input parameters and combined with the type characteristics of the target immune cells. The formula includes a specific concentration combination of cryoprotectants, nutrients, antioxidants and buffer systems.

[0043] Meanwhile, the low-temperature storage optimization algorithm calculates the optimal freezing and cooling rate and the rewarming and thawing rate based on the final total number of cells collected and the cell viability, and generates the individualized cooling program, which includes the dwell time of multiple temperature platforms and the cooling gradient.

[0044] Furthermore, based on the percentage of early apoptotic cells, the cryogenic storage optimization algorithm determines the individualized long-term storage strategy, which includes a recommended long-term storage temperature, whether to use a gas-liquid phase mixed storage mode, and a planned cell viability verification time point.

[0045] As a further aspect of the present invention, the execution steps of the individualized cooling program include:

[0046] The enriched target immune cell suspension, which is suspended in the individualized cryopreservation solution, is dispensed into a cryopreservation container.

[0047] Place the low-temperature storage container in the sample chamber of the programmed cooling instrument and start the cooling process;

[0048] According to the individualized cooling program, the programmable cooling device is controlled to first cool from room temperature to a preset first temperature plateau above freezing point at a preset initial cooling rate, and maintain a preset equilibrium time on the first temperature plateau.

[0049] After balancing, the programmed cooling device is controlled to cool from the first temperature platform to the preset transfer temperature at the optimal freezing and cooling rate.

[0050] Once the transfer temperature is reached, the cryogenic storage container is quickly transferred to a long-term cryogenic storage device for long-term preservation.

[0051] As a further aspect of the present invention, it also includes dynamic monitoring and risk assessment steps during cell storage:

[0052] After the enriched target immune cell suspension is placed into long-term low-temperature storage, representative samples are extracted from the long-term storage device at preset intervals for rapid miniaturized viability detection to obtain cell viability data during the storage period.

[0053] The cell viability data during the storage period is compared longitudinally with the cell quality inspection data set at the time of cell entry into the warehouse to analyze the trend of cell viability decline.

[0054] Simultaneously, historical data of the storage device's operating parameters are recorded, including temperature fluctuation range and liquid nitrogen replenishment records.

[0055] The cell viability decline trend and historical data of the storage device operating parameters are input into the storage risk prediction model;

[0056] The storage risk prediction model outputs a report on the remaining estimated effective storage time and risk level assessment of the current batch of cells.

[0057] Based on the aforementioned risk assessment report, a decision will be made as to whether to adjust the long-term storage strategy or to plan the use of cell products in advance.

[0058] As a further aspect of the present invention, the workflow of the storage risk prediction model includes:

[0059] Receive the input cell viability decline trend data, which includes the rate of change in cell survival rate and the proportion of apoptotic cells for each sampling test;

[0060] Receive historical data of the operating parameters of the storage device, especially the number and duration of abnormal temperature events exceeding a preset threshold;

[0061] The cell viability decline trend data and the historical data of the storage device operating parameters are matched and fitted with the model’s built-in database of viability decline curves for different cell types under different storage conditions.

[0062] Based on the fitting results, the cell survival rate and functional activity indicators are predicted at a specific time point in the future, and the time point when the cell survival rate drops to a preset critical value is extrapolated and calculated. The time point is the remaining expected effective storage time.

[0063] Based on the remaining estimated effective storage time, the severity of abnormal temperature events, and the current rate of vitality decay, a risk assessment report with high, medium, and low risk levels is output through risk matrix calculation.

[0064] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0065] By combining donor health status information with a sample cell composition atlas including the absolute percentage of lymphocyte subsets, monocyte purity, and granulocyte contamination ratio, a pre-defined data analysis model is used to calculate and determine the expected optimal collection volume of target immune cells and the corresponding most suitable cell collection process. The initiation of supplementary collection is determined based on the comparison between the cell quality inspection data set and the expected optimal collection volume. The collection volume and cell collection process settings are tailored to the actual composition characteristics of the cell sample; the collection execution logic is adapted to the individual health status of the donor; the operational parameters of gradient density centrifugation and specific magnetic bead sorting are consistent with the proportion characteristics of various cell types in the sample; the control of granulocyte contamination and the regulation of monocyte purity during cell enrichment are consistent with the inherent composition of the sample; the initiation of supplementary collection is based on quantitative comparison results; and the quantity and composition of the target immune cell suspension after enrichment are matched with the pre-defined results calculated by the model.

[0066] Based on the final total cell collection and cell quality control data, a cryopreservation optimization algorithm is invoked to calculate individualized cryopreservation solution formulations, cooling procedures, and long-term storage strategies for the enriched target immune cell suspension. The components and ratios of the cryopreservation solution formulation are adapted to the actual total cell collection and viability test results; the cooling procedure's rate, stages, and other parameters are matched to the cell physiological state corresponding to the cell quality control data; and the parameters of the long-term storage strategy are configured to fit the actual quality and quantity characteristics of the cell suspension. The parameter settings throughout the cryopreservation process directly correspond to the individual attributes of the cell suspension, the storage conditions are consistent with the actual state of the enriched cells, and the parameter adaptability during cell cryopreservation is correlated with the individual characteristics of the samples. Attached Figure Description

[0067] Figure 1 This is a flowchart of the method for efficient collection and cryogenic storage of peripheral blood immune cells described in this invention;

[0068] Figure 2 A flowchart for calculating the optimal data acquisition volume and process scheme for a data analysis model;

[0069] Figure 3 This is a graph showing the trend of cell viability changes across multiple batches.

[0070] Figure 4 Results of early apoptosis rate detection for immune cells;

[0071] Figure 5 Heatmap of the correlation between risk factors and storage. Detailed Implementation

[0072] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0073] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and 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, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0074] See Figure 1 The overall implementation scheme of the method for efficient collection and cryogenic storage of peripheral blood immune cells provided by the present invention is as follows:

[0075] Peripheral blood samples are collected, and donor health status information associated with these samples is obtained. Fully automated hematology analysis is performed on the peripheral blood samples to generate a sample cell composition atlas containing the absolute percentage of lymphocyte subsets, the purity of monocytes, and the proportion of granulocyte contamination. Based on donor health status information and the sample cell composition atlas, automated cell morphology analysis is applicable in scenarios requiring the differentiation of monocytes from lymphocytes within morphologically similar mononuclear cell populations based on cell light scattering characteristics and image features. It also provides a second dimension of judgment based on cell size and granularity for subsequent calculation of the granulocyte contamination ratio. This fully automated hematology analysis serves as the overall framework for initial sample evaluation, integrating subsequent test results and generating a structured sample cell composition atlas. It does not directly generate raw cell count and morphology data but instead constructs the atlas by calling the results of multicolor flow cytometry and automated cell morphology analysis. A preset data analysis model is used to calculate and determine the expected optimal collection volume of target immune cells and the corresponding most suitable cell collection process. According to the optimal cell collection process, peripheral blood samples are subjected to gradient density centrifugation and specific magnetic bead sorting to generate an enriched target immune cell suspension. The viability and activity of the enriched target immune cell suspension are then tested to generate a cell quality control dataset. Based on the comparison between the cell quality control dataset and the expected optimal collection volume, it is determined whether to initiate a supplementary collection process. Based on the final total cell collection volume and the cell quality control dataset, a cryogenic storage optimization algorithm is invoked to calculate an individualized cryopreservation solution formulation, cooling procedure, and long-term storage strategy for the enriched target immune cell suspension.

[0076] In one embodiment of the present invention, the donor's health status information includes the donor's age, recent medication records, and the total white blood cell count in a routine blood test. A first subsample for testing is extracted from the peripheral blood sample. This subsample is then subjected to multicolor flow cytometry analysis of cell surface markers to identify the absolute numbers of T cell subsets, B cells, and NK cells, and to calculate the percentage of each subset relative to nucleated cells. Multicolor flow cytometry analysis is applicable in scenarios requiring precise classification and absolute counting of different lymphocyte subsets through specific fluorescent labeling of cell surface markers. It also provides a first dimension of judgment based on marker expression intensity for subsequent calculation of granulocyte contamination ratio. A second subsample for testing is extracted from the peripheral blood sample. This second subsample undergoes automated cell morphology analysis based on cell light scattering characteristics to distinguish the morphological differences between monocytes and lymphocytes, and to calculate the purity of monocytes. By combining multicolor flow cytometry analysis results with automated cell morphology analysis results, the contamination ratio of granulocytes is assessed through a pre-defined contamination judgment logic. This logic determines contamination based on cell size, granularity, and the expression intensity of specific markers. The absolute percentage of lymphocyte subsets, the purity of monocytes, and the contamination ratio of granulocytes are integrated to form a structured sample cell composition atlas.

[0077] In practice, donor health information includes donor age, recent medication records, and total white blood cell count in routine blood tests. A first subsample for testing is extracted from the peripheral blood sample and subjected to multicolor flow cytometry analysis of cell surface markers to identify the absolute numbers of T cell subsets, B cells, and NK cells, and to calculate the percentage of each subset relative to nucleated cells. The antibody combination used in the multicolor flow cytometry analysis includes fluorescently labeled antibodies against surface markers such as CD3, CD4, CD8, CD19, and CD56. The volume of the first subsample for testing is typically 50 to 100 microliters. The percentage of each subset relative to nucleated cells is calculated by dividing the absolute number of a specific lymphocyte subset by the total absolute number of nucleated cells.

[0078] In some embodiments, a second subsample for testing is extracted from a peripheral blood sample. An automated cell morphology analysis based on cell light scattering characteristics is performed on the second subsample to distinguish morphological differences between monocytes and lymphocytes, and the purity of monocytes is calculated. The automated cell morphology analysis is performed on a platform integrating a digital microscope and an image analysis system. The system classifies cells based on cell size and forward and side scattering signal characteristics. The purity of monocytes is equal to the number of cells identified as monocytes divided by the total number of cells identified as mononuclear cells.

[0079] In practice, the contamination ratio of granulocytes is assessed by combining the results of multicolor flow cytometry and automated cell morphology analysis using a pre-defined contamination judgment logic. This logic is based on cell size, granularity, and the expression intensity of specific markers. The pre-defined contamination judgment rules include: in multicolor flow cytometry analysis, events with high CD15 or CD16 expression and large cell size are counted as granulocytes; in automated cell morphology analysis, events with high granularity scattering characteristics and lobed nuclei are counted as granulocytes. The contamination ratio of granulocytes is calculated by weighting the results of the two analysis techniques, using the following formula:

[0080]

[0081] in: Indicates the proportion of granulocyte contamination. This represents the granulocyte count as analyzed by flow cytometry. This represents the total count of nucleated cells analyzed by flow cytometry. Granulocyte count representing morphological analysis Total cell count representing morphological analysis. and These are preset weighting coefficients, and they satisfy... This is understandable; the weighting coefficients... and The values ​​are set based on the comparison results of instrument calibration data and historical data.

[0082] The absolute percentage counts of lymphocyte subsets, monocyte purity, and granulocyte contamination ratio are integrated to form a structured sample cell composition atlas. The sample cell composition atlas is stored as a data file containing donor identifiers, detection timestamps, and percentage values ​​for each lymphocyte subset, monocyte purity, and granulocyte contamination ratio. In some embodiments, the data file is in JSON or XML format. Optionally, the sample cell composition atlas also includes storage path indexes for the raw fluorescence intensity data files from multicolor flow cytometry analysis and cell image snapshot files from automated cell morphology analysis.

[0083] In one embodiment of the present invention, see [reference] Figure 2The donor's age, recent medication records, and total white blood cell count are input into the donor status assessment submodule of the data analysis model to calculate the donor cell collection suitability score. The absolute percentage of lymphocyte subsets, monocyte purity, and granulocyte contamination ratio in the sample cell composition atlas are input into the cell quality prediction submodule of the data analysis model to predict the potential viability and proliferation capacity of the collected cell product. By integrating the donor cell collection suitability score with the predicted potential viability and proliferation capacity of the collected cell product, a target cell yield optimization algorithm is used to calculate the expected optimal collection volume of target immune cells while ensuring cell quality. Based on the expected optimal collection volume, the type of target immune cells, and the granulocyte contamination ratio, the most suitable cell collection process is matched from a pre-set process library, including gradient density centrifugation parameters, magnetic bead sorting antibody combinations, and sorting buffer formulations. Following the gradient density centrifugation parameters specified in the most suitable cell collection process, including centrifugation force and time, the peripheral blood sample is subjected to stratified centrifugation to separate a cell layer rich in mononuclear cells. A cell layer rich in mononuclear cells was collected and washed and resuspended using the sorting buffer specified in the optimal cell collection protocol to prepare a pre-sorting cell suspension. Following the magnetic bead sorting antibody combination specified in the optimal cell collection protocol, specific magnetic beads were incubated with the pre-sorting cell suspension to form a magnetic bead-target cell complex. The incubated mixed suspension was placed on a magnetic sorting rack to separate the target cell population bound to the magnetic beads, and unbound cells were removed. The separated target cell population was then dissociated and washed to remove the magnetic beads, finally yielding a concentrated target immune cell suspension suspended in a preservation solution.

[0084] In practice, the donor's age, recent medication records, and total white blood cell count (from a routine blood test) are input into the donor status assessment submodule of the data analysis model to calculate a donor cell collection suitability score. The donor status assessment submodule calculates this score based on a multivariate function, which takes the donor's age, medication record quantification, and total white blood cell count as input variables. The medication record quantification is assigned values ​​based on specific items such as recent use of immunosuppressants or chemotherapy drugs. In some embodiments, the formula for calculating the donor cell collection suitability score is as follows:

[0085]

[0086] in: The donor cell collection suitability score represents the appropriateness score. Represents the donor's age. Represents the quantitative value of medication records. Represents the total number of white blood cells, function , and These are standardized mapping functions for age, medication history, and total white blood cell count, respectively, with coefficients... , and The preset weighting factors satisfy the following conditions: .

[0087] The donor status assessment submodule employs a three-layer feedforward neural network structure. The first layer is the input layer, containing three input nodes that receive the donor's age A, the quantified value of medication records M, and the total white blood cell count W, respectively. The second layer is the hidden layer, containing five neurons. Each neuron performs a weighted summation of the input vectors and then a non-linear transformation using a sigmoid activation function. The mathematical form of the sigmoid activation function is:

[0088] ,

[0089] in Represented by natural constant The function is an exponential function with base $\frac{\pi}{\mathbf{a}$, where $z$ is the weighted sum of inputs to the neuron. The hidden layer is... The output of each neuron The calculation method is as follows:

[0090]

[0091] in Indicates the donor's age , Indicates the quantitative value of medication records , Indicates the total number of white blood cells , Indicates the input layer's first... The node is connected to the hidden layer. The connection weight of each node Indicates the hidden layer number 1 The bias term of each node, This represents the summation operation. This represents a multiplication operation. The third layer is the output layer, containing one output node, which linearly weights and sums the outputs of all neurons in the hidden layers to obtain the donor cell collection suitability score. :

[0092]

[0093] in Indicates the hidden layer number 1 The connection weights from each node to the output layer This represents the output layer bias term. The connection weights of the donor state evaluation submodule. , and bias terms , The training dataset was determined through supervised learning on a labeled dataset. The dataset consisted of donor health information collected from multiple centers and corresponding records of actual collection results, containing at least 500 donor samples. Each record was labeled with a suitability level assessed by experienced technicians based on the collection results. The levels were divided into five grades and quantified as continuous values ​​between 0 and 1. The training employed a backpropagation algorithm with mean squared error as the loss function, a learning rate of 0.001, and 2000 training iterations. Training was terminated early when the validation set loss no longer decreased after 50 consecutive iterations. (The text then mentions quantified values ​​for medication records.) The assignment rule is as follows: if the donor has not used immunosuppressants, glucocorticoids, or chemotherapy drugs within the past thirty days, then... The value is assigned to 0; if any of the above-mentioned drugs have been used but have been discontinued for more than seven days, then... Assign a value of 1; if any of the above-mentioned drugs are currently in use, then The value is assigned to 2. Total white blood cell count. Standardized mapping function The raw white blood cell count values ​​are mapped to the 0-1 range, as follows: ,in Values ,when Exceed hour The value is 1. Donor age Standardized mapping function for: ,when hour Take 1, when hour Take 0.

[0094] In practice, the absolute percentage of lymphocyte subsets, the purity of monocytes, and the proportion of granulocyte contamination in the sample cell composition atlas are input into the cell quality prediction submodule of the data analysis model to predict the potential viability and proliferation capacity of the collected cell products. The cell quality prediction submodule has a built-in regression prediction model that correlates cell composition with the final product quality. The model inputs include the percentage of CD4+ T cells, the percentage of CD8+ T cells, the purity of monocytes, and the proportion of granulocyte contamination. The regression prediction model uses a multiple linear regression structure. The model input vector contains four features: the proportion of CD4+ T cells is denoted as... The proportion of CD8+ T cells is recorded as Mononuclear cell purity is denoted as The proportion of granulocyte contamination is recorded as The model output is a cell quality prediction score. A comprehensive evaluation value characterizing the potential viability and proliferation capacity of cell products after collection. The value range is from 0 to 1. The calculation formula is:

[0095]

[0096] in, , , , For regression coefficients, For the intercept term, Indicates multiplication operation. This represents an addition operation. The regression coefficients and intercept term were determined by fitting the data to the training dataset using the least squares method. The training dataset consisted of at least 300 batches of historical cell collection and quality control records. Each batch included the four input features mentioned above, as well as the measured cell viability and fold increase after cryopreservation and thawing. During fitting, a weighted combination of cell viability and fold increase after cryopreservation and thawing was used as the regression target value. The weights for the weighted combination were 0.6 for viability and 0.4 for fold increase, and the combination result was normalized to the interval between 0 and 1. Regression coefficients , , , Intercept term The above values ​​are default parameter values ​​obtained by fitting the training dataset, and can be updated and calibrated regularly on a quarterly basis as new data accumulates.

[0097] By integrating donor cell collection suitability scores with predicted potential viability and proliferation capacity of post-collection cell products, a target cell yield optimization algorithm is used to calculate the expected optimal collection quantity of target immune cells while ensuring cell quality. The algorithm optimizes the collection quantity by ensuring the predicted cell viability does not fall below a certain threshold, with the constraint of maximizing the predicted cell viability. The algorithm is executed through constrained optimization calculations. The optimization objective is to maximize the collection quantity, and the constraint is to ensure the donor cell collection suitability score is met. With cell quality prediction score The fusion score is not lower than the preset quality threshold. Fusion score The calculation method is as follows:

[0098]

[0099] in To integrate the weighting coefficients, , Indicates a multiplication operation. Preset quality threshold. Expected optimal collection volume The calculation formula is:

[0100]

[0101] in, Reference values ​​for basic data collection. One cell; Assess the suitability of donor cell collection; Assess cell quality prediction scores; This is the quality constraint threshold; For safety margin coefficient, The fraction bar indicates a division operation. This represents a multiplication operation. When... At that time, the conservative data collection strategy is triggered. Pick 0.5 times.

[0102] Based on the expected optimal collection volume, the type of target immune cells, and the proportion of granulocyte contamination, the most suitable cell collection process is matched from a pre-defined process library. The most suitable cell collection process includes gradient density centrifugation parameters, magnetic bead sorting antibody combinations, and sorting buffer formulations. It can be understood that when the target immune cells are CD4+ T cells and the proportion of granulocyte contamination is high, the process library will match a higher centrifugation force and a magnetic bead sorting antibody combination containing CD4 antibodies.

[0103] In some embodiments, peripheral blood samples are subjected to stratified centrifugation according to the gradient density centrifugation parameters specified in the optimal cell collection process, including centrifugation force and centrifugation time, to separate a cell layer rich in mononuclear cells. Gradient density centrifugation uses Ficoll isodense gradient medium, with centrifugation force ranging from 400×g to 800×g and centrifugation time ranging from 20 minutes to 30 minutes. The gradient medium used for gradient density centrifugation is Ficoll-Paque PLUS with a density of 1.077 g / mL. When the target immune cells are CD4+ T cells, the matched centrifugation force is... The centrifugation time was 25 minutes. When the target immune cells were CD8+ T cells, the appropriate centrifugal force was... The centrifugation time was 25 minutes. When the target immune cells were NK cells, the matching centrifugal force was... The centrifugation time was 20 minutes. When the target immune cells were B cells, the appropriate centrifugal force was... Centrifugation time was 30 minutes. Centrifugation temperature was uniformly controlled at 20°C, with the centrifuge's acceleration setting set to level 5 and the deceleration setting set to natural deceleration without braking. If the granulocyte contamination ratio... If the value exceeds 0.10, then increase the centrifugal force based on the above-mentioned matching force. After centrifugation, use a pipette to aspirate the white film layer between the plasma layer and the Ficoll medium layer. Move the pipette tip slowly along the tube wall, and control the aspirated volume to be between 2 mL and 3 mL.

[0104] A cell layer rich in mononuclear cells is collected and washed and resuspended using the sorting buffer specified in the optimal cell collection protocol to prepare a pre-sorting cell suspension. The sorting buffer is typically a phosphate buffer containing proteins (such as bovine serum albumin) and ethylenediaminetetraacetic acid (EDTA). Following the magnetic bead sorting antibody combination specified in the optimal cell collection protocol, specific magnetic beads are incubated with the pre-sorting cell suspension to form a magnetic bead-target cell complex. Optionally, the magnetic bead sorting antibody combination is a conjugate of surface antibodies against CD3, CD4, CD8, CD19, or CD56 with superparamagnetic microspheres. The detailed configuration of the magnetic bead sorting antibody combination is as follows: When the target immune cells are CD4+ T cells, superparamagnetic magnetic beads conjugated with anti-CD4 antibodies are used. The magnetic bead matrix consists of iron oxide nanoparticles encapsulated with a dextran shell, and the bead diameter ranges from 50 nm to 100 nm. The antibody-to-magnetic bead conjugate ratio is 10 μg antibody per milligram of magnetic beads. When the target immune cells are CD8+ T cells, magnetic beads conjugated with anti-CD8 antibodies are used. When the target immune cells are NK cells, magnetic beads conjugated with anti-CD56 antibodies are used. When the target immune cells are B cells, magnetic beads conjugated with anti-CD19 antibodies are used. The incubation conditions are as follows: magnetic beads are incubated with the pre-sorted cell suspension at 2°C to 8°C for 20 minutes. The binding ratio of magnetic beads to target cells is [missing information - likely a specific ratio]. Each target cell was suspended in 20 μL of magnetic beads. During incubation, the cells were gently inverted and mixed every 5 minutes. The incubation container was continuously rotated on a rotary mixer at 10 rpm. The dissociation method was as follows: The target cell population bound to the magnetic beads was resuspended in dissociation buffer (phosphate buffer containing 2 mmol / L EDTA, pH 7.2). The cells were incubated at room temperature for 10 minutes, gently pipetting and mixing five times every 3 minutes. The suspension was then placed on a magnetic sorting rack to re-adsorb the magnetic beads for 2 minutes. The dissociated target cells in the supernatant were collected and washed twice with sorting buffer, with a centrifugal force of [insert value here] for each wash. Centrifugation time is 5 minutes.

[0105] The incubated mixed suspension was placed on a magnetic sorting rack to separate the target cell population bound to magnetic beads, and unbound cells were removed. The separated target cell population was then dissociated and washed to remove the magnetic beads, finally yielding an enriched target immune cell suspension suspended in a preservation solution.

[0106] In one embodiment of the present invention, samples are taken from the enriched target immune cell suspension and subjected to trypan blue rejection assay. The number of viable cells and the total number of cells are calculated using a cell counter to obtain the cell viability. Flow cytometry analysis is performed on the sampled cells to detect the expression ratio of early apoptosis markers and calculate the percentage of early apoptotic cells. The sampled cells are then stained and analyzed for target functional markers to detect the proportion of cell subpopulations with specific functions among the viable cells. The cell viability, the percentage of early apoptotic cells, and the proportion of cell subpopulations with specific functions are combined to form a cell quality control dataset. The cell viability is read from the cell quality control dataset. If the cell viability is lower than a preset viability threshold, the current collection is considered a failure, and a new round of complete sample collection is required. If the cell viability is not lower than the preset viability threshold, the total number of viable cells actually collected is compared with the expected optimal collection amount. If the total number of viable cells actually collected is lower than the expected optimal collection amount but higher than the minimum effective therapeutic dose, the cell number difference is calculated, and it is determined whether the donor meets the physiological conditions for supplementary collection. If the donor meets the physiological conditions for supplementary collection, the volume of peripheral blood to be supplemented is recalculated based on the difference in cell count and the sample cell composition atlas, and the supplementary collection process is initiated. During this integration process, the fully automated blood cell analysis system, acting as a unified data aggregation and atlas generation platform, integrates lymphocyte subset percentage data and granulocyte marker expression data from multicolor flow cytometry analysis, as well as monocyte purity data and granulocyte morphology characteristic data from automated cell morphology analysis. Based on a preset contamination judgment logic, it performs a weighted calculation of the granulocyte contamination ratio, ultimately outputting a structured sample cell composition atlas containing all detection indicators. If the actual total number of live cells collected reaches or exceeds the expected optimal collection volume, or if the donor does not meet the supplementary collection conditions, the collection process ends, and the process enters the cryogenic storage preparation stage.

[0107] In practice, the cell quality inspection and supplementary collection process involves sampling from the enriched target immune cell suspension, performing a trypan blue rejection assay, and calculating the number of viable cells and the total number of cells using a cell counter to obtain the cell viability. Flow cytometry analysis is then performed on the sampled cells to detect the expression ratio of early apoptosis markers and calculate the percentage of early apoptotic cells. This percentage is determined by the proportion of Annexin V-positive and PI-negative cells in the total cell count. The sampled cells are then stained and analyzed for target functional markers, including cytokines or activation markers, to detect the proportion of cell subsets with specific functions within the viable cells. The cell viability, percentage of early apoptotic cells, and proportion of cell subsets with specific functions are combined to form a cell quality inspection dataset. An example of a cell quality inspection dataset is shown in Table 1.

[0108] Table 1: Example of a Cell Quality Inspection Data Set

[0109] project numerical values unit Cell viability 95.5 % Percentage of cells in early apoptosis 3.2 % The proportion of CD8+ T cells that secrete IFN-γ 15.8 % Proportion of CD4+ T cells expressing CD25 8.4 %

[0110] In practice, cell viability is read from the cell quality control dataset. If the cell viability is lower than a preset viability threshold, the current collection is considered a failure, and a new round of complete sample collection is required. This preset viability threshold is typically set between 80% and 90%. Specifically, the preset viability threshold is set to 0.85. If the cell viability read from the cell quality control dataset is lower than 0.85, the current collection is considered a failure. The minimum effective treatment dose is set according to the target immune cell type: the minimum effective treatment dose for CD4+ T cells is... The minimum effective therapeutic dose of CD8+ T cells per live cell is [number]. The minimum effective therapeutic dose of NK cells is [number] live cells. The minimum effective therapeutic dose for B cells is [number] live cells. One living cell.

[0111] If the cell viability rate is not lower than the preset viability threshold, the total number of live cells actually collected is compared with the expected optimal collection amount. The total number of live cells actually collected is calculated by multiplying the cell concentration measured by a cell counter by the total volume of the enriched target immune cell suspension. If the total number of live cells actually collected is lower than the expected optimal collection amount but higher than the minimum effective therapeutic dose, the cell number difference is calculated, and it is determined whether the donor meets the physiological conditions for supplementary collection. The expected optimal collection amount is calculated by taking the expected optimal collection amount of target immune cells output by the data analysis model. This value is used as the basis for calculating the total number of live cells actually collected.

[0112]

[0113] in This represents the actual total number of living cells. Cell concentration as measured by a cell counter. This represents the total volume of the enriched target immune cell suspension. This refers to the cell viability rate, expressed as a decimal, recorded in the cell quality inspection data set. The formula for calculating the difference in cell number is:

[0114]

[0115] The minimum effective therapeutic dose is a preset constant. If satisfied If the donor meets the physiological conditions for supplementary collection, then the decision to collect additional blood is initiated. The determination of whether the donor meets the physiological conditions for supplementary collection is based on the following quantifiable indicators: diastolic blood pressure higher than 60 mmHg and systolic blood pressure lower than 140 mmHg, heart rate between 60 and 100 beats per minute, peripheral blood oxygen saturation higher than 95%, and no subjective symptoms such as dizziness or pallor. The formula for recalculating the volume of peripheral blood to be collected is:

[0116]

[0117] in To supplement the volume of peripheral blood collected, This is the difference in the number of cells mentioned above. This refers to the absolute count concentration of the target cell subpopulation in peripheral blood, determined based on the sample cell composition atlas. This is the preset recovery rate constant for this subpopulation of cells in the gradient density centrifugation and magnetic bead sorting process.

[0118] The formula for calculating the difference in cell number is expressed as follows:

[0119]

[0120] in: Indicates the difference in cell number. This indicates the expected optimal collection volume. This indicates the total number of live cells actually collected. In some embodiments, the minimum effective therapeutic dose is a minimum cell count standard set for a specific cell therapy product.

[0121] If the donor meets the physiological conditions for supplementary collection, the volume of peripheral blood to be collected will be recalculated based on the difference in cell count and the sample cell composition profile, and the supplementary collection process will be initiated. When recalculating the volume of peripheral blood to be collected, the percentage of the target cell subset in the sample cell composition profile will be considered, along with the expected recovery rate of the collection and sorting process. The criteria for determining whether a donor meets the physiological conditions for supplementary collection include stable vital signs, hemoglobin levels above the safe threshold, and no discomfort symptoms during the collection interval. Specific quantitative indicators for meeting the physiological conditions for supplementary collection include: systolic blood pressure not lower than 90 mmHg and not higher than 160 mmHg, diastolic blood pressure not lower than 60 mmHg and not higher than 100 mmHg, heart rate between 60 and 100 beats / min, peripheral blood hemoglobin concentration not lower than 120 g / L, the donor reporting no dizziness, fatigue, or palpitations, and the time interval since the last blood collection being at least 48 hours. If any of the above six indicators are not met, the donor is deemed not to meet the physiological conditions for supplementary collection, the collection process is terminated, and the patient enters the cryogenic storage preparation stage.

[0122] Optionally, the volume of peripheral blood collected during the supplementary collection process shall not exceed the upper limit of the safe blood collection volume for a single session. If the total number of viable cells actually collected reaches or exceeds the expected optimal collection volume, or if the donor does not meet the conditions for supplementary collection, the collection process ends and enters the cryogenic storage preparation stage. It can be understood that when the total number of viable cells actually collected exceeds the expected optimal collection volume, the enriched target immune cell suspension can directly proceed to the subsequent cryogenic storage steps.

[0123] See Figure 3 In the analysis of cell viability trends across multiple batches, the cell viability of each batch remained consistently above the preset 85% viability threshold, demonstrating the robustness of the peripheral blood immune cell collection and enrichment process. Specifically, the cell viability of the eight tested batches ranged from 93% to 96%, exhibiting slight fluctuations but consistently maintaining a high level, with no instances falling below the threshold. This indicates that the process can reliably guarantee cell viability in repeated production. The trend analysis showed that the fluctuation range of viability between batches was controlled within 3 percentage points, reflecting a high degree of standardization in the collection, sorting, and quality control processes, effectively reducing the impact of operational variations on cell viability. This high and stable viability performance provides a reliable input basis for the subsequent optimization algorithm in the low-temperature storage stage, and also verifies the effectiveness of the current cell collection and quality control process in ensuring the quality of cell products.

[0124] In one embodiment of the present invention, the final total number of cells collected, along with the cell viability and percentage of early apoptotic cells from the cell quality inspection data set, are input into a cryogenic storage optimization algorithm. Based on the input parameters and combined with the type characteristics of the target immune cells, the cryogenic storage optimization algorithm selects and determines an individualized cryopreservation solution formulation from multiple candidate formulations. This formulation includes a specific concentration combination of cryoprotectants, nutrients, antioxidants, and buffer systems.

[0125] The individualized adaptation rule complement is a recipe matching mapping function:

[0126]

[0127] in, This represents the survival rate of the input cells. Represents the percentage of cells in early apoptosis. The type of target immune cell is encoded. The first one selected from the candidate formulation library Each formula identifier. When the low-temperature storage optimization algorithm is executed, it first loads a rule base containing multiple sets of preset conditions. Each set of rules consists of the correspondence between survival rate ranges, apoptosis rate thresholds, cell type codes, and formula identifiers. The algorithm then inputs... , , The value is sequentially matched with the survival rate range, apoptosis rate threshold, and cell type code in the rule base. When all three satisfy the set conditions under the same rule, the recipe identifier mapped by that rule is determined. This is the output result. Regarding the calculation rules for the cooling program, the optimal freezing and cooling rate... After the solution is obtained, the algorithm uses the rate value and the input cell survival rate as a basis. The algorithm looks up a pre-stored segmented cooling strategy table, which defines the number of temperature plateau stages corresponding to the rate and survival rate combination, the target temperature value of each plateau, and the residence time, thereby generating an individualized cooling program. The storage strategy is determined based on the input percentage of early apoptotic cells. The storage strategy configuration file performs threshold checks, defining long-term storage temperature values ​​for different apoptosis percentage ranges, Boolean values ​​for whether to use gas-phase or liquid-phase storage modes, and the number of days for viability retesting. The algorithm output satisfies the current... The configuration items for the value serve as an individualized long-term storage strategy.

[0128] The low-temperature storage optimization algorithm consists of three submodules: a formulation matching submodule, a cooling program calculation submodule, and a storage strategy formulation submodule. The formulation matching submodule has a built-in formulation decision table based on cell viability. Proportion of early apoptotic cells and target immune cell types Mapping to the corresponding candidate recipes. The construction logic of the recipe decision table is as follows: setting a high activity threshold. and low apoptosis threshold .when and When, matching formulation A is used, wherein the concentration of dimethyl sulfoxide is 1.41 mol / L, the concentration of serum substitute is 0.02 v / L, the concentration of trehalose is 50 mmol / L, the antioxidant is propyl gallate at a concentration of 0.1 mmol / L, and the concentration of HEPES buffer is 10 mmol / L. and When using formulation C, the concentration of dimethyl sulfoxide is 2.12 mol / L, the concentration of serum substitute is 0.10 v / L, the concentration of trehalose is 150 mmol / L, the antioxidant is reduced glutathione at a concentration of 5 mmol / L, and the concentration of HEPES buffer is 25 mmol / L. Formula B was used, comprising: 1.41 mol / L dimethyl sulfoxide, 0.05% serum substitute, 100 mmol / L trehalose, 0.5 mmol / L ascorbic acid as the antioxidant, and 25 mmol / L HEPES buffer. Additionally, when the target immune cell type was T cell subsets, recombinant human interleukin-2 at a concentration of 20 ng / mL was added to the formula; when the target immune cell type was NK cells, recombinant human interleukin-15 at a concentration of 10 ng / mL was added to the formula.

[0129] The cooling program calculation submodule is based on the final total number of cells collected. and cell survival rate Calculate the optimal freezing and cooling rate The calculation formula is:

[0130]

[0131] in The optimal freezing and cooling rate is expressed in K / min. As the baseline cooling rate, ; This refers to the final total number of cells collected. As a baseline cell number, ; Cell viability is expressed as a decimal between 0 and 1; As the baseline survival rate, ; The fraction bar represents the natural logarithm function; the fraction bar represents division. This represents a multiplication operation. When the result is... When the speed is below 0.3 K / min, take 0.3 K / min. For speeds above 3.0 K / min, use 3.0 K / min. The refreezing rate is fixed at 100 K / min.

[0132] Simultaneously, the cryogenic storage optimization algorithm calculates the optimal freezing and thawing rates based on the final total cell collection and cell viability, generating a personalized cooling program. This program includes the residence time and cooling gradient of multiple temperature plateaus. Furthermore, based on the percentage of early apoptotic cells, the algorithm determines a personalized long-term storage strategy, including a recommended long-term storage temperature, whether to use a gas-liquid phase mixed storage mode, and planned cell viability verification time points. The enriched target immune cell suspension, suspended in the personalized cryopreservation solution, is aliquoted into cryogenic storage containers. The cryogenic storage containers are placed in the sample chamber of a programmed cooling instrument, and the cooling process is initiated. According to the personalized cooling program, the programmed cooling instrument first cools from room temperature to a preset first temperature plateau above freezing at a preset initial cooling rate, maintaining this plateau at a preset equilibrium time. After equilibrium, the programmed cooling instrument cools from the first temperature plateau to a preset transfer temperature at the optimal freezing and cooling rate. Once the transfer temperature is reached, the cryogenic storage containers are rapidly transferred to a long-term cryogenic storage device for long-term preservation.

[0133] In practice, the final total number of cells collected, along with the cell viability and percentage of early apoptotic cells from the cell quality control data set, are input into the cryopreservation optimization algorithm. Based on the input parameters and the type characteristics of the target immune cells, the algorithm selects and determines an individualized cryopreservation solution formulation from multiple candidate formulations. The individualized adaptation of the cryopreservation solution formulation incorporates cell type correction rules. After the basic formulation is output from the formulation decision table, the following corrections are applied based on the target immune cell type: when the target immune cell is NK cells, the dimethyl sulfoxide concentration in the basic formulation is reduced by 0.35 mol / L, while the trehalose concentration is increased by 50 mmol / L; when the target immune cell is B cells, the serum substitute concentration in the basic formulation is increased by 0.03 volume fraction. This correction rule is based on the differences in osmotic pressure tolerance and membrane fluidity of different immune cell subsets to cryoprotectants.

[0134] The individualized adaptation of the optimal cooling rate introduces two correction factors. The basic optimal freezing cooling rate is obtained in the cooling program calculation submodule. Then, the following correction steps were added: when the proportion of early apoptotic cells... At that time, Multiplied by apoptosis correction factor , To slow down the cooling rate and thus reduce the risk of membrane damage to cells in the early apoptotic state during freezing; when the final total number of cells collected... At that time, Multiply by the total amount correction factor , This is to compensate for the temperature gradient effect caused by the delay in heat conduction within large-volume samples. The two correction factors can be used together; when combining them, the product of the two factors is summed. Multiplying these values ​​yields the final optimal freezing rate.

[0135] The individualized adaptation rule for storage temperature and retesting cycle is as follows: the storage temperature is selected based on the proportion of early apoptotic cells. . The storage temperature was set to −150°C, and gaseous liquid nitrogen was used for storage. The storage temperature was set to −150°C, and gaseous liquid nitrogen was used for storage, but the first verification cycle was shortened. The storage temperature was set to −196°C, and liquid nitrogen was used for storage. The determination of the review cycle was also related to… Storage temperature: upon receipt For each batch, the initial verification period is 90 days after warehousing, and subsequent verification cycles are every 180 days; upon warehousing... For each batch, the initial verification period is 60 days after warehousing, and subsequent verification cycles are every 120 days; upon warehousing... For each batch, the first verification is conducted on the 30th day after storage, and subsequent verifications are conducted every 90 days. The dynamic adjustment rule is as follows: if the cell viability rate measured at each verification decreases by more than 0.10 compared to the previous verification, the subsequent verification cycle will be shortened to half of the current cycle, with a minimum verification cycle of no less than 30 days.

[0136] Individualized cryopreservation solutions include specific concentrations of cryoprotectants, nutrients, antioxidants, and buffer systems. The concentrations of components in candidate solutions vary depending on the cell type and state, as shown in Table 2.

[0137] Table 2: Examples of Candidate Formulations for Personalized Cryopreservation Solutions

[0138] Ingredient Category Formula A (High Viability Cells) Formula B (Standard Cells) Formula C (cells with high apoptosis tendency) Dimethyl sulfoxide concentration 5% 10% 7.5% serum substitute concentration 2% 5% 10% Trehalose concentration 50mM 100mM 150mM Antioxidant types propyl gallate ascorbic acid Glutathione HEPES buffer concentration 10mM 25mM 25mM

[0139] In practical implementation, the cryogenic storage optimization algorithm calculates the optimal freezing-cooling rate and thawing rate based on the final total number of cells collected and the cell viability, generating an individualized cooling program. This individualized cooling program includes the residence time and cooling gradient of multiple temperature plateaus. One empirical formula for calculating the optimal freezing-cooling rate is expressed as:

[0140]

[0141] in: This represents the calculated optimal freezing rate. This indicates the final total number of cells collected. This indicates the cell viability rate in the cell quality inspection dataset. This is a correction factor related to the target immune cell type. It is understood that the thawing rate is typically set to a fixed value much higher than the freezing rate, such as 100°C / min. Furthermore, based on the percentage of early apoptotic cells, the cryogenic storage optimization algorithm determines an individualized long-term storage strategy. This individualized strategy includes a recommended long-term storage temperature, whether to use a gas-liquid phase mixed storage mode, and a planned cell viability retest time point. In some embodiments, when the percentage of early apoptotic cells is higher than a set threshold, the long-term storage strategy recommends using a lower storage temperature and shortening the interval between the first viability retests.

[0142] In some embodiments, the personalized cooling procedure includes aliquoting a suspension of enriched target immune cells in a personalized cryopreservation solution into a cryopreservation container. The cryopreservation container may be a cryopreservation bag or cryovial. The cryopreservation container is placed in the sample chamber of the programmed cooling instrument, and the cooling process is initiated. Following the personalized cooling procedure, the programmed cooling instrument is controlled to first cool from room temperature to a preset first temperature plateau above freezing at a preset initial cooling rate, and maintain this plateau at a preset equilibration time. For example, the initial cooling rate is set to -1°C / min, cooling from room temperature (approximately 25°C) to a first temperature plateau of 4°C, and equilibrating at 4°C for 10 minutes. After equilibration, the programmed cooling instrument is controlled to cool from the first temperature plateau to a preset transfer temperature at an optimal freezing rate. The transfer temperature is typically set between -80°C and -120°C. Once the transfer temperature is reached, the cryopreservation container is rapidly transferred to a long-term cryopreservation device for long-term storage. Optionally, the transfer operation is performed on dry ice to prevent temperature rebound. It is understandable that the specific values ​​of the first temperature plateau, equilibration time, and transfer temperature are output by the low-temperature storage optimization algorithm based on the cell type and state.

[0143] See Figure 4In the analysis of core indicators for immune cell quality control, early apoptosis rate is a key quality control indicator for assessing cell functional status and storage suitability. Specifically, the enriched target immune cell suspension was sampled, and the expression ratio of early apoptosis markers was detected by flow cytometry. Early apoptosis rate data for each sample were obtained and plotted as a trend curve. The curve shows that the early apoptosis rate of samples 1 to 4 gradually increased, with sample 4 reaching a peak of 10.2%, indicating potential stress damage during collection or processing. The apoptosis rate of sample 5 decreased to 4.1%, indicating a significant improvement in cell status after process optimization. This apoptosis rate data will be used as a core input parameter in a cryogenic storage optimization algorithm to match individualized cryopreservation solution formulations, design cooling procedures, and formulate long-term storage strategies. When the early apoptosis rate exceeds a preset threshold, the algorithm will prioritize formulations containing higher concentrations of cryoprotectants and antioxidants (such as formulation C in Table 2) and recommend lower storage temperatures and more frequent viability retesting cycles to minimize the risk of cell viability decline during storage.

[0144] In one embodiment of the present invention, after the enriched target immune cell suspension enters long-term cryogenic storage, representative samples are extracted from the long-term storage device at preset intervals for rapid miniaturized viability testing to obtain cell viability data during the storage period. The cell viability data during the storage period is longitudinally compared with the cell quality inspection data set at the time of cell entry into the storage facility to analyze the cell viability decay trend. Simultaneously, historical data of the storage device's operating parameters are recorded, including temperature fluctuation range and liquid nitrogen replenishment records. The cell viability decay trend and the historical data of the storage device's operating parameters are input into a storage risk prediction model. The storage risk prediction model outputs a report on the remaining estimated effective storage time and risk level assessment of the current batch of cells. Based on the risk level assessment report, a decision is made on whether to adjust the long-term storage strategy or to advance the usage plan for the cell products. The storage risk prediction model receives input cell viability decay trend data, which includes the rate of change in cell survival rate and the proportion of apoptotic cells in each sampling test. The storage risk prediction model also receives input historical data of the storage device's operating parameters, particularly the number and duration of abnormal temperature events exceeding preset thresholds. The storage risk prediction model matches and fits cell viability decay trend data with historical data on storage device operating parameters to a built-in database of viability decay curves for different cell types under different storage conditions. Based on the fitting results, it predicts cell survival rate and functional activity indicators at a specific future time point and extrapolates the time point when cell viability drops to a preset critical value; this time point represents the remaining estimated effective storage time. The viability decay curve database uses an exponential decay model to parameterize historical storage data. The mathematical expression of the exponential decay model is:

[0145]

[0146] in, Indicates storage time Predicted cell viability at that time; Indicates the initial cell viability upon entry into the storage facility; Indicates storage time, in days; Represents the decay constant, in days. ; Represented by natural constant An exponential function with base 0; This represents a multiplication operation. Decay constants for different cell types under different storage conditions. The method was determined by fitting historical batch data: For each historical batch, the cell viability rate of each sampling test and the corresponding storage time were fitted using the exponential decay model described above with nonlinear least squares to obtain the data for that batch. Value; for all batches of the same cell type and under the same storage conditions. The values ​​are calculated to be the mean and standard deviation, and stored as database entries. Each record in the database contains the cell type identifier, storage temperature, cryopreservation solution formula number, and corresponding decay constant. Mean and standard deviation. The matching process involves retrieving the corresponding values ​​from the database based on the cell type, storage temperature, and cryopreservation solution formula number used for the current batch. The mean was used as the initial value. The fitting process was as follows: all cell viability data points measured in the current batch were fitted to the exponential decay model using the Levenberg-Marquardt algorithm with nonlinear least squares fitting, and the data was then used in the database. The mean is used as the initial value, and the personalized decay constant for the current batch is obtained through iterative fitting. .

[0147] Taking into account the remaining estimated effective storage time, the severity of abnormal temperature events, and the current rate of vitality decay, a risk matrix is ​​used to calculate and output a risk level assessment report that includes high, medium, and low risk levels.

[0148] In practice, after the enriched target immune cell suspension is placed into long-term cryogenic storage, representative samples are extracted from the long-term storage device at preset intervals for rapid miniaturized viability testing to obtain cell viability data during the storage period. This rapid miniaturized viability testing is performed using a fluorescent dye-based live cell counting chamber or a microfluidic cell viability analysis chip. The cell viability data during the storage period is longitudinally compared with the cell quality inspection data set at the time of cell entry into the storage facility to analyze the cell viability decay trend. The cell viability decay trend is characterized by calculating the change in the ratio of cell viability at each sampling test to the baseline viability at the time of entry into the storage facility and fitting a time function. Simultaneously, historical data on the storage device's operating parameters are recorded. This historical data includes temperature fluctuation range and liquid nitrogen replenishment records. The temperature fluctuation range is recorded at fixed time intervals using a built-in temperature recorder in the storage device, and the liquid nitrogen replenishment records include the date, amount, and liquid level of each replenishment.

[0149] In practice, historical data on cell viability decay trends and storage device operating parameters are input into the storage risk prediction model. The model outputs a report on the remaining estimated effective storage time and risk level assessment for the current batch of cells. The workflow of the storage risk prediction model includes receiving input cell viability decay trend data, which includes the rate of change in cell viability and the proportion of apoptotic cells in each sampling test. The model receives historical data on storage device operating parameters, particularly the number and duration of temperature anomaly events exceeding preset thresholds. These preset thresholds are set for specific storage devices, and the complete settings are as follows: For storage devices using gas-phase liquid nitrogen storage, the preset temperature threshold is −140°C. A temperature anomaly event is recorded when any temperature sensor reading within the storage device exceeds −140°C for more than 5 minutes. For storage devices using liquid-phase liquid nitrogen storage, the preset temperature threshold is −185°C. A temperature anomaly event is recorded when any temperature sensor reading within the storage device exceeds −185°C for more than 5 minutes. The average duration of a single temperature anomaly event is also specified. The time from the trigger threshold to recovery below the threshold is recorded in minutes. For example, in a storage environment of -150°C, a temperature above -140°C is considered an abnormal event. The storage risk prediction model matches and fits cell viability decay trend data with historical data of storage device operating parameters to a database of viability decay curves for different cell types under different storage conditions built into the model. It can be understood that the viability decay curve database is derived from research data of historical storage batches.

[0150] Based on the fitting results, the storage risk prediction model predicts the cell survival rate and functional activity indicators at a specific future time point, and extrapolates the time point at which the cell survival rate drops to a preset critical value; this time point is the remaining estimated effective storage time. Taking into account the length of the remaining estimated effective storage time, the severity of abnormal temperature events, and the current rate of viability decay, a risk matrix is ​​calculated to output a risk level assessment report including high, medium, and low risk levels. The risk level can be quantitatively scored using the following formula:

[0151]

[0152] in: Represents a quantitative risk score. This represents the calculated remaining estimated effective storage time. The expected storage time represents the standard for this cell type. This represents the number of times abnormal temperature events occur. This represents the average duration of an abnormal temperature event. Represents the absolute value of the current rate of cell viability decay. , and These are preset weighting coefficients. Based on risk quantification scoring. The numerical range in which it falls corresponds to three risk levels: high, medium, and low.

[0153] Based on the risk assessment report, a decision is made as to whether to adjust the long-term storage strategy or to advance the usage plan for cell products. It is understood that when the risk assessment report indicates a high risk, possible adjustments may include transferring cells from gaseous nitrogen storage to liquid nitrogen storage, increasing the frequency of viability testing, or notifying the user to prioritize the use of that batch of cells. In some embodiments, the risk assessment report is associated with cell storage records and triggers automatic alarms or task prompts. Instructions for adjusting the long-term storage strategy are executed automatically by the system or manually after confirmation by management personnel.

[0154] See Figure 5In the correlation analysis of factors influencing storage risk, the linear correlation between variables was quantified using the Pearson correlation coefficient. Specifically, the cell viability decay showed a very strong negative correlation with the number of temperature anomalies (correlation coefficient of -0.98), indicating that the higher the frequency of temperature anomalies, the slower the rate of cell viability decay. This phenomenon is directly related to the damage mechanism of temperature fluctuations to cell viability during low-temperature storage. Meanwhile, the cell viability decay showed a very strong positive correlation with the duration of anomalies (0.92) and the risk score (0.87), indicating that the longer the duration of anomalies and the higher the risk score, the more significant the trend of cell viability decay. Further analysis showed that the number of temperature anomalies also showed a strong negative correlation with the duration of anomalies (-0.85) and the risk score (-0.80), reflecting an inverse correlation between the number of anomalies and their duration and risk level. The duration of anomalies showed a very strong positive correlation with the risk score (0.94), meaning that the duration of anomalies is one of the core factors driving the increase in the risk score. The correlation heatmap intuitively reveals the synergistic and antagonistic relationships among various driving factors of storage risk, providing a quantitative basis for feature selection, weight allocation and risk matrix construction of subsequent storage risk prediction models, and can effectively guide the optimization of low-temperature storage strategies and the setting of risk warning thresholds.

[0155] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for efficient collection and cryogenic storage of peripheral blood immune cells, characterized in that, The method includes: Collect peripheral blood samples and obtain donor health status information associated with the peripheral blood samples; The peripheral blood sample is subjected to fully automated blood cell analysis to generate a sample cell composition atlas, which includes the absolute percentage of lymphocyte subsets, the purity of monocytes, and the contamination ratio of granulocytes. Based on the donor's health status information and the sample cell composition map, calculations are performed using a preset data analysis model to determine the expected optimal collection amount of target immune cells and the corresponding most suitable cell collection process. According to the optimal cell collection process, the peripheral blood sample is subjected to gradient density centrifugation and specific magnetic bead sorting to generate an enriched target immune cell suspension. The viability and activity of the enriched target immune cell suspension are detected to generate a cell quality control data set. Based on the comparison between the cell quality control data set and the expected optimal collection amount, it is determined whether to initiate a supplementary collection process. Based on the final total number of cells collected and the cell quality inspection data set, a low-temperature storage optimization algorithm is invoked to calculate an individualized low-temperature preservation solution formula, cooling procedure, and long-term storage strategy for the enriched target immune cell suspension in a -150°C storage environment.

2. The method for efficient collection and cryogenic storage of peripheral blood immune cells according to claim 1, characterized in that, The peripheral blood sample is subjected to fully automated hematology analysis to generate a sample cell composition atlas, including: The donor's health status information includes the donor's age, recent medication records, and total white blood cell count in routine blood tests. A first subsample for testing is separated from the peripheral blood sample. The first subsample for testing is subjected to multicolor flow cytometry analysis of cell surface markers to identify the absolute number of T cell subsets, B cells, and NK cells, and to calculate the percentage of each subset relative to nucleated cells. A second subsample for testing is extracted from the peripheral blood sample. An automated cell morphology analysis based on cell light scattering characteristics is performed on the second subsample to distinguish the morphological differences between the monocytes and lymphocytes, and to calculate the purity of the monocytes. Combining the results of multicolor flow cytometry analysis with the results of automated cell morphology analysis, the contamination ratio of the granulocytes is assessed by a preset contamination judgment logic, which makes judgments based on cell size, granularity, and the expression intensity of specific markers. The absolute percentage count of the lymphocyte subsets, the purity of the monocytes, and the contamination ratio of the granulocytes are integrated to form a structured atlas of the sample cell composition.

3. The method for efficient collection and cryogenic storage of peripheral blood immune cells according to claim 2, characterized in that, Based on the donor's health status information and the sample cell composition atlas, a preset data analysis model is used to calculate and determine the expected optimal collection amount of target immune cells and the corresponding most suitable cell collection process, including: The donor's age, recent medication records, and total white blood cell count are input into the donor status assessment submodule of the data analysis model to calculate the donor cell collection suitability score. The absolute percentage of lymphocyte subsets, the purity of monocytes, and the contamination ratio of granulocytes in the sample cell composition atlas are input into the cell quality prediction submodule of the data analysis model to predict the potential viability and proliferation capacity of the collected cell products. By integrating the donor cell collection suitability score with the predicted results of the potential viability and proliferation capacity of the collected cell products, the expected optimal collection amount of the target immune cells is calculated using a target cell yield optimization algorithm, while ensuring cell quality. Based on the expected optimal collection volume, the type of target immune cells, and the contamination ratio of granulocytes, the most suitable cell collection process scheme, including gradient density centrifugation parameters, magnetic bead sorting antibody combination, and sorting buffer formulation, is matched from a preset process library.

4. The method for efficient collection and cryogenic storage of peripheral blood immune cells according to claim 3, characterized in that, The peripheral blood sample was subjected to gradient density centrifugation and specific magnetic bead sorting to generate an enriched suspension of target immune cells, including: According to the gradient density centrifugation parameters specified in the optimal cell collection process, including centrifugation force and centrifugation time, the peripheral blood sample is subjected to stratified centrifugation to separate a cell layer rich in mononuclear cells. The cell layer rich in mononuclear cells was collected and washed and resuspended using the sorting buffer specified in the optimal cell collection process to prepare a pre-sorting cell suspension. According to the magnetic bead sorting antibody combination specified in the optimal cell collection process, the specific magnetic beads are incubated with the pre-sorted cell suspension to form a magnetic bead-target cell complex. The incubated mixed suspension was placed on a magnetic sorting rack to separate the target cell population bound to magnetic beads, and the unbound cells were removed. The isolated target cell population was dissociated and washed to remove magnetic beads, ultimately yielding the enriched target immune cell suspension suspended in the preservation solution.

5. The method for efficient collection and cryogenic storage of peripheral blood immune cells according to claim 4, characterized in that, The enriched target immune cell suspension was subjected to viability and activity testing to generate a cell quality control data set, including: Samples were taken from the enriched target immune cell suspension and subjected to trypan blue rejection assay. The number of live cells and the total number of cells were calculated using a cell counter to obtain the cell viability. Flow cytometry analysis was performed on sampled cells to detect the expression ratio of early apoptosis markers and calculate the percentage of early apoptotic cells. Staining and analyzing target functional markers on sampled cells to detect the proportion of cell subpopulations with specific functions in living cells; The cell quality control data set is composed of the cell survival rate, the percentage of early apoptotic cells, and the proportion of cell subpopulations with specific functions.

6. The method for efficient collection and cryogenic storage of peripheral blood immune cells according to claim 5, characterized in that, Based on the comparison between the cell quality inspection data set and the expected optimal collection volume, determine whether to initiate a supplementary collection process, including: The cell viability rate is read from the cell quality inspection data set. If the cell viability rate is lower than the preset viability rate threshold, the current collection is determined to be a failure and a new round of complete sample collection process needs to be performed. If the cell survival rate is not lower than the preset survival rate threshold, the total number of live cells actually collected is compared with the expected optimal collection amount. If the actual total number of live cells collected is lower than the expected optimal collection amount, but higher than the minimum effective treatment dose, the cell number difference is calculated, and it is determined whether the donor meets the physiological conditions for supplementary collection. If the donor meets the physiological conditions for supplementary collection, the volume of peripheral blood to be supplemented is recalculated based on the difference in cell number and the cell composition profile of the sample, and the supplementary collection process is initiated. If the total number of live cells actually collected reaches or exceeds the expected optimal collection amount, or if the donor does not meet the conditions for supplementary collection, the collection process ends and enters the low-temperature storage preparation stage.

7. The method for efficient collection and cryogenic storage of peripheral blood immune cells according to claim 6, characterized in that, The process of invoking a low-temperature storage optimization algorithm to calculate an individualized low-temperature preservation solution formulation, cooling procedure, and long-term storage strategy for the enriched target immune cell suspension includes: The final total number of cells collected, along with the cell viability and percentage of early apoptotic cells from the cell quality inspection data set, are input into the low-temperature storage optimization algorithm. The low-temperature storage optimization algorithm selects and determines the individualized low-temperature preservation solution formula from multiple candidate formulas based on input parameters and combined with the type characteristics of the target immune cells. The formula includes a specific concentration combination of cryoprotectants, nutrients, antioxidants and buffer systems. Meanwhile, the low-temperature storage optimization algorithm calculates the optimal freezing and cooling rate and the rewarming and thawing rate based on the final total number of cells collected and the cell viability, and generates an individualized cooling program, which includes the dwell time of multiple temperature platforms and the cooling gradient. Furthermore, based on the percentage of early apoptotic cells, the low-temperature storage optimization algorithm determines an individualized long-term storage strategy, which includes a recommended long-term storage temperature, whether to use a gas-liquid phase mixed storage mode, and a planned cell viability verification time point.

8. The method for efficient collection and cryogenic storage of peripheral blood immune cells according to claim 7, characterized in that, The individualized cooling procedure includes the following steps: The enriched target immune cell suspension, which is suspended in the individualized cryopreservation solution, is dispensed into a cryopreservation container. Place the low-temperature storage container in the sample chamber of the programmed cooling instrument and start the cooling process; According to the individualized cooling program, the programmable cooling device is controlled to first cool from room temperature to a preset first temperature plateau above freezing point at a preset initial cooling rate, and maintain a preset equilibrium time on the first temperature plateau. After balancing, the programmed cooling device is controlled to cool from the first temperature platform to the preset transfer temperature at the optimal freezing and cooling rate. Once the transfer temperature is reached, the cryogenic storage container is quickly transferred to a long-term cryogenic storage device for long-term preservation.

9. The method for efficient collection and cryogenic storage of peripheral blood immune cells according to claim 8, characterized in that, It also includes dynamic monitoring and risk assessment steps during cell storage: After the enriched target immune cell suspension is placed into long-term low-temperature storage, representative samples are extracted from the long-term storage device at preset intervals for rapid miniaturized viability detection to obtain cell viability data during the storage period. The cell viability data during the storage period is compared longitudinally with the cell quality inspection data set at the time of cell entry into the warehouse to analyze the trend of cell viability decline. Simultaneously, historical data of the storage device's operating parameters are recorded, including temperature fluctuation range and liquid nitrogen replenishment records. The cell viability decline trend and historical data of the storage device operating parameters are input into the storage risk prediction model; The storage risk prediction model outputs a report on the remaining estimated effective storage time and risk level assessment of the current batch of cells. Based on the aforementioned risk assessment report, a decision will be made as to whether to adjust the long-term storage strategy or to plan the use of cell products in advance.

10. The method for efficient collection and cryogenic storage of peripheral blood immune cells according to claim 9, characterized in that, The workflow of the storage risk prediction model includes: Receive the input cell viability decline trend data, which includes the rate of change in cell survival rate and the proportion of apoptotic cells for each sampling test; Receive historical data of the operating parameters of the storage device, especially the number and duration of abnormal temperature events exceeding a preset threshold; The cell viability decline trend data and the historical data of the storage device operating parameters are matched and fitted with the model’s built-in database of viability decline curves for different cell types under different storage conditions. Based on the fitting results, the cell survival rate and functional activity indicators are predicted at a specific time point in the future, and the time point when the cell survival rate drops to a preset critical value is extrapolated and calculated. The time point is the remaining expected effective storage time. Based on the remaining estimated effective storage time, the severity of abnormal temperature events, and the current rate of vitality decay, a risk assessment report with high, medium, and low risk levels is output through risk matrix calculation.

Images