Immune density and centrifugation technology combined blood leukocyte accurate separation method

Abstract

The embodiment of the invention provides a blood leukocyte precise separation method combining immune density and a centrifugation technology. The blood leukocyte precise separation method comprises the following steps: performing multicolor immunofluorescence labeling on leukocyte subgroups in a blood sample; performing flow cytometry analysis on the marked sample, and establishing an initial multicolor fluorescence compensation matrix for the current sample individual; dynamically adjusting and generating a corrected compensation matrix of the individual specificity based on the deviation degree of the initial compensation matrix in a sample group reference range; carrying out association mapping on a specific physical sorting interval which is determined after the signal recalculation and the enclosure and is rich in the target leukocyte subgroup and a density gradient centrifugal separation condition; and directly carrying out density gradient centrifugation operation on the blood sample subjected to immunolabeling by utilizing the association mapping. According to the scheme of the embodiment of the invention, the high-purity specific leukocyte subgroup can be efficiently separated from a complex blood sample on the premise of keeping high cell activity and flux.

Description

Technical Field

[0001] This application relates to the field of biomedical engineering technology, specifically to a method for precise separation of blood leukocytes that combines immune density and centrifugation techniques. Background Technology

[0002] The main purpose of blood leukocyte separation in biomedical engineering is to improve the purity and recovery rate of leukocytes. However, in practical applications, challenges remain, namely how to efficiently separate specific leukocyte subpopulations with high purity from complex blood samples while maintaining high cell viability and processing throughput. This involves several key technical aspects, such as antibody selection, centrifugation condition optimization, and cell damage control. Summary of the Invention

[0003] In view of this, the present disclosure provides a method for precise separation of blood leukocytes by combining immune density and centrifugation techniques, which at least partially solves the problems existing in the prior art.

[0004] A method for precise separation of blood leukocytes combining immunodiffusion and centrifugation techniques includes the following steps: Multicolor immunofluorescence labeling of leukocyte subsets in blood samples; The labeled samples were analyzed by flow cytometry, and an initial multicolor fluorescence compensation matrix was established for the current individual sample based on the real-time fluorescence signal data obtained from the analysis. Based on the deviation of the initial compensation matrix from the reference range of the sample population, a correction compensation matrix specific to the individual is dynamically adjusted and generated, and the correction compensation matrix is ​​used to recalculate and gating the flow cytometry analysis signal of the sample. The specific physical sorting interval rich in the target leukocyte subpopulation, determined after signal recalculation and gatening, is correlated and mapped with the density gradient centrifugation separation conditions. Using the aforementioned association mapping, the immunolabeled blood sample is directly subjected to density gradient centrifugation, thereby achieving high-purity separation of the target leukocyte subset based on cell physical density.

[0005] In one specific implementation, after obtaining the initial multicolor fluorescence compensation matrix, the median signal intensity of the cell subpopulation present in the current individual sample in the fluorescence channel is identified, and all compensation coefficients involving that channel in the initial compensation matrix are proportionally scaled and corrected based on the degree of deviation of the median signal intensity from the population reference baseline.

[0006] In one specific implementation, after generating an initial multicolor fluorescence compensation matrix, an event confidence matrix is ​​created in parallel. The event confidence matrix is ​​used to evaluate the Mahalanobis distance of each cell event in the compensated space after initial compensation. If a cell event exceeds a predetermined threshold and exhibits a Mahalanobis distance exceeding the threshold, a backup compensation matrix is ​​generated as the corrected matrix.

[0007] In one specific implementation, the recalculation of the flow cytometry signal of the sample and the cell subpopulation gating using the corrected compensation matrix include: Recalculate the normalized fluorescence intensity N_i = (I_i-minI) / (maxI-minI) for each event point, where I_i is the original fluorescence intensity of the event point, minI is the minimum fluorescence intensity of all events in the channel, and maxI is the maximum fluorescence intensity. Compare N_i with a pre-set multicolor fluorescent labeling threshold; The signal deviation coefficient S is calculated using the following formula: S = |N_i-T_i| / T_i, where T_i is the expected ideal normalized intensity. If S < ε, where ε is the upper limit of the allowable deviation coefficient, then the event is considered to belong to the target subgroup.

[0008] In one specific implementation, the step of associating the specific physical sorting interval rich in the target leukocyte subpopulation, determined after signal recalculation and gating, with the density gradient centrifugation separation conditions includes: The mean density value D_mean of the target leukocyte subset; The ideal centrifugal gradient range G = [D_mean-ΔD, D_mean + ΔD] is calculated using the empirical formula, where ΔD represents the allowable density deviation range. Match this interval with the interlayer interface in the density gradient column; By controlling the centrifugation speed V and time t, effective sorting of corresponding density regions can be achieved.

[0009] In one specific embodiment, the step of directly performing density gradient centrifugation on the immunolabeled blood sample using the association mapping includes: The centrifugation rate v is set as a calculation formula based on cell mass size: v = k × m^α, where m is the white blood cell mass, k is a proportionality constant, and α is a power exponent. Record the sample liquid volume V_sample before each centrifugation; The required centrifuge tube filling rate F = (V_selected / V_cap) is calculated according to the following formula, where V_selected is the theoretical volume of the selected range and V_cap is the maximum capacity of the centrifuge tube; If F ≤ F_max, where F_max is the design fill limit, then centrifugation is performed; otherwise, the sample size is reduced before proceeding.

[0010] In one specific implementation, the step of dynamically adjusting and generating an individual-specific corrected compensation matrix based on the deviation of the initial compensation matrix from the reference range of the sample population includes: Perform independent standard deviation analysis on each fluorescent label to obtain the standard deviation vector σ_i; The adjustment coefficient R is constructed according to the following formula: R = α × Σ(σ_i) / max(σ), where α is the weighting factor; R is incorporated into the adjustment factor of the final correction matrix; If R > R_threshold, where R_threshold is the preset significance threshold, then the adjustment is applied; otherwise, the original compensation matrix is ​​retained.

[0011] In one specific implementation, the step of associating the specific physical sorting interval rich in the target leukocyte subpopulation, determined after signal recalculation and gating, with the density gradient centrifugation separation conditions includes: Establish migration curves M(v) of the target cell subpopulation under different centrifugal forces; The optimal sorting region is predicted using the fitted polynomial M(v) = a×v³ + b×v² + c×v + d; The optimal centrifugation rate is selected according to the following formula: v_opt = argmax{M(v)}, that is, to find the v that maximizes M(v); Use v_opt as the actual centrifugation parameter for the sorting operation.

[0012] In one specific embodiment, the step of directly performing density gradient centrifugation on the immunolabeled blood sample using the association mapping includes: Set the centrifuge speed V_set and the sample filling height H_set; The sorting efficiency is evaluated using the following formula: Efficiency = (V_output × t_eff) / (V_input × t_total), where V_output is the volume of successfully sorted cells, t_eff is the effective processing time, V_input is the original sample volume, and t_total is the total time consumed in the entire experiment. If Efficiency ≥ Eff_threshold, the process is considered controllable.

[0013] In one specific embodiment, the step of directly performing density gradient centrifugation on the immunolabeled blood sample using the association mapping includes: Based on the current performance of the centrifuge, set the maximum centrifugal pressure that it can withstand, P_max. Based on the cell viability detection indicators before and after centrifugation, the cell viability index A_index = (A_after / A_before) × 100%, where A_before is the cell viability before centrifugation and A_after is the cell viability after centrifugation. The critical function for cell activity is defined as C = A_index × (1+λ × (F_actual-F_design)²), where λ is the pressure sensitivity factor and F_design is the standard centrifugal force. If C ≥ C_threshold, where C_threshold is the minimum survival requirement, then the centrifugation procedure is executed.

[0014] This disclosure provides a method for precise separation of blood leukocytes combining immunofluorescence and centrifugation techniques, comprising the following steps: multicolor immunofluorescence labeling of leukocyte subpopulations in a blood sample; flow cytometry analysis of the labeled sample; establishing an initial multicolor fluorescence compensation matrix for the current individual sample based on the real-time fluorescence signal data obtained from the analysis; dynamically adjusting and generating an individual-specific corrected compensation matrix based on the deviation of the initial compensation matrix from the reference range of the sample population; recalculating and gating the flow cytometry analysis signal of the sample using the corrected compensation matrix; mapping the specific physical sorting interval rich in the target leukocyte subpopulation determined after signal recalculation and gating to density gradient centrifugation conditions; and directly performing density gradient centrifugation on the immunolabeled blood sample using the mapping, thereby achieving high-purity separation of the target leukocyte subpopulation based on cell physical density. The solution of this disclosure addresses how to efficiently separate high-purity specific leukocyte subpopulations from complex blood samples while maintaining high cell viability and throughput. Attached Figure Description

[0015] In the accompanying drawings, unless otherwise specified, the same reference numerals throughout the various drawings denote the same or similar parts or elements. These drawings are not necessarily drawn to scale. It should be understood that these drawings depict only some embodiments disclosed in this application and should not be construed as limiting the scope of this application.

[0016] Figure 1 This is a flowchart of the precise blood leukocyte separation method combining immune density and centrifugation techniques described in this application; Figure 2 This application describes the process of recalculating the flow cytometry analysis signal of the sample and gating the cell subpopulation using the corrected compensation matrix. Figure 3 This application describes a process for mapping a specific physical sorting region rich in the target leukocyte subpopulation, determined after signal recalculation and gating, to density gradient centrifugation separation conditions. Figure 4 This is a flowchart of the density gradient centrifugation operation performed directly on immunolabeled blood samples using correlation mapping in this application. Figure 5 This is a flowchart illustrating how the initial compensation matrix deviates from the reference range of the sample population, and how the individual-specific corrected compensation matrix is ​​dynamically adjusted and generated. Figure 6 This is a flowchart illustrating the association and mapping of the specific physical sorting interval rich in the target leukocyte subpopulation, determined after signal recalculation and gating, with density gradient centrifugation separation conditions. Figure 7 This is a flowchart illustrating the use of correlation mapping in this application to directly perform density gradient centrifugation on immunolabeled blood samples. Figure 8 This is a flowchart illustrating the use of association mapping in this application to directly perform density gradient centrifugation on immunolabeled blood samples. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the embodiments of this disclosure will be further described in detail below with reference to the accompanying drawings. The illustrative implementation methods and descriptions of the embodiments of this disclosure are only used to explain the embodiments of this disclosure and are not intended to limit the embodiments of this disclosure.

[0018] Next, referring to the accompanying drawings, a method for precise separation of blood leukocytes combining immunodensity and centrifugation techniques according to the present invention will be described. This method aims to achieve efficient and high-purity separation of target leukocyte subsets. Figure 1 As shown, the method of this application includes: S101: Multicolor immunofluorescence labeling of leukocyte subsets in blood samples. Typically, antibodies are designed for different types of leukocyte subsets and labeled with different fluorescent dyes, such as FITC, PE, and PerCP. This allows each leukocyte to produce a unique fluorescence signal at a specific excitation wavelength, enabling differentiation of different cell types in subsequent flow cytometry analysis. This multicolor fluorescent labeling allows for the simultaneous detection of multiple surface markers, improving the accuracy and efficiency of cell classification. For example, when detecting CD4+ T cells, a combination of antibodies against CD3, CD4, and CD8 might be used for labeling to ensure the identification and counting of specific cell subsets.

[0019] S102: Perform flow cytometry analysis on the labeled samples, and establish an initial multicolor fluorescence compensation matrix for the current sample individual based on the real-time fluorescence signal data obtained from the analysis.

[0020] Flow cytometry analysis uses laser excitation to detect fluorescence signals in samples, collects and converts them into digital signals for software analysis. Because different fluorescent dyes have spectral overlap, signals from adjacent channels can interfere with each other, necessitating adjustment using a compensation matrix. The initial multicolor fluorescence compensation matrix is ​​the result of calculations and optimizations for all possible cross-interference, reducing signal confusion in subsequent data processing and ensuring accurate separation of cells with different fluorescent labels.

[0021] In one example, after multicolor immunofluorescence labeling of leukocyte subsets in a blood sample (e.g., using CD3-FITC, CD4-PE, CD8-APC, and CD45-PerCP) is completed, the sample is subjected to flow cytometry analysis. First, the lymphocyte population is initially delineated by acquiring forward scattered (FSC) and side scattered (SSC) signals. Subsequently, the instrument sequentially acquires the raw signals of each cell in each fluorescence detection channel (e.g., FL1-FITC, FL2-PE, FL3-APC, FL4-PerCP). Intensity; Based on these real-time acquired raw fluorescence signal data, the flow cytometry software first uses the pre-set standard spectral overlap coefficients of single positive control samples (such as samples labeled only with CD3-FITC, samples labeled only with CD4-PE, etc.) to construct an initial fluorescence compensation matrix. This matrix quantitatively describes the degree of "leakage" of each fluorescent dye signal to other detection channels in mathematical form (usually an n×n square matrix, where n is the number of fluorescent dyes). For the specific individual sample to be analyzed, this initial matrix serves as the basis for subsequent dynamic adjustments and signal demixing.

[0022] S103: Based on the degree of deviation of the initial compensation matrix within the reference range of the sample population, dynamically adjust and generate an individual-specific corrected compensation matrix, and use the corrected compensation matrix to recalculate and gating the flow cytometry analysis signal of the sample.

[0023] This process involves real-time optimization of the compensation matrix to accommodate differences in fluorescence characteristics in individual samples. For example, some individuals may have a high background of red blood cells or platelets, leading to significant signal fluctuations. In such cases, by comparing the individual's compensation value with the population average, the system can automatically adjust the compensation parameters to improve data reliability. This corrected compensation matrix allows for more precise regional segmentation of the target white blood cell subpopulation, ensuring accurate sorting. Specifically, in practice, when the fluorescence signal in certain channels of an individual sample shows a shift, the algorithm can be used for local adjustments to bring the entire dataset closer to the ideal state.

[0024] In one specific embodiment, after obtaining the initial multicolor fluorescence compensation matrix, specific cell subpopulations (such as lymphocytes) in the current individual sample are combined in key fluorescence channels (e.g., FITC vs. The median signal distribution on the PE channel is compared in real time with a pre-established database of "preset population baselines" containing a large amount of data from healthy or typical populations. If a systematic shift is found in the specific combination of fluorescence signals in the individual sample (such as the median fluorescence intensity of CD4-PE) compared to the population baseline (e.g., the overall shift exceeds a preset 15% threshold), it is determined that the leakage coefficient of the initial compensation matrix on that channel may not be applicable to this sample. Subsequently, based on the direction and magnitude of the shift, the corresponding compensation parameters in the initial compensation matrix are dynamically fine-tuned at a preset step size (e.g., by 5% of the shift) (e.g., reducing the compensation value k_{PE-FITC} from the PE channel to the FITC channel), generating a corrected compensation matrix specific to the current individual sample. Finally, this corrected compensation matrix is ​​applied to all cellular events of the sample for signal recalculation and compensation, and based on the compensated and more accurate fluorescence signals, "virtual clustering" and "gate identification" of cell subpopulations are re-performed on the flow cytometry scatter plot (e.g., accurately distinguishing CD4+ T cells from CD8+ T cells), thereby providing accurate target cell population identification for subsequent centrifugation parameter mapping.

[0025] In another embodiment, after acquiring the initial multicolor fluorescence compensation matrix, the median signal intensity of a cell subpopulation present in all current individual samples in the fluorescence channel is identified. Based on the deviation of this median signal intensity from the population reference baseline, all compensation coefficients in the initial compensation matrix involving that channel are proportionally scaled and corrected. That is, after acquiring the initial compensation matrix, the system does not immediately apply it to all cells. Instead, it first identifies a known, stable, and clearly present "anchored cell subpopulation" in the current individual sample (e.g., a positive signal cluster corresponding to CD45 molecules highly expressed in all nucleated leukocytes). Subsequently, the system compares the median signal intensity of the "anchored subpopulation" in the sample in key fluorescence channels (e.g., the median fluorescence intensity MFI of CD45-PerCP) with a pre-established "population reference baseline" containing a large amount of data from healthy individuals. If a systematic shift is found in the anchored subpopulation MFI of the individual compared to the population reference (e.g., shift ΔMFI > 15%), it is determined that the signal baseline drift is caused by instrument status or sample-specific factors (e.g., hemoglobin interference). At this point, the system will dynamically scale and correct all compensation coefficients related to the channel in the initial compensation matrix according to the proportion of ΔMFI (for example, if the PerCP channel signal is 20% stronger overall, the compensation value of other dyes to the PerCP channel will be reduced accordingly), thereby generating an individual-specific corrected compensation matrix, and applying this matrix to perform signal recalculation and precise gating.

[0026] In another embodiment, after generating the initial multicolor fluorescence compensation matrix, an event confidence matrix is ​​created in parallel. This event confidence matrix is ​​used to evaluate the Mahalanobis distance of each cell event in the compensated space after initial compensation. If cell events exceeding a predetermined threshold exhibit Mahalanobis distances exceeding the threshold, a backup compensation matrix is ​​generated as the corrected matrix. In other words, after generating the initial compensation matrix, the system creates an "event confidence matrix" in parallel. This matrix is ​​not used to calculate fluorescence intensity, but rather to evaluate the mathematical "confidence" of the signal combination of each cell event after initial compensation. The principle is that normal, single-positive cell populations will cluster in specific coordinate regions in the compensated multidimensional fluorescence space. The system calculates the Mahalanobis distance of each cell event to these known, stable cell populations (such as lymphocytes and monocytes) in the compensated space. If a large number of cell events exhibit abnormally large Mahalanobis distances (i.e., become "outliers"), it indicates that the initial compensation matrix may be severely inapplicable to the individual sample, causing "distortion" of the cell population in multidimensional space. At this point, the system will not directly modify the optical compensation coefficient, but will trigger a protection mechanism: either prompting the operator to recheck the sample and control, or starting a more conservative backup compensation matrix based on robust statistics as the corrected matrix, thereby ensuring that cell subpopulation gating is performed in a reliable and undistorted data space.

[0027] S104: The specific physical sorting intervals rich in the target leukocyte subsets, determined after signal recalculation and gating, are mapped to density gradient centrifugation conditions. At this stage, it is necessary to clarify the physical properties of the leukocyte subsets exhibited during density gradient centrifugation and design centrifugation parameters accordingly, such as centrifugal force, centrifugation time, and temperature. Different types of leukocytes have varying buoyancy densities due to differences in cytoplasm, cell membrane, and internal structure, allowing them to form clear interface layers in centrifugation media of different densities. For example, human peripheral blood mononuclear cells (PBMCs) are commonly used in density gradient centrifugation, using Ficoll-Paque solution to separate lymphocytes and monocytes, while other cell types sink to the bottom. This method allows for high-purity cell separation.

[0028] For example, in one embodiment, after completing signal recalculation and virtual gatening based on the corrected compensation matrix, the system locks the specific physical sorting interval (i.e., an electron gate with a specific range of FSC and SSC signal values) occupied by the target leukocyte subset (such as CD4+ T cells) on the scatter plots of front-scattered light (FSC-A) and side-scattered light (SSC-A) during flow cytometry. Subsequently, the physical parameters of this "electron gate" (mainly the ratio of the median of FSC-A to the median of SSC-A or its position vector in a preset FSC / SSC two-dimensional coordinate system) are correlated with density gradient centrifugation conditions: specifically, this range of physical parameters is mapped to the specific centrifugal force (e.g., 400×g to 600×g), centrifugation time (e.g., 20 to 30 minutes), and density of the separation solution used (e.g., choosing 1.077 g / mL Ficoll solution instead of 1.119 g / mL) required for performing density gradient centrifugation. The Percoll solution (g / mL) provides precise instructions for the next step of physical separation, ensuring that after centrifugation, the target cell subpopulation is enriched in a specific interfacial layer of the separation solution due to its unique physical density and size.

[0029] S105: Using the aforementioned correlation mapping, the immunolabeled blood sample is directly subjected to density gradient centrifugation, thereby achieving high-purity separation of the target leukocyte subset based on cell physical density. In this step, the data from the completed fluorescent labeling and flow cytometry analysis are used as input, combined with the pre-established density gradient centrifugation mapping relationship, to guide the specific centrifugation procedure. This method not only reduces errors caused by human judgment but also ensures that the target cells are maximized enriched during centrifugation. A typical example is the separation of tumor-infiltrating T cells (TILs) in experiments. Flow cytometry results are used to find a suitable density cutoff point, and then the gradient is set according to this point during centrifugation, thereby obtaining highly purified TILs in a single operation. This method significantly improves separation efficiency while ensuring cell viability and function.

[0030] The applicant recognizes that differences in CD molecule expression levels between individuals lead to subtle but crucial changes in their physical density. Traditional, fixed centrifugation protocols cannot adapt to these dynamic changes. This application utilizes high-precision flow cytometry information to guide and optimize coarse centrifugation separation to achieve centrifugation purity. Specifically, through "dynamic compensation" and "signal recalculation / virtual gating," more precise "position" information about the target cells in physical space (e.g., forward scattering / side scattering maps) is obtained. Then, the results of this "virtual sorting" are "correlatedly mapped" or "transformed" into specific centrifugation parameters, establishing an intelligent bridge from the "optical information domain" to the "physical separation domain." Furthermore, this application uses centrifugation, a traditional technique, as the final high-purity separation method, overcoming the traditional notion that "centrifugation cannot achieve high-purity leukocyte subpopulation separation." In summary, the proposed solution achieves near-flow cytometry (FACS) purity while retaining the advantages of centrifugation—high throughput, low cost, and minimal damage to cell viability.

[0031] Next, the steps of the present invention are described, which involve dynamically adjusting and generating an individual-specific corrected compensation matrix based on the degree of deviation of the initial compensation matrix within the reference range of the sample population.

[0032] First, the average signal intensity ΔI of each fluorescence channel in the current sample is obtained, which is the difference between the current sample and the population reference value. ΔI reflects the abnormal signal level of an individual sample in a specific fluorescence channel, and its value directly affects the accuracy of the compensation effect.

[0033] Secondly, the standard deviation σ of ΔI relative to the population reference value is calculated, where σ represents the magnitude of normal signal variation among samples. Quantifying the degree of variation through standard deviation allows for an objective assessment of the authenticity of signal deviations. Then, based on the formula Δσ = |ΔI - μ| / σ, it is determined whether the normal fluctuation threshold is exceeded. μ represents the population reference value, representing the expected value of the normal signal; Δσ characterizes the relative deviation of an individual sample from the population average level, with a larger value indicating a more significant deviation. If Δσ is greater than Tth (Tth is recommended to be set to 2.5, indicating exceeding 99% of the normal range), the correction matrix is ​​updated to accommodate sample-specific changes; otherwise, the initial matrix remains unchanged. This formula design is based on the Z-score principle in statistics, effectively distinguishing between systematic errors and random fluctuations. For example, in a precise blood leukocyte separation method combining immunodensity and centrifugation techniques, if an atypical enhancement of the fluorescence signal in a sample occurs, Δσ will significantly increase, triggering a compensation matrix update to ensure signal consistency among different samples and improve separation accuracy. This dynamic adjustment method can reduce the impact of inter-individual biological differences on the results, improving measurement stability and experimental repeatability.

[0034] Next, refer to Figure 2 The present invention describes the steps of recalculating the flow cytometry analysis signal of the sample and gating the cell subpopulation using the corrected compensation matrix.

[0035] S201: Recalculate the normalized fluorescence intensity N_i = (I_i - minI) / (maxI - minI) for each event point, where I_i is the original fluorescence intensity of the event point, minI is the minimum fluorescence intensity of all events in that channel, and maxI is the maximum fluorescence intensity. N_i ranges from [0,1], with an optimal value close to 1 indicating a high fluorescence intensity for that event. This step aims to standardize the fluorescence signal scale across different samples for easier subsequent analysis.

[0036] S202: Compare N_i with a pre-set multicolor fluorescent labeling threshold to screen target cells. For example, in one embodiment, when detecting T cells, an anti-CD3 fluorescent label is used, and the preset threshold is set to 0.75, meaning that only when N_i exceeds 0.75 is it initially considered a T cell. This step can effectively filter out non-target cells.

[0037] S203: Calculate the signal deviation coefficient S = |N_i - T_i| / T_i according to the following formula, where T_i is the expected ideal normalized intensity, representing the theoretical target fluorescence signal intensity, and S ranges from [0, ∞), with an optimal value close to 0. If S < ε (ε is usually set to 0.1 or 0.05), the event is considered to belong to the target subgroup. The purpose of this formula is to quantify the degree of deviation between the actual signal and the ideal signal, and to ensure higher accuracy of the results by setting a small upper limit for the allowable deviation coefficient.

[0038] In a precise blood leukocyte separation method combining immunodensity and centrifugation techniques, specifically, when analyzing the centrifuged single-cell suspension by flow cytometry, a corrected compensation matrix is ​​used to correct for multicolor fluorescence overlap effects. Subsequently, each cell undergoes normalized fluorescence processing, is compared with a set threshold, and its identification as a target cell is assessed using the signal deviation coefficient. This method effectively improves the accuracy and reproducibility of flow cytometry data, reduces false positives or false negatives, and enhances the reliability and stability of leukocyte subset classification.

[0039] Next, refer to Figure 3 This invention describes a technical solution for associating and mapping a specific physical sorting interval rich in the target leukocyte subpopulation, determined after signal recalculation and gating, with density gradient centrifugation separation conditions.

[0040] S301: The mean density value D_mean of the target leukocyte subset. D_mean represents the theoretical sedimentation position of the target cells during density gradient centrifugation. This value is usually obtained through pre-set biophysical data or experimental tests, and the range is generally from 1.06 g / cm³ to 1.12 g / cm³. The optimal value depends on the specific leukocyte type. For example, the mean density of neutrophils is close to 1.075 g / cm³.

[0041] S302: The ideal centrifugation gradient range G = [D_mean - ΔD, D_mean + ΔD] is calculated based on the empirical formula, where ΔD represents the allowable density deviation range, typically ±0.01 g / cm³. This range ensures that most target cells can still be captured even with individual differences.

[0042] S303: Match this interval with the interlayer interface in the density gradient column. This means aligning the expected density location of the cells after centrifugation with the layered structure of the experimental setup, so that the target cells are precisely located in a specific layer.

[0043] S304: By controlling the centrifugation speed (V) and time (t), effective sorting of corresponding density regions can be achieved. The typical range for V is 400 to 800 r / min, and t is usually set to 20 to 30 minutes. This ensures efficient enrichment of target cells while excluding other non-target components. For example, in the process of isolating memory T cells, D_mean can be set to 1.08 g / cm³, G can be set to [1.07, 1.09], and centrifugation parameters can be configured according to this range to ensure efficient acquisition of high-purity memory T cells. This correlation mapping method helps improve sorting efficiency and target cell purity, reduces non-specific contamination, and significantly improves the effectiveness of subsequent research or clinical applications.

[0044] In one specific embodiment, after completing signal recalculation and gating and identifying the target leukocyte subset (e.g., CD8+ cytotoxic T cells), the average physical density value D_mean (e.g., 1.075 g / mL) of the target subpopulation is first identified and calculated based on the typical distribution of this subset in flow cytometry forward scatter (FSC-A, related to cell size) and side scatter (SSC-A, related to intracellular complexity / granularity) data, combined with a pre-stored empirical database. Subsequently, the ideal centrifugation gradient range is calculated according to the preset empirical formula G = [D_mean - ΔD, D_mean + ΔD], where the allowable density deviation range ΔD is set to 0.005 g / mL according to the target purity and recovery requirements, thus obtaining the target cell enrichment density range G of [1.070, 1.080] g / mL. Next, this target density range is compared with the density gradient separation solution used (e.g., Ficoll-Paque PLUS, density 1.077 g / mL). Precise matching is performed at a specific interlayer interface (g / mL) to determine that the target cells will be enriched above the separation medium interface. Finally, by controlling the centrifuge speed V (e.g., calculated and set to 400 × g) and centrifugation time t (e.g., set to 30 minutes according to Stokes' law and empirical data), cells of different densities in the sample migrate and stabilize at the separation medium layer corresponding to their own density in the centrifugal force field, thereby achieving the purpose of effectively separating the target CD8+ T cells from the specific density region.

[0045] Next, refer to Figure 4 This describes the steps of the present invention to perform density gradient centrifugation on immunolabeled blood samples using correlation mapping.

[0046] S401: The centrifugation rate v is set using the formula v = k × m^α, which is based on cell mass, where m is the white blood cell mass, k is a proportionality constant, and α is a power exponent. This formula aims to adjust the appropriate centrifugation speed according to different white blood cell masses, ensuring precise separation of white blood cells within a density gradient. The range of parameter m is determined based on the specific cell type, typically between 1 × 10^-12 kg and 5 × 10^-12 kg. α generally ranges from 0.5 to 1.5, with an optimal value of approximately 1.0, ensuring a linear change in centrifugation rate with cell mass. The value of k needs to be experimentally calibrated to match equipment performance and sample characteristics.

[0047] S402: Record the sample liquid volume V_sample before each centrifugation. This is to ensure the accuracy of subsequent fill rate calculations.

[0048] S403: Calculate the required centrifuge tube filling rate F = (V_selected / V_cap) using the following formula, where V_selected is the theoretical volume of the selected range, and V_cap is the maximum capacity of the centrifuge tube. F represents the degree to which the centrifuge tube is filled, and should be controlled between 0.3 and 0.8, with an optimal value close to 0.6, to avoid overflow and reduce mixing effects. If F ≤ F_max (F_max is usually set to 0.8), then perform the centrifugation step; otherwise, reduce the sample volume before proceeding. This ensures both centrifugation effectiveness and equipment safety.

[0049] For example, in one embodiment, when processing approximately 2 mL of immunolabeled blood sample, if the centrifuge tube capacity is 4 mL, then according to the preset value of V_selected (2.4 mL), F = 2.4 / 4 = 0.6 is calculated, which is less than the upper limit of 0.8, allowing direct centrifugation. This operation method can improve the separation efficiency and purity of leukocytes, while avoiding uneven centrifugation or equipment damage caused by overfilling. This technical solution improves the controllability and repeatability of separation by dynamically adjusting centrifugation parameters, thereby enhancing the reliability of clinical applications.

[0050] Next, refer to Figure 5 This describes the steps of the present invention to dynamically adjust and generate an individual-specific corrected compensation matrix based on the degree of deviation of the initial compensation matrix within the reference range of the sample population.

[0051] S501: Perform independent standard deviation analysis for each fluorescent label to obtain the standard deviation vector σ_i. This step quantifies the fluctuation of each fluorescent signal in the sample population, where σ_i represents the standard deviation of the i-th fluorescent label. The numerical range is determined by the experimental data and is usually non-negative. Its magnitude reflects the dispersion of the label signal. The larger the standard deviation, the more significant the inter-individual variation. For example, in the process of separating blood leukocytes, the CD4 or CD8 fluorescent signals of different individuals may exhibit different standard deviations.

[0052] S502: Construct the adjustment coefficient R = α × Σ(σ_i) / max(σ) according to the following formula, where α is a weighting factor. The value of α is generally in the range of [0,1], used to balance the influence of overall fluctuations. Σ(σ_i) is the sum of the standard deviations of all fluorescence signals, while max(σ) represents the maximum standard deviation. This formula means to comprehensively evaluate the overall deviation of all fluorescence channels and calculate the adjustment magnitude in a proportional form to ensure that the adjustment is representative. For example, when multiple fluorescence signals of a sample generally deviate from the reference range, this coefficient will amplify the adjustment.

[0053] S503: Incorporate R into the adjustment factor of the final correction matrix. This means that the adjustment coefficient, as a correction parameter, affects the optimization result of the subsequent compensation matrix. If R > R_threshold, where R_threshold is a significance threshold set based on statistical or empirical values ​​(such as 0.5 or 1.0), then correction is performed; otherwise, the original compensation matrix is ​​retained. This step avoids unnecessary adjustments, reduces over-correction caused by small fluctuations, and improves the stability and reliability of the algorithm.

[0054] For example, after establishing an initial fluorescence compensation matrix based on a single positive control sample, the system performs independent statistical analysis on the fluorescence intensity values ​​of all positive events in the main detection channel for each fluorescent marker (e.g., CD3-FITC, CD4-PE, CD8-APC) based on the measured data of the current individual sample. This yields the standard deviation vector σ_i, characterizing the dispersion of the marker signal distribution (e.g., σ_FITC = 5200, σ_PE = 8100, σ_APC = 4500). Subsequently, a comprehensive adjustment coefficient is constructed according to the formula R = α × Σ(σ_i) / max(σ), where the weighting factor α is empirically set to 0.5, and max(σ) is the maximum value among all standard deviations (i.e., σ_PE = 8100). The calculated R = 0.5 × (5200 + 8100 + 4500) / 8100 ≈ 1.10; This adjustment factor R is incorporated into the generation process of the final correction matrix. For example, each compensation factor k_ij in the initial compensation matrix is ​​multiplied by a factor related to R (such as the square root of R √1.10≈1.05) for fine-tuning. Finally, the calculated R value is compared with a preset significance threshold R_threshold (e.g., 1.05). Since 1.10 > 1.05, the adjustment condition is met. Therefore, the system will apply this adjusted, individual-specific corrected compensation matrix to recalculate all data of the current sample. Conversely, if R ≤ R_threshold, the original compensation matrix is ​​retained without modification.

[0055] Next, refer to Figure 6 The present invention describes the steps of associating and mapping a specific physical sorting interval rich in the target leukocyte subpopulation, determined by signal recalculation and gating, with density gradient centrifugation separation conditions.

[0056] S601: Establish migration curves M(v) of the target cell subpopulation under different centrifugal forces. The purpose of this step is to obtain the migration characteristics of the target cells at various centrifugation rates through experimental measurement or simulation calculation, and to reflect their distribution in the form of a function. Here, v represents the centrifugation rate in revolutions per minute (rpm). M(v) reflects the cell's migration ability in density gradient centrifugation, and its numerical range depends on the experimental conditions and cell type. For example, in one embodiment, assuming the target cells are CD14+ monocytes, M(v) first increases and then decreases with increasing v, indicating the existence of optimal centrifugation conditions. This function is used for subsequent parameter optimization.

[0057] S602: The optimal sorting region is predicted using a fitting polynomial M(v) = a×v³ + b×v² + c×v + d. This polynomial has high-order terms, enabling a more accurate fit to actual experimental data. Here, 'a' is typically negative, giving the curve an inverted parabolic shape; 'b' is the second-highest parameter, controlling the inflection point of the curve; 'c' and 'd' represent the linear and constant parts, respectively, adjusting the overall distribution trend. Parameter values ​​depend on the specific migration behavior of the target cells and must be obtained using fitting methods such as least squares.

[0058] S603: Select the optimal centrifugation rate v_opt = argmax{M(v)} according to the following formula, that is, find the v that maximizes M(v). This step aims to mathematically determine the centrifugation rate that maximizes the migration curve as the optimal sorting condition.

[0059] S604: Use v_opt as the actual centrifugation parameter for the sorting operation.

[0060] The maximum value of M(v) represents the location where the target cells are most concentrated in the density gradient, improving separation efficiency. For example, in a specific case, it was calculated that v_opt = 3500 rpm, corresponding to M(v) = 95%, indicating that the recovery rate of mononuclear cells reaches its highest at this rate. Ultimately, v_opt is used as the actual centrifugation parameter for the sorting operation.

[0061] The beneficial effects of this technical solution are that it enables precise customization of centrifugation parameters, optimizes the separation process by combining cell characteristics and experimental data, improves sorting purity and efficiency, reduces repetitive operations, and shortens the experimental cycle.

[0062] Next, refer to Figure 7 This describes the steps of the present invention to perform density gradient centrifugation on immunolabeled blood samples using correlation mapping.

[0063] S701: Set the centrifuge speed V_set and sample filling height H_set. V_set is the centrifuge rotation speed, usually expressed in revolutions per minute (rpm). Its value range varies depending on the target cell type, generally between 500-1500 rpm, with the optimal value adjusted based on sample type and cell density. H_set is the sample height within the centrifuge tube, typically recommended to be between one-third and one-half full to ensure uniform liquid flow and avoid vigorous mixing during centrifugation.

[0064] S702: The sorting efficiency is evaluated using the following formula: Efficiency = (V_output × t_eff) / (V_input × t_total), where V_output is the volume of successfully sorted cells, t_eff is the effective processing time in seconds or minutes, V_input is the original sample volume, and t_total is the total experimental time. Efficiency measures the system's ability to efficiently separate target cells from the sample. By setting Eff_threshold, it can be determined whether the current experiment has achieved the ideal separation effect, ensuring that the operation is controllable and the results are stable.

[0065] S703: If Efficiency ≥ Eff_threshold, the process is considered controllable. This criterion ensures that the separation process is executed within preset standards, reducing invalid experiments and improving the success rate. For example, in one embodiment, CD34+ positive hematopoietic stem cell separation is performed on human peripheral blood samples. When V_set is set to 1000 rpm and H_set to 1.5 cm, F ≈ 230 g is calculated, and the Efficiency is found to be 85% using the formula, exceeding the 75% set for Eff_threshold. This result indicates that under the set conditions, the sample can be effectively separated.

[0066] Combining immunodensity and centrifugation techniques can significantly improve the accuracy and stability of leukocyte separation, reduce human intervention, and enhance the consistency and reproducibility of sample processing, making it suitable for large-scale cell analysis and applications in scientific research and clinical fields.

[0067] Next, refer to Figure 8 This describes the steps of the present invention to perform density gradient centrifugation on immunolabeled blood samples using correlation mapping.

[0068] S801: Set the maximum centrifugation pressure P_max based on the current centrifuge performance. This parameter represents the highest pressure value that the centrifuge can provide, and is usually set in the range of 5000 to 15000 rpm, subject to the specific equipment parameters, to prevent equipment overload and cell rupture.

[0069] S802: Calculate the cell viability index A_index = (A_after / A_before) × 100% based on the cell viability test indicators before and after centrifugation. Where A_before is the cell viability before centrifugation and A_after is the cell viability after centrifugation. This index reflects the degree of cell damage, and it is generally recommended to be no less than 85%.

[0070] S803: Define the critical cell viability function C = A_index × (1 + λ × (F_actual - F_design)²), where λ is the pressure sensitivity factor, typically ranging from 0.1 to 0.5; F_design is the standard centrifugal force, set to 1000–2000 rpm; and F_actual is the actual applied centrifugal force. This formula indicates the degree to which cell viability is affected by pressure when the actual centrifugal force deviates from the standard value.

[0071] S804: If C ≥ C_threshold (C_threshold is the minimum survival requirement, such as 80%), then execute the centrifugation procedure. If the condition is not met, adjust the parameters and try again.

[0072] For example, in one embodiment, assuming a batch of samples is immunolabeled and P_max is set to 12000 rpm, with A_after being 90% and A_before being 95%, then A_index is 94.7%. If F_actual is set to 11000 rpm, the difference from F_design is 1000 rpm, and λ is 0.3, substituting into the formula, we get C approximately 94.7 × (1 + 0.3 × 1000²) = 94.7 × (1 + 3000000) ≈ 2841000. This value is clearly higher than C_threshold, therefore centrifugation can be safely performed. This process ensures cell integrity and improves the efficiency and accuracy of leukocyte separation by precisely adjusting centrifugation parameters.

[0073] The present invention provides a precise blood leukocyte separation method combining immunodensity and centrifugation techniques, comprising: firstly, multicolor immunofluorescence labeling of leukocyte subsets in a blood sample, using specific antibodies to label target cells; subsequently, flow cytometry analysis of the labeled sample to obtain real-time fluorescence signal data, and establishing an initial multicolor fluorescence compensation matrix for the current individual sample; to improve the accuracy of subsequent analysis, the system dynamically adjusts and generates an individual-specific corrected compensation matrix based on the deviation of the initial compensation matrix from the reference range of the sample population; then, the corrected compensation matrix is ​​used to recalculate the flow cytometry signal and perform cell subset gating to accurately identify regions rich in target leukocyte subsets; further, the specific physical sorting intervals determined after signal recalculation and gating are correlated and mapped with density gradient centrifugation conditions to achieve precise matching; finally, density gradient centrifugation is performed directly using this correlation mapping, relying on the different physical density characteristics of cells to achieve high-purity, high-throughput separation of target leukocyte subsets. This invention organically combines the high sensitivity of immunoassay with the high purity of density centrifugation technology, significantly improving the separation efficiency and purity of specific leukocyte subsets while maintaining high cell viability. It effectively solves the technical bottleneck problems caused by strong cell heterogeneity, low sorting efficiency and impaired cell viability in traditional methods.

[0074] The above description is the preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this invention, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for precise separation of blood leukocytes combining immunodiffusion and centrifugation techniques, characterized in that, Includes the following steps: Multicolor immunofluorescence labeling of leukocyte subsets in blood samples; The labeled samples were analyzed by flow cytometry, and an initial multicolor fluorescence compensation matrix was established for the current individual sample based on the real-time fluorescence signal data obtained from the analysis. Based on the deviation of the initial compensation matrix from the reference range of the sample population, a correction compensation matrix specific to the individual is dynamically adjusted and generated, and the correction compensation matrix is ​​used to recalculate and gating the flow cytometry analysis signal of the sample. The specific physical sorting interval rich in the target leukocyte subpopulation, determined after signal recalculation and gatening, is correlated and mapped with the density gradient centrifugation separation conditions. Using the aforementioned association mapping, the immunolabeled blood sample is directly subjected to density gradient centrifugation, thereby achieving high-purity separation of the target leukocyte subset based on cell physical density.

2. The method for precise separation of blood leukocytes combining immunodensity and centrifugation techniques according to claim 1, characterized in that, After obtaining the initial multicolor fluorescence compensation matrix, the median signal intensity of the cell subpopulation present in the current individual sample in the fluorescence channel is identified, and all compensation coefficients involving that channel in the initial compensation matrix are proportionally scaled and corrected based on the degree of deviation of the median signal intensity from the population reference baseline.

3. The method for precise separation of blood leukocytes combining immunodiffusion and centrifugation techniques according to claim 1, characterized in that, After generating the initial multicolor fluorescence compensation matrix, an event confidence matrix is ​​created in parallel. The event confidence matrix is ​​used to evaluate the Mahalanobis distance of each cell event in the compensated space after the initial compensation. If a cell event exceeds a predetermined threshold and exhibits a Mahalanobis distance exceeding the threshold, a backup compensation matrix is ​​generated as the corrected matrix.

4. The method for precise separation of blood leukocytes combining immunodiffusion and centrifugation techniques according to claim 1, characterized in that, The process of recalculating the flow cytometry signal of the sample and gating the cell subpopulation using the corrected compensation matrix includes: Recalculate the normalized fluorescence intensity N_i = (I_i-minI) / (maxI-minI) for each event point, where I_i is the original fluorescence intensity of the event point, minI is the minimum fluorescence intensity of all events in the channel, and maxI is the maximum fluorescence intensity. Compare N_i with a pre-set multicolor fluorescent labeling threshold; The signal deviation coefficient S is calculated using the following formula: S = |N_i-T_i| / T_i, where T_i is the expected ideal normalized intensity. If S < ε, where ε is the upper limit of the allowable deviation coefficient, then the event is considered to belong to the target subgroup.

5. The method for precise separation of blood leukocytes combining immunodiffusion and centrifugation techniques according to claim 1, characterized in that, The step of mapping the specific physical sorting interval rich in the target leukocyte subpopulation, determined after signal recalculation and gating, to the density gradient centrifugation separation conditions includes: The mean density value D_mean of the target leukocyte subset; The ideal centrifugal gradient range G = [D_mean-ΔD, D_mean + ΔD] is calculated using the empirical formula, where ΔD represents the allowable density deviation range. Match this interval with the interlayer interface in the density gradient column; By controlling the centrifugation speed V and time t, effective sorting of corresponding density regions can be achieved.

6. The method for precise separation of blood leukocytes combining immunodiffusion and centrifugation techniques according to claim 1, characterized in that, The step of directly performing density gradient centrifugation on the immunolabeled blood sample using the association mapping includes: The centrifugation rate v is set as a calculation formula based on cell mass size: v = k × m^α, where m is the white blood cell mass, k is a proportionality constant, and α is a power exponent. Record the sample liquid volume V_sample before each centrifugation; The required centrifuge tube filling rate F = (V_selected / V_cap) is calculated according to the following formula, where V_selected is the theoretical volume of the selected range and V_cap is the maximum capacity of the centrifuge tube; If F ≤ F_max, where F_max is the design fill limit, then centrifugation is performed; otherwise, the sample size is reduced before proceeding.

7. The method for precise separation of blood leukocytes combining immunodiffusion and centrifugation techniques according to claim 1, characterized in that, The process of dynamically adjusting and generating an individual-specific corrected compensation matrix based on the deviation of the initial compensation matrix from the reference range of the sample population includes: Perform independent standard deviation analysis on each fluorescent label to obtain the standard deviation vector σ_i; The adjustment coefficient R is constructed according to the following formula: R = α × Σ(σ_i) / max(σ), where α is the weighting factor; R is incorporated into the adjustment factor of the final correction matrix; If R > R_threshold, where R_threshold is the preset significance threshold, then the adjustment is applied; otherwise, the original compensation matrix is ​​retained.

8. The method for precise separation of blood leukocytes combining immunodensity and centrifugation techniques according to claim 7, characterized in that, The step of mapping the specific physical sorting interval rich in the target leukocyte subpopulation, determined after signal recalculation and gating, to the density gradient centrifugation separation conditions includes: Establish migration curves M(v) of the target cell subpopulation under different centrifugal forces; The optimal sorting region is predicted using the fitted polynomial M(v) = a×v³ + b×v² + c×v + d; The optimal centrifugation rate is selected according to the following formula: v_opt = argmax{M(v)}, that is, to find the v that maximizes M(v); Use v_opt as the actual centrifugation parameter for the sorting operation.

9. The method for precise separation of blood leukocytes combining immunodensity and centrifugation techniques according to claim 8, characterized in that, The step of directly performing density gradient centrifugation on the immunolabeled blood sample using the association mapping includes: Set the centrifuge speed V_set and the sample filling height H_set; The sorting efficiency is evaluated using the following formula: Efficiency = (V_output × t_eff) / (V_input × t_total), where V_output is the volume of successfully sorted cells, t_eff is the effective processing time, V_input is the original sample volume, and t_total is the total time consumed in the entire experiment. If Efficiency ≥ Eff_threshold, the process is considered controllable.

10. The method for precise separation of blood leukocytes combining immunodensity and centrifugation techniques according to claim 9, characterized in that, The step of directly performing density gradient centrifugation on the immunolabeled blood sample using the association mapping includes: Based on the current performance of the centrifuge, set the maximum centrifugal pressure that it can withstand, P_max. Based on the cell viability detection indicators before and after centrifugation, the cell viability index A_index = (A_after / A_before) × 100%, where A_before is the cell viability before centrifugation and A_after is the cell viability after centrifugation. The critical function for cell activity is defined as C = A_index × (1+λ × (F_actual-F_design)²), where λ is the pressure sensitivity factor and F_design is the standard centrifugal force. If C ≥ C_threshold, where C_threshold is the minimum survival requirement, then the centrifugation procedure is executed.

Images