Blood cell analysis system based on image flow cytometry

Abstract

The invention belongs to the technical field of biomedical detection, and discloses a blood cell analysis system based on image flow cytometry, which comprises a cell staining module, the dyeing module simultaneously comprises a first fluorescent probe capable of being specifically combined with components in a cell nucleus or enriched in the cell nucleus, a second fluorescent probe capable of being specifically combined with cell nucleus DNA and a third fluorescent probe capable of being specifically combined with a blood cell differentiation antigen; an image flow cytometer; and the image processing module is used for calculating the kernel detection rate and / or the average kernel number. According to the present invention, the combination of the specific types of the cell staining dyes is adopted, the image flow cytometry is matched, and the optimized multi-dimensional fluorescence staining combination and the matched automatic image analysis process are adopted to achieve the synchronous, high-throughput and accurate quantitative analysis of the blood cell immune phenotype and the cell morphology key characteristics at the single cell level;

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of biomedical detection, and more particularly relates to a blood cell analysis system based on image flow cytometry, which utilizes image flow cytometry and multi-dimensional fluorescence staining technology to perform synchronous high-throughput and quantitative analysis on the immunophenotype and cell morphological characteristics of a single blood cell, and is particularly suitable for the auxiliary diagnosis of blood system malignancies characterized by the proliferation of primitive cells and the alteration of nucleoli, such as but not limited to acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL), and high-risk types of myelodysplastic syndrome (MDS). BACKGROUND

[0002] Common clinical blood detection methods include blood smears and flow cytometry. Morphological examination of blood cells is the cornerstone of blood disease diagnosis. With the development of technology, new blood analyzers can perform semi-automatic morphological classification of cells in blood smears in combination with automatic analysis and artificial intelligence (AI) algorithms [1][2] , but there are two inherent technical bottlenecks that restrict it from becoming the most ideal diagnostic solution. One is that the information dimension is single, and the "gold standard" data source of the AI model has problems of reliability and explainability [3] . The bone marrow blood smear cell 23 classifier DeepHeme developed by the team of Professor Gregory Goldgof of the University of California, San Francisco [4] , the fully automatic cell morphological analyzer MC-100i of Mindray [5] , the DI-60 of the Japanese Sysmex company, and the DM-96 of the Swedish CellaVision company, and the Sight OLO Analyzer of the Israeli Scopio Labs [6] , which are representative of automatic blood smear morphological analyzers, still rely on two-dimensional fixed blood smear microscopic pictures. Blood smear preparation is an end-point detection, and only has morphological single-dimensional information. The annotation process of the AI model corresponding to the blood smear lacks cross-validation of indicators such as immunophenotype, resulting in limited explainability and reliability [7][8][9] . The second is low detection throughput and insufficient sensitivity. The low detection throughput of blood smear analysis results in a cell sample amount of usually only 100-500, resulting in a high coefficient of variation (CV) and low sensitivity of detection

[10] , which makes it difficult to effectively detect rare abnormal cells such as MRD monitoring and early diagnosis.

[0003] High-throughput flow cytometry is another commonly used instrument for existing clinical peripheral blood detection, which performs outstandingly in high-throughput (>10,000 cells / second) and high-parameter immunotyping (>40 colors), and is the cornerstone of immunological research and diagnosis

[11]

[12] The instrument can indirectly infer cell size and granularity through forward / side scatter light (FSC / SSC), and cannot obtain the "gold standard" morphological information for blood disease diagnosis such as cell area, cell roundness and shape, nuclear morphology, nuclear area, nuclear-cytoplasmic ratio, chromatin density, cytoplasmic RNA content, cytoplasmic content, cell granularity, whether nuclear shift occurs, nucleolus number and average nucleolus number, and cannot meet the comprehensive requirements of clinical diagnosis.

[0004] Primitive cells in peripheral blood can be used for diagnosis of leukemia and other blood diseases

[13]

[14] In cell morphology, the nucleolus, as the center of ribosome biosynthesis, its number, size and morphology are a direct reflection of cell proliferation activity, and are also a classic pathological feature of tumor cells

[15]

[16] In blood system diseases such as AML

[17] , ALL [7] and high-risk types of MDS

[18] , primitive cells, as immature cells with rapid proliferation, need a large amount of synthesized ribosomes and proteins, so their nucleoli are usually very active in metabolism, and morphologically show one or more clear and obvious nucleoli, which can be used as one of the characteristics of identifying primitive cells

[19] .

[0005] In summary, although morphological diagnosis of blood smears and immunological diagnosis of flow cytometry are outstanding in single morphological and immunological dimensions respectively, blood smears have the problems of low throughput and inability to provide immunological information, and immunotyping has the problems of no morphological information and inability to realize cell-level correspondence with blood smears. This leads to the fact that single detection results cannot provide the comprehensive information required for clinical diagnosis, resulting in problems such as high cost, low efficiency and complex diagnostic criteria for cross-platform detection.

[0006] As a new generation of high-parameter flow cytometer, the image flow cytometer (IFC) combines the high-throughput analysis of flow cytometry with the high-resolution imaging of fluorescence microscopy for the first time, opening up a new dimension of cell analysis

[11] . The currently disclosed applications of IFC in hematology are mostly limited to specific and single-dimensional morphological analysis, such as using only DAPI and other DNA dyes for cell cycle or apoptosis analysis, or only observing the intracellular localization of a specific protein

[20] These methods fail to establish a comprehensive staining and analysis scheme to systematically and specifically 'translate' the morphological parameters focused on by traditional blood smear microscopy and combine them with immunological analysis to achieve integrated immunomorphological diagnosis. After searching, there is currently no solution to the problems of low throughput of blood smear morphological analysis, inability of individual cells to correspond one by one with immunological information, lack of morphological information in traditional flow cytometry, lack of a system in the field that can correlate immunophenotype with comprehensive and detailed blood cell pathological morphology in high-throughput and automated analysis, and other issues. In view of this, the present application develops a new technology for simultaneous high-throughput and quantitative analysis of the immunophenotype and morphological characteristics of individual blood cells using image flow cytometry and multidimensional fluorescence staining technology.

[0007] The reference list is as follows: [1] G. Zini, F. Mancini, E. Rossi, S. Landucci, G. d'Onofrio. Artificial Intelligence and the Blood Film: Performance of the Mc-80 Digital Morphology Analyzer in Samples with Neoplastic and Reactive Cell Types. International Journal of Laboratory Hematology. 2023, 45(6): 881-889 [2] Y. Xing, X. Liu, J. Dai, X. Ge, Q. Wang, Z. Hu, et al. Artificial Intelligence of Digital Morphology Analyzers Improves the Efficiency of Manual Leukocyte Differentiation of Peripheral Blood. BMC Medical Informatics and Decision Making. 2023, 23(1): [3] J. Laosai, K. Chamnongthai, editors. Deep-Learning-Based Acute Leukemia Classification Using Imaging Flow Cytometry and Morphology. 2018 international symposium on intelligent signal processing and communication systems (ISPACS); 2018: IEEE; [4] S. Sun, Z. Yin, J. G. Van Cleave, L. Wang, B. Fried, K. H. Bilal, et al. Deepheme, a High-Performance, Generalizable Deep Ensemble for Bone Marrow Morphometry and Hematologic Diagnosis. Science Translational Medicine. 2025, 17(802): [5] H. Jiang, W. Xu, W. Chen, J. He, H. Jiang, Z. Mao, et al. Performance of the Digital Cell Morphology Analyzer Mc-100i in a Multicenter Study in Tertiary Hospitals in China. Clinica Chimica Acta. 2024, 555: 117801 [6] N. Bachar, D. Benbassat, D. Brailovsky, Y. Eshel, D. Glück, D. Levner, et al. An Artificial Intelligence-Assisted Diagnostic Platform for Rapid near-Patient Hematology. 2021, 96(10): 1264-1274 [7]G. Yan, G. Mingyang, S. Wei, L. Hongping, Q. Liyuan, L. Ailan, etal. Diagnosis and Typing of Leukemia Using a Single Peripheral Blood Cellthrough Deep Learning. Cancer Science. 2024, 116(2): 533-543 [8]A. R. Ahmad, N. A. Zainudin, M. D. I. M. Radzi, N. H. A. Halim, M.K. Osman, Z. Saad. Cnn-Based Classification of Acute Myeloid Leukemia BloodSamples.2024 IEEE 14th International Conference on Control System, Computingand Engineering (ICCSCE)2024. p. 186-191 [9]J. E. Lewis, L. A. D. Cooper, D. L. Jaye, O. Pozdnyakova.Automated Deep Learning-Based Diagnosis and Molecular Characterization ofAcute Myeloid Leukemia Using Flow Cytometry. Modern Pathology. 2024, 37(1):

[10] C. Matek, S. Krappe, C. Münzenmayer, T. Haferlach, C. Marr.Highly Accurate Differentiation of Bone Marrow Cell Morphologies Using DeepNeural Networks on a Large Image Data Set. Blood. 2021, 138(20): 1917-1927

[11] M. Spasic, E. R. Ogayo, A. M. Parsons, E. A. Mittendorf, P. vanGalen, S. S. McAllister. Spectral Flow Cytometry Methods and Pipelines forComprehensive Immunoprofiling of Human Peripheral Blood and Bone Marrow.Cancer Research Communications. 2024, 4(3): 895-910

[12] J. R. Brestoff. Full Spectrum Flow Cytometry in the ClinicalLaboratory. International Journal of Laboratory Hematology. 2023, 45(S2): 44-49

[13] C. D. Godwin, Y. Zhou, M. Othus, M. M. Asmuth, C. M. Shaw, K. M.Gardner, et al. Acute Myeloid Leukemia Measurable Residual Disease Detectionby Flow Cytometry in Peripheral Blood Vs Bone Marrow. Blood. 2021, 137(4):569-572

[14] T. Kewan, W. S. Bahaj, C. Gurnari, O. D. Ogbue, S. Mukherjee, A.Advani, et al. Clinical and Molecular Characteristics of Extramedullary AcuteMyeloid Leukemias. Leukemia. 2024, 38(9): 2032-2036

[15] A. Corman, O. Sirozh, V. Lafarga, O. Fernandez-Capetillo.Targeting the Nucleolus as a Therapeutic Strategy in Human Disease. Trends inBiochemical Sciences. 2023, 48(3): 274-287

[16] M. Jo, S. Kim, J. Park, Y.-T. Chang, Y. Gwon. Reduced Dynamicityand Increased High-Order Protein Assemblies in Dense Fibrillar Component ofthe Nucleolus under Cellular Senescence. Redox Biology. 2024, 75:

[17] J. T. Butler, W. M. Yashar, R. Swords. Breaking the Bone MarrowBarrier: Peripheral Blood as a Gateway to Measurable Residual DiseaseDetection in Acute Myelogenous Leukemia. American Journal of Hematology.2025, 100(4): 638-651

[18] M. G. Della Porta, J. P. Bewersdorf, Y. H. Wang, R. P.Hasserjian. Future Directions in Myelodysplastic Syndromes / Neoplasms andAcute Myeloid Leukaemia Classification: From Blast Counts to Biology.Histopathology. 2024, 86(1): 158-170

[19] C. Jia, J. Gao, D. Xie, J.-Y. Wang. Tumor Diagnosis Based on Nucleolus Labeling. Sensors&Diagnostics. 2024, 3(11): 1807-1821

[20] D. Schraivogel, TM Kuhn, B. Rauscher, M. Rodríguez-Martínez, M. Paulsen, K. Owsley, et al. High-Speed ​​Fluorescence Image–Enabled CellSorting. Science. 2022, 375(6578): 315-320 Summary of the Invention To address the aforementioned deficiencies or improvement needs of existing technologies, the present invention aims to provide a blood cell analysis system based on image flow cytometry. This system employs a combination of specific cell staining dyes (i.e., a first fluorescent probe that specifically binds to or accumulates in the nucleolus, a second fluorescent probe that specifically binds to nuclear DNA, and a third fluorescent probe that specifically binds to blood cell differentiation antigens), in conjunction with an image flow cytometer. Utilizing optimized multidimensional fluorescent staining combinations and a supporting automated image analysis workflow, it achieves simultaneous, high-throughput, and precise quantitative analysis of key features of blood cell immunophenotype and morphology at the single-cell level. This addresses the technical challenge of separating morphological and immunophenotypic information in existing technologies.

[0008] To achieve the above objectives, according to one aspect of the present invention, a blood cell analysis system based on image flow cytometry is provided, characterized in that it comprises: The cell staining module includes a first fluorescent probe, a second fluorescent probe, and a third fluorescent probe for staining a blood cell sample to be analyzed. The first fluorescent probe specifically binds to or accumulates within the nucleolus of the cell; the second fluorescent probe specifically binds to nuclear DNA; and the third fluorescent probe specifically binds to blood cell differentiation antigens. An imaging flow cytometer is used to simultaneously acquire multiple fluorescence images and bright-field images for each stained cell, including at least one fluorescent probe, a second fluorescent probe, and a third fluorescent probe. An image processing module is used to identify and delineate a target cell subpopulation based on the image signal corresponding to the third fluorescent probe; within the target cell subpopulation, to define the nucleus region of each cell based on the image corresponding to the second fluorescent probe; within the nucleus region, to identify and segment the nucleolar region based on the image corresponding to the first fluorescent probe, and to calculate the nucleolar detection rate and / or the average number of nucleoli; wherein: Nucleolus detection rate = (Number of cells with nucleoli / Total number of nucleated cells); Average number of nucleoli = (total number of nucleoli / total number of nucleated cells).

[0009] As a further preferred embodiment of the present invention, the imaging flow cytometer is used to simultaneously acquire bright-field images, immunofluorescence images, and morphological images of each stained cell; wherein the morphological images include both DNA fluorescence images and RNA fluorescence images.

[0010] As a further preferred embodiment of the present invention, the first fluorescent probe is a fluorescent dye for labeling the nucleolus, the second fluorescent probe is a DNA fluorescent dye, and the third fluorescent probe is an immunomarker probe coupled with a fluorescein. The fluorescent dye used to label the nucleolus includes one of RNA fluorescent dyes and nucleolar protein-specific antibodies conjugated with fluorescein.

[0011] As a further preferred embodiment of the present invention, the immunomarker probe is a fluorescein-conjugated anti-human leukocyte differentiation cluster (CD) monoclonal antibody for delineating target leukocyte populations; preferably, it is a fluorescein-conjugated anti-human leukocyte differentiation cluster (CD) monoclonal combination antibody for delineating target leukocyte populations, containing at least a fluorescein-labeled CD45 antibody. The DNA fluorescent dye is any one of DRAQ5, 7-AAD, and DAPI; preferably DRAQ5, and more preferably a final concentration of 5 μM.

[0012] As a further preferred embodiment of the present invention, the image processing module is also used to analyze at least one morphological parameter among cell area, cell roundness and shape, cell nucleus morphology, nuclear area, nucleocytoplasmic ratio, chromatin density, cytoplasmic RNA content, cytoplasmic content, cell granularity, and whether nuclear shift has occurred. Preferably, the image processing module is also capable of performing correlation analysis between the morphological parameters and target cell subpopulations with specific immune phenotypes.

[0013] As a further preferred embodiment of the present invention, the imaging flow cytometer is a Cytek® Amnis® ImageStream MKII series or a BD FACSDiscover™ S8 series imaging flow cytometer.

[0014] As a further preferred embodiment of the present invention, the blood cell sample to be analyzed is a peripheral blood mononuclear cell (PBMC) sample or a bone marrow blood mononuclear cell (BM-MNC) sample.

[0015] As a further preferred embodiment of the present invention, it also includes: Sample preprocessing module: used to perform red blood cell lysis and centrifugation on whole blood samples to remove red blood cells, thereby obtaining a sample containing blood cells to be analyzed.

[0016] According to another aspect of the present invention, the present invention provides an auxiliary diagnostic device for hematological diseases, characterized in that it includes the above-mentioned blood cell analysis system based on image flow cytometry.

[0017] As a further preferred embodiment of the present invention, the hematological disease is at least one of acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL), and myelodysplastic syndrome (MDS).

[0018] According to another aspect of the present invention, the present invention provides the application of the above-described image flow cytometry-based blood cell analysis system in blood cell analysis for non-diagnostic purposes.

[0019] Compared with the prior art, the above-described technical solutions conceived in this invention can achieve the following beneficial effects: 1. Achieved comprehensive digital translation of morphology: This invention combines a third fluorescent probe that specifically binds to blood cell differentiation antigens with a first fluorescent probe (e.g., RNA fluorescent dye) and a second fluorescent probe (e.g., DNA fluorescent dye) to stain cells. Combined with imaging flow cytometry, this enables integrated detection of the morphology and immunity of samples containing blood cells. Through a simple nucleic acid staining protocol, this invention successfully and reproducibly quantifies key morphological features observed under a microscope in traditional blood smears—mean nucleolar number and nucleolar detection rate—on an imaging flow cytometry platform. This represents a major paradigm shift in the field of blood cell analysis (the nucleolus is the most prominent substructure within the cell nucleus, serving as the factory for ribosome biosynthesis and the hub for cellular stress perception; its state is closely related to cell proliferation and metabolic activity. For example, the number of nucleoli and the mean number of nucleoli reflect the number of primitive cells and are crucial for the diagnosis of acute leukemia). Furthermore, it can be further correlated with other morphological characteristics, including cell area, cell roundness and shape, nuclear morphology, nuclear area, nucleocytoplasmic ratio, chromatin density, cytoplasmic RNA content, cytoplasmic content, cell granularity, and whether nuclear shift has occurred, for further analysis.

[0020] 2. Generation of new data: This invention integrates multi-dimensional information such as the immunophenotype, overall morphology, nuclear morphology, chromatin structure and content, nucleolar characteristics and cytoplasmic state of a single cell at the single-cell level in a high-throughput manner, so that morphology and immunity can correspond one-to-one at the cell level, providing new data information for a single blood cell.

[0021] 3. Technological Breakthrough Achieved: This invention can be applied to the auxiliary diagnosis of hematologic malignancies characterized by primitive cell proliferation and nucleolar alterations. This invention achieves high-throughput, objective, and quantitative analysis of the nucleolar morphology, a key diagnostic indicator for acute leukemia, and directly correlates it with immunophenotype, filling a technological gap in high-throughput morphoimmunoassay.

[0022] 4. High throughput and minimally invasive: Taking the analysis of peripheral blood mononuclear cell (PBMC) samples as an example, only a small amount of peripheral blood is needed to complete the comprehensive analysis of tens of thousands of cells within minutes, which greatly improves the detection efficiency and the ability to capture rare cells, and is especially suitable for MRD monitoring.

[0023] In summary, the blood cell analysis system based on image flow cytometry in this invention can simultaneously achieve immunophenotyping and quantitative analysis of multidimensional morphological characteristics at the single-cell level, and has the characteristics of high throughput, objectivity and accuracy. Attached Figure Description

[0024] Figure 1 This is a flowchart illustrating an embodiment of the present invention; Figure 2 This is a flow cytometry image of a typical healthy person's peripheral blood mononuclear cells from Example 1; Figure 3 This is a flow cytometry image of peripheral blood blasts typical of AML patients in Example 2; Figure 4 This is a statistical comparison chart of nucleolar quantitative parameters between the healthy group (denoted as H in the figure) in Example 1 and the AML group (denoted as AML in the figure) in Example 2; wherein, Figure 4 a) in the figure represents a comparison of nucleolus detection rates. Figure 4 b) in the figure represents the comparison of the average number of nucleoli. Detailed Implementation

[0025] 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. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0026] Based on this invention, in practical applications, such as Figure 1As shown, the following steps may be included: Step 1: Sample Processing (a) Sample pretreatment and multidimensional fluorescence staining: For EDTA-anticoagulated whole blood or bone marrow blood samples, white blood cells are collected by centrifugation, preferably treated with 1× red blood cell lysis buffer, and stained with a reagent combination containing the following components: i. Immunoblot probe: A fluorescein-conjugated anti-human leukocyte differentiation cluster (CD) monoclonal antibody, such as CD45, for delineating target leukocyte populations. Preferably, the cells are surface stained in staining buffer BD Stain Buffer (FBS), preferably incubated at 4°C in the dark for 30 minutes.

[0027] ii. DNA fluorescent dye: A DNA-specific fluorescent dye (e.g., DRAQ5, 7-AAD, DAPI). For example, DRAQ5 at a final concentration of 5 μM can be used and incubated at 37°C in the dark for 20 minutes.

[0028] iii. RNA fluorescent dye: A fluorescent probe capable of labeling intracellular RNA. During the development of this invention, it was discovered that although RNA probes such as SYTO RNASelect can bind to all intracellular RNA, in rapidly proliferating cells such as primitive cells, the nucleolus, as the ribosome synthesis center, has an extremely high local concentration of rRNA. Staining with this probe can form well-defined, high-brightness fluorescent spots within the cell nucleus that can be identified by image algorithms. The location, number, and intensity of these spots highly correspond to the nucleolar characteristics under traditional microscopy, thus serving as an effective digital characterization of the nucleolus. Preferably, the RNA dye SYTO® RNASelect™ Green is used, preferably at a concentration of 500 nM, and preferably incubated at 37°C in the dark for 20 minutes. As another embodiment of this invention, nucleolar protein-specific antibodies conjugated to fluorescein, such as specific antibodies against fibrillarin (FBL) and nucleolar phosphoprotein (NPM1) conjugated to fluorescein (e.g., Alexa Fluor® 488 Conjugate), can also be used for labeling. The results can be corroborated with the labeling results based on the RNA probe.

[0029] (b) Pretreatment before cell loading: After washing the cells, prepare at least 50 μL of phosphate-buffered saline (PBS) to resuspend the cells at a concentration of 1-2 × 10⁻⁶. 7 The sample was collected at cells / mL and filtered through a 70-mesh filter.

[0030] Step 2: Imaging Flow Cytometry Data Acquisition An imaging flow cytometer (IFC) is a flow cytometer with imaging capabilities, comprising a fluid flow system, an optical system, a detection system, and a data processing system. Using an imaging flow cytometer, such as the Cytek® Amnis® ImageStream MKII series or the BD FACSDiscover™ S8 series from BD, bright-field images, immunofluorescence images, and morphological images (i.e., DNA fluorescence images and RNA fluorescence images) are simultaneously acquired for each stained cell. For example, data acquisition from a prepared sample can be performed using the Amnis® ImageStream MKII imaging flow cytometer. The following settings can be configured: i. Instrument setup: Collect data from blank controls and single-stain compensation tubes to complete instrument voltage settings, fluorescence compensation matrix calculations, and gate boundary determination.

[0031] ii. Channel acquisition: Bright field images, immunofluorescence images, DNA fluorescence images, and RNA fluorescence images are acquired simultaneously for each cell.

[0032] iii. Data volume: For all stained samples, collect a predetermined number of nucleated single cells (e.g., at least 10,000 nucleated single cells).

[0033] Step 3: Automated quantitative analysis of multidimensional cell morphological characteristics Taking the Amnis® ImageStream MKII imaging flow cytometer as an example, one can refer to existing technology reports (e.g., KA Elfimov, DA Baboshko, NM Gashnikova. Imaging Flow Cytometry in Hiv Infection Research: Advantages and Opportunities. Methods and Protocols. 2025, 8(1)) and use IDEAS. ® Data analysis software was used for reproduction. This software can automatically calculate more than 200 morphological and photometric characteristics and related statistical data for each cell, identifying unique cell populations not only based on their fluorescence intensity characteristics but also on their morphological characteristics. The analysis steps can be as follows: i. Adjust compensation using single-positive sample data; ii. Use Gradient RMS values ​​to identify well-focused cells. The higher the Gradient RMS value, the better the focusing effect. Gradient RMS > 60 is preferred. iii. Use scatter plots of Area and Aspect Ratio to delineate single cells, excluding red blood cells and cell debris, and select cell populations with moderate Area and Aspect Ratio close to 1; iv. In the scatter plots of Threshold Area of ​​Nucleus and Bright Detail Intensity of Nucleus, select the cell population with a smaller Threshold Area of ​​Nucleus value and a higher Bright Detail Intensity of Nucleus value, and remove apoptotic cells; v. Based on the signals of immune marker probes and forward and side-scattered light, the user can delineate the cell subpopulations of interest. This experiment only delineates peripheral blood mononuclear cells by double positivity for CD45 and DNA. Optionally, the user can also delineate the subpopulations of interest, such as eosinophils, monocytes, and platelets, according to the purpose. vi. The user analyzes the subpopulation of cells within the phylum in step v, creates a cell mask based on the bright field gradient, creates a nuclear mask based on the fluorescence gradient of the DNA fluorescence channel, and obtains the cytoplasmic mask through Boolean subtraction between the cell mask and the nuclear mask; vii. Apply IDEAS to nucleolar fluorescence images within the nuclear mask region. ® The SpotLocalization algorithm of the data analysis software automatically identifies green RNA fluorescent spots in the "nucleolar mask" and calculates the number of nucleoli per cell based on the Spot Count value. It also calculates the nucleolar detection rate and the average number of nucleoli.

[0034] Nucleolus detection rate = (Number of cells with nucleoli / Total number of nucleated cells); Average number of nucleoli = (total number of nucleoli / total number of nucleated cells).

[0035] viii. Optional, users can use IDEAS ®Data analysis software assists in extracting cell size and texture, dark field intensity and granularity, as well as nuclear fluorescence intensity, texture and shape, and other commonly used analytical features of the target cell subpopulation of interest. It quantitatively analyzes one or more of the following: cell area (cell mask area), cell roundness and shape (by calculating cell mask roundness, aspect ratio, and concavity / convexity), nuclear morphology (by calculating nuclear roundness, aspect ratio, and concavity / convexity), nuclear area (nuclear mask area), nucleocytoplasmic ratio (nuclear area / cell area), chromatin density (relative heterochromatin fraction, calculated by thresholding high-brightness DNA regions within the nucleus), cytoplasmic RNA content (such as the total fluorescence intensity of SYTO® RNASelect™ Green within the cytoplasmic mask), cytoplasmic content (difference between cell mask area and nuclear mask area), cell granularity (distribution of side-scattered light SSC & CD45 scatter plot), and nuclear shift (eccentricity of the nuclear mask within the cell mask).

[0036] Example 1: Analysis of peripheral blood samples from healthy controls based on the present invention Sample source: Six healthy volunteers (numbered H1-H6) were recruited, and 2 mL of peripheral blood EDTA anticoagulated samples were collected from each of them.

[0037] Sample processing: Follow the steps outlined above, i.e., based on the results of the complete blood count (CBC) and white blood cell count, collect approximately 200 μL of peripheral blood (controlling the white blood cell count to 1 × 10⁻⁶). 6 -1×10 7 Between 5 × 10⁶ / mL, preferably 5 × 10⁶ / mL. 6 Peripheral blood mononuclear cells (PBMCs) were collected by centrifugation after adding 1× red blood cell lysis buffer to lyse the red blood cells. The cells were shaken well and incubated in the dark for 15 minutes. PBMCs were then collected by centrifugation. 100 μL of fetal bovine serum-FBS buffer containing APCCY7-labeled CD45 antibody (BD Pharmingen, catalog number 557833) was added to the PBMCs, and the cells were incubated at 4°C in the dark for 30 minutes for surface staining. After washing, 200 μL of PBS containing 500 nM SYTO® RNASelect™ Green dye (Molecular Probes, catalog number S32703) and 5 µM DRAQ5 dye (Abcam, catalog number ab108410) was added to the PBMCs, and the cells were incubated at 37°C in the dark for 20 minutes. Finally, the cells were resuspended in PBS and the concentration was adjusted to 1-2 × 10⁶ cells / mL. 7 cells / mL, filtered through a 70-mesh filter before being used in the instrument.

[0038] Image flow cytometry data acquisition: In this embodiment, the Amnis ImageStreamX Mk II instrument was used to acquire data according to the configuration in step two above. More than 10,000 cell events were acquired for each sample, and bright field (Ch1, 488 nm laser) images and fluorescence images of SYTO® RNASelect™ Green fluorescence (Ch2, 488 nm laser), APCCY7 fluorescence (Ch6, 642 nm laser), and DRAQ5 fluorescence (Ch5, 642 nm laser) were obtained.

[0039] Automated quantitative analysis of multidimensional cell morphological characteristics: Analysis was performed using IDEAS® data analysis software. Following step three above, the cell morphological characteristics were quantitatively analyzed. a) Adjustment and compensation: Blank control, SYTO®RNASelect™ Green fluorescence, CD45-APC-CY7 fluorescence, and DRAQ5 fluorescence single label were entered into the adjustment and compensation. b) Cell delineation: Based on Gradient RMS parameters, cells with clear focus are screened (Gradient RMS > 60); then, morphologically normal single cells are delineated in the Area Ch1 vs Aspect Ratio Ch1 scatter plot, preferably with an Aspect Ratio > 0.5; c) Removal of apoptotic cells: In the Threshold Area of ​​Nucleus and Bright Detail Intensity of Nucleus scatter plots, cell populations with lower Threshold Area of ​​Nucleus values ​​and higher Bright Detail Intensity of Nucleus values ​​are removed; d) Gating: In the Intensity-MC-Ch6 vs Intensity-MC-Ch5 scatter plot, double-positive cells are delineated to gate leukocytes. e) Nucleolar quantification: First, a nuclear mask is created using the DRAQ5 signal. Within the nuclear mask, the SYTO signal is analyzed using the "Spot Count" function. The nucleoli are automatically identified and counted by the software as bright spots within the cell nuclei.

[0040] Result: As Figure 2 As shown, Figure 2From left to right, the images show bright field (BF), CD45 (magenta), RNA (green), DNA (red), DNA / RNA combined channel, and DNA / BF combined channel. The top left corner of each row indicates the cell acquisition number. SYTO® RNASelect™ Green is an RNA dye; RNA mainly aggregates in the nucleolus within the cell nucleus, appearing as yellow dots on the image. Morphological results showed that healthy human PBMCs lacked aggregated green nucleolar signals in the cell nuclei; nucleolar signals were generally weak or absent, with an average nucleolar detection rate of only 4.03% (P<0.01), and typically only one nucleolus was detected, with a faint signal. The average number of nucleoli was 0.042 (P<0.01). In addition, nuclear shift (e.g., Figure 2 3104, 8933), cell nucleus shape (e.g. Figure 2 1420, 1504), cell size (e.g. Figure 2 3104, 1504), nucleocytoplasmic ratio (e.g.) Figure 2 It exhibits several typical PBMC imaging features, such as 8933 and 7929. Immunologically, the presence or absence of CD45 fluorescence expression and its intensity can be used to determine whether a cell belongs to the white blood cell group.

[0041] Conclusion: Based on the analysis of peripheral blood samples from healthy controls using this invention, it can be concluded that this scheme can successfully achieve integrated detection of morphology and immunity in normal human samples. At the same time, it can quantitatively analyze the immune information and morphological information of individual cells, such as cell area, cell roundness and shape, nuclear morphology, nuclear area, nucleocytoplasmic ratio, chromatin density, cytoplasmic RNA content, cytoplasmic content, cell granularity, whether nuclear shift has occurred, number of nucleoli and average number of nucleoli.

[0042] Example 2: Analysis of peripheral blood samples from AML patients based on the present invention Sample source: Six newly diagnosed AML patients (numbered AML1-AML6) were collected, and 2 mL of peripheral blood EDTA anticoagulated samples were collected from each of them.

[0043] Sample processing: Proceed as described in step one above, i.e., based on the results of the complete blood count and white blood cell count, obtain 200 μL of peripheral blood (controlling the white blood cell count to 1×10⁻⁶). 6 -1×10 7 Between 5 × 10⁶ / mL, preferably 5 × 10⁶ / mL. 6Red blood cells were lysed with 1× red blood cell lysis buffer ( / mL), shaken well, and incubated in the dark for 15 minutes. PBMCs were then collected by centrifugation. 100 μL of fetal bovine serum-FBS buffer containing PE-labeled CD45 antibody (BD Pharmingen, catalog number 560975) was added to the PBMCs, and the cells were incubated at 4°C in the dark for 30 minutes for surface staining. After washing, 200 μL of PBS containing 500 nM SYTO® RNASelect™ Green dye (Molecular Probes, catalog number S32703) and 5 µM DRAQ5 dye (Abcam, catalog number ab108410) was added to the PBMCs, and the cells were incubated at 37°C in the dark for 20 minutes. Finally, the cells were resuspended in PBS and the concentration was adjusted to 1-2 × 10⁻⁶. 7 cells / mL, filtered through a 70-mesh filter before being used in the instrument.

[0044] Image flow cytometry data acquisition: In this embodiment, the Amnis ImageStreamX Mk II instrument was used to acquire data according to the configuration in step two above. More than 10,000 cell events were acquired for each sample, and bright field (Ch1, 488 nm laser) images and fluorescence images of SYTO® RNASelect™ Green fluorescence (Ch2, 488 nm laser), PE fluorescence (Ch3, 488 nm laser), and DRAQ5 fluorescence (Ch5, 642 nm laser) were obtained.

[0045] Automated quantitative analysis of multidimensional cell morphology characteristics: Analysis was performed using IDEAS® software. Following step three above, the quantitative analysis of cell morphology characteristics was conducted as follows: a) Adjustment compensation: Blank control, SYTO® RNASelect™ Green fluorescence, CD45-PE fluorescence, and DRAQ5 fluorescence single label were entered for adjustment compensation. b) Cell delineation: Cells with clear focus were screened based on Gradient RMS parameters (Gradient RMS > 60); then, morphologically normal single cells were delineated in the Area Ch1 vs Aspect Ratio Ch1 scatter plot, preferably with an Aspect Ratio > 0.5. c) Removal of apoptotic cells: In the Threshold Area of ​​Nucleus and Bright Detail Intensity of Nucleus scatter plots, cell populations with lower Threshold Area of ​​Nucleus values ​​and higher Bright Detail Intensity of Nucleus values ​​were removed. d) Gating: In the Intensity-MC-Ch3 vs Intensity-MC-Ch5 scatter plot, double-positive cells were delineated to gate leukocytes. e) Nucleolar quantification: First, a nuclear mask is created using the DRAQ5 signal. Within the nuclear mask, the SYTO signal is analyzed using the "Spot Count" function. The nucleoli are the bright spots in the nuclei that are automatically identified and counted by the software.

[0046] Results: Nucleolar signaling was significantly enhanced in PBMCs of AML patients, with a significantly increased proportion of primitive cell populations possessing bright, clean nucleoli. Figure 3 As shown, Figure 3 From left to right, the images show bright field, CD45 (magenta), RNA (green), DNA (red), DNA / RNA combined channel, and DNA / BF combined channel, with white arrows indicating clearly visible nucleoli. The results show aggregated nucleolar signals within the nuclei of PBMCs from AML patients. The average nucleolar detection rate was as high as 38.40% (P<0.01), significantly higher than the 4.03% in the healthy control group (P<0.01). Furthermore, the nuclei generally exhibited 1-4 clear, bright nucleolar fluorescent signals, with an average of 0.857 nucleoli (P<0.01), significantly higher than the 0.042 in the normal group (P<0.01). In addition, the cells and nuclei were mostly round, with less cytoplasm, exhibiting typical primitive cell morphological characteristics. Immunologically, the presence and intensity of CD45 fluorescence expression can be used to determine whether a cell belongs to the leukocyte group. These observations are highly consistent with the clear nucleoli characteristic of primitive cells observed under traditional bone marrow smear microscopy.

[0047] The nucleolus is the most prominent substructure within the cell nucleus. As the factory for ribosome biosynthesis and a hub for cellular stress sensing, its state is closely related to cell proliferation and metabolic activity. Blood cells in the peripheral blood of healthy individuals are mature and lack proliferative capacity. However, in peripheral blood samples from leukemia patients, tumor cells undergo intense ribosome synthesis to meet their rapid proliferation needs, often exhibiting pathological features such as enlarged nucleoli, increased number of nucleoli, and abnormal morphology. A statistical comparison was made between the nucleolus detection rate and average number of nucleoli obtained from the healthy group (as a control) in Example 1 and the AML patient group in Example 2. The results are as follows: Figure 4 As shown, it is easy to see that the nucleolus detection rate and the average number of nucleoli in the peripheral blood of AML patients are higher than those of normal people. Moreover, this method can combine the morphology and immunology of blood cells and can analyze the main morphological and immunological characteristics of each cell in a high-throughput and quantitative manner. In Example 2, the average nucleolus detection rate of the AML patient group was as high as 38.40%, which is consistent with the clinical characteristic that the proportion of primitive cells in the peripheral blood of newly diagnosed AML patients is >20%.

[0048] The above embodiments are merely examples, and only use CD45 and DNA double positivity to delineate peripheral blood mononuclear cells. Optionally, users can also delineate subpopulations of interest, such as eosinophils, monocytes, lymphocytes, etc. For example, the first fluorescent probe in this invention is a nucleolar probe, capable of specifically labeling nucleolar structures or components. Besides RNA-specific dyes, nucleolar-specific fluorescent probes can also be nucleolar protein-specific antibodies conjugated with fluorescein, including fluorescently labeled antibodies against FBL, NPM1, or nucleolin (NCL). For example, in addition to CD45 antibodies as immunophenotypic antibodies, depending on the target cell subpopulation, fluorescein-conjugated anti-human leukocyte differentiation cluster (CD) monoclonal combination antibodies can be used, especially combinations of fluorescein-labeled CD45 antibodies with other CD antibodies (e.g., combinations of CD45 antibodies with CD117 and CD34 antibodies).

[0049] In addition to its use in the auxiliary diagnosis of hematologic diseases such as AML, ALL, and MDS, this invention can also be applied to non-diagnostic scenarios such as high-throughput drug screening (HTS) and mechanism research, quality control of cell therapy products such as CAR-T, and basic immunology research such as lymphocyte activation and cell interaction analysis.

[0050] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An image-based flow cytometry blood cell analysis system, comprising: The method comprises the following steps: a cell staining module, which comprises a first fluorescent probe, a second fluorescent probe and a third fluorescent probe, is used to stain a sample containing blood cells to be analyzed; wherein the first fluorescent probe can specifically bind to or be enriched in components within the nucleolus; the second fluorescent probe can specifically bind to cell nucleus DNA; and the third fluorescent probe can specifically bind to blood cell differentiation antigens; an image flow cytometer is used to synchronously collect at least a plurality of fluorescent images corresponding to the first fluorescent probe, the second fluorescent probe and the third fluorescent probe and a bright field image for each cell after staining; an image processing module is used to identify and delineate a target cell subpopulation based on an image signal corresponding to the third fluorescent probe; within the target cell subpopulation, a cell nucleus region of each cell is defined based on an image corresponding to the second fluorescent probe; and within the cell nucleus region, a nucleolus region is identified and segmented based on an image corresponding to the first fluorescent probe, and a nucleolus detection rate and / or an average number of nucleoli are calculated; wherein: the nucleolus detection rate = (number of cells with nucleoli / total number of cells with nuclei); the average number of nucleoli = (total number of nucleoli / total number of cells with nuclei).

2. The system of claim 1, wherein, The image flow cytometer is used to synchronously collect a bright field image, an immunofluorescence image and a morphological image for each cell after staining; wherein the morphological image comprises a DNA fluorescent image and an RNA fluorescent image.

3. The system of claim 1, wherein, The first fluorescent probe is a fluorescent dye for labeling nucleoli, the second fluorescent probe is a DNA fluorescent dye, and the third fluorescent probe is an immunomarker probe coupled with fluorescein; The fluorescent dye for labeling nucleoli comprises one of an RNA fluorescent dye and a nucleolin-specific antibody coupled with fluorescein.

4. The system of claim 3, wherein, The immunomarker probe is a fluorescein-coupled anti-human leukocyte differentiation cluster (CD) monoclonal antibody for delineating a target leukocyte population; preferably, it is a fluorescein-coupled anti-human leukocyte differentiation cluster (CD) monoclonal combination antibody for delineating a target leukocyte population, and at least contains a fluorescein-labeled CD45 antibody. The DNA fluorescent dye is any one of DRAQ5, 7-AAD and DAPI; preferably, it is DRAQ5, and the final concentration is more preferably 5 μM.

5. The system of claim 1, wherein, The image processing module is further used to analyze at least one morphological parameter of a cell area, a cell circularity and shape, a nucleus shape, a nucleus area, a nucleus-cytoplasm ratio, a chromatin density, a cytoplasmic RNA content, a cytoplasmic content, a cell granularity and whether nuclear deviation occurs. Preferably, the image processing module can further perform correlation analysis on the morphological parameters and a target cell subpopulation with a specific immunophenotype.

6. The system of claim 1, wherein, The image flow cytometer is an image flow cytometer of the Cytek® Amnis® ImageStream series or a BD FACSDiscover™ S8 series.

7. The system of claim 1, wherein, The sample containing blood cells to be analyzed is a peripheral blood mononuclear cell (PBMC) sample to be analyzed or a bone marrow blood mononuclear cell (BM-MNC) sample to be analyzed.

8. The system of claim 1, wherein, Also included are: A sample pre-processing module for performing red blood cell lysis processing and centrifugal separation on a whole blood sample to remove red blood cells, thereby obtaining a sample containing blood cells to be analyzed.

9. An auxiliary diagnostic device for hematological diseases, characterized in that, The blood cell analysis system based on image flow cytometry according to any one of claims 1-8; Preferably, the blood system disease is at least one of acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL), and myelodysplastic syndrome (MDS).

10. Use of the blood cell analysis system based on image flow cytometry according to any one of claims 1-8 in blood cell analysis for non-diagnostic purposes.

Images