Cell classification and function evaluation method and system based on deep learning nucleolus morphology

Abstract

The invention relates to the field of cell classification, in particular to a cell classification and function evaluation method and system based on deep learning nucleolus morphology. The method provided by the invention comprises the following steps: S101, acquiring microscopic image data of a biological sample; s102, acquiring a first region of interest, a second region of interest and a third region of interest; s103, extracting kernel feature data; s104, carrying out first classification on cells in the biological sample; and S105, calculating the number and / or proportion of the A-type cells and the B-type cells in the biological sample so as to evaluate the functional state of the tissues or organs corresponding to the biological sample. The method can accurately and sensitively reflect and evaluate the functional state of the tissue or organ corresponding to the biological sample, and is especially suitable for the biological sample which is lack of a specific marker and has strong heterogeneity.

Description

[0001] This application is a divisional application based on the Chinese invention patent application with application number 202510300854.X, application date March 14, 2025, and invention name “A method and system for cell classification and function evaluation based on nucleolar morphology”. Technical Field

[0002] The present invention relates to the field of cell classification, and in particular to a method and system for cell classification and function evaluation based on deep learning of nucleolar morphology. Background Art

[0003] Traditional disease characterization methods rely on specific disease markers and are often limited by factors such as high tissue heterogeneity and low marker expression levels. In particular, when screening for drug efficacy and molecular targets, existing methods are often limited to assessing toxicity indicators by measuring cell viability or apoptosis levels, which makes it difficult to reflect changes in the response of key cell subpopulations to drugs.

[0004] As a crucial subcellular structure within the cell, the nucleolus undertakes multiple core functions, including rRNA transcription, ribosome biogenesis, cellular stress and response, and cell cycle regulation. The nucleolus plays a key role in cell growth and is closely associated with cell growth, proliferation, differentiation, and dysfunction. Recent studies have shown that changes in nucleolus morphology can reflect cellular health and are particularly relevant to the development and progression of cancer, immune disorders, and cardiovascular diseases.

[0005] Although existing technologies have certain evaluation criteria for changes in nucleolar morphology, such as the enlargement of the nucleoli of cancer cells, they lack clear judgment criteria, and the evaluation of nucleolar morphology results is highly subjective. For more complex biological samples (such as tissue samples with ambiguous nucleolar phenotypes and high heterogeneity), the identification accuracy is low and the repeatability is poor. For example, Chinese patent application CN112239455A discloses an RNA fluorescent probe that uses nucleolar morphology changes to quickly distinguish cancer from normal tissue. The interpretation of the nucleolar staining results not only relies on manual interpretation, but can only distinguish between breast tissue and normal tissue, making it difficult to characterize the specific pathological state corresponding to the biological sample. Chinese patent application CN111886630A discloses an automated method for analyzing biological tissues and cells. This method is based on three-dimensional images, which is difficult and complex to image. It requires combining multiple geometric morphological features of the nucleolus to achieve cell classification, and it is difficult to sensitively reflect the changing trends of the nucleolus. As a result, it is difficult to apply to application scenarios such as early disease diagnosis and high-throughput drug screening. Summary of the Invention

[0006] In a first aspect, the present invention provides a method for cell classification and function assessment based on nucleolar morphology, comprising the following steps: S101 acquires microscopic image data of biological samples; In some embodiments, the microscopic image data includes optical microscope images with an objective lens of 40 times or more and / or super-resolution microscopic image data.

[0007] In some embodiments, the biological sample comprises a tissue or cell culture comprising at least one cell.

[0008] In some embodiments, the biological sample is labeled with nucleolar markers, and the nucleolar markers include at least a dense fiber component marker and a granular component marker.

[0009] In some embodiments, the dense fiber component marker comprises fibrinogen (FBL).

[0010] In some embodiments, the granule component marker comprises nucleophosmin 1 (NPM1).

[0011] In some embodiments, the biological sample is further labeled with a nuclear marker.

[0012] In some embodiments, the nuclear marker comprises 4',6-diamidino-2-phenylindole (DAPI).

[0013] S102: Acquire a first region of interest, a second region of interest, and a third region of interest in the microscopic image data; In some embodiments, the first region of interest is a cell nucleus region delineated by a fluorescent channel corresponding to the cell nucleus marker.

[0014] In some embodiments, the second region of interest is a nucleolus region delineated using a fluorescent channel corresponding to the dense fiber component marker.

[0015] In some embodiments, the third region of interest is a nucleolus region delineated using a fluorescent channel corresponding to the particle component marker.

[0016] S103 extracting nucleolus feature data of the second region of interest and the third region of interest, wherein the nucleolus feature data includes line scan data and region area data; S104: performing a first classification on cells in the biological sample based on the line scan data; In some embodiments, the line scan data includes fluorescence intensity distribution curves of the second region of interest and the third region of interest.

[0017] In some embodiments, the first classification method includes: When the fluorescence intensity distribution curve of the third region of interest surrounds the fluorescence intensity distribution curve of the second region of interest, classifying the cell as a type A cell; When the fluorescence intensity distribution curve of the second region of interest overflows outside the fluorescence intensity distribution curve of the third region of interest, the cell is classified as a type C cell.

[0018] In some embodiments, the spillover assessment method includes: Analyze whether the fluorescence intensity change trends of the fluorescence intensity distribution curve of the second region of interest and the fluorescence intensity distribution curve of the third region of interest at the first and last non-overlapping areas formed outside the fluorescence intensity distribution curve of the third region of interest are the same; if not, overflow occurs; if so, no overflow occurs.

[0019] S105 performing a second classification on the type C cells in the biological sample; In some embodiments, the second classification method includes: The ratio of the area of ​​the second region of interest to the area of ​​the third region of interest in the C-type cells is analyzed to determine whether it is abnormal, so as to perform a second classification on the C-type cells in the biological sample.

[0020] In some embodiments, if the ratio of the area of ​​the second region of interest to the third region of interest in the C-type cells is abnormal, the classification of the C-type cells is adjusted to B-type cells; if it is not abnormal, the classification of the C-type cells is adjusted to A-type cells.

[0021] In some embodiments, when the ratio of the area of ​​the second region of interest to the area of ​​the third region of interest shows an increasing trend, it is considered abnormal.

[0022] S106 Calculate the number and / or ratio of the type A cells to the type B cells in the biological sample, and then evaluate the functional status of the tissue or organ corresponding to the biological sample.

[0023] In some embodiments, the biological sample includes a biological sample treated with a drug and a biological sample not treated with a drug.

[0024] In some embodiments, the method further includes: S107 analyzing an average diameter of spots formed by the dense fiber component marker in the second region of interest in the B-type cells; when the average diameter is less than a first threshold, the biological sample is assessed as being in a malignant proliferation state.

[0025] In some embodiments, the first threshold comprises 0.5-2 μm.

[0026] In a second aspect, the present invention provides a cell classification and function assessment system based on nucleolar morphology, the system comprising: An image acquisition module 102 is configured to acquire microscopic image data of a biological sample; In some embodiments, the microscopic image data includes optical microscope images with an objective lens of 40 times or more and / or super-resolution microscopic image data.

[0027] In some embodiments, the biological sample comprises a tissue or cell culture comprising at least one cell.

[0028] In some embodiments, the biological sample is labeled with nucleolar markers, and the nucleolar markers include at least a dense fiber component marker and a granular component marker.

[0029] In some embodiments, the dense fiber component marker comprises fibrinogen (FBL).

[0030] In some embodiments, the granule component marker comprises nucleophosmin 1 (NPM1).

[0031] In some embodiments, the biological sample is further labeled with a nuclear marker.

[0032] In some embodiments, the nuclear marker comprises 4',6-diamidino-2-phenylindole (DAPI).

[0033] The first processing module is configured to acquire a first region of interest, a second region of interest, and a third region of interest in the microscopic image data.

[0034] In some embodiments, the first region of interest is a cell nucleus region delineated by a fluorescent channel corresponding to a cell nucleus marker.

[0035] In some embodiments, the second region of interest is a nucleolus region delineated by a fluorescent channel corresponding to a dense fiber component marker.

[0036] In some embodiments, the third region of interest is a nucleolus region delineated using a fluorescent channel corresponding to a particle component marker.

[0037] A second processing module is configured to extract nucleolus feature data of the second region of interest and the third region of interest, wherein the nucleolus feature data includes line scan data and region area data; a first classification module configured to perform a first classification on cells in the biological sample based on the line scan data; In some embodiments, the line scan data includes fluorescence intensity distribution curves of the second region of interest and the third region of interest.

[0038] In some embodiments, the first classification method includes: When the fluorescence intensity distribution curve of the third region of interest surrounds the fluorescence intensity distribution curve of the second region of interest, classifying the cell as a type A cell; When the fluorescence intensity distribution curve of the second region of interest overflows outside the fluorescence intensity distribution curve of the third region of interest, the cell is classified as a type C cell.

[0039] In some embodiments, the spillover assessment method includes: Analyze whether the fluorescence intensity change trends of the fluorescence intensity distribution curve of the second region of interest and the fluorescence intensity distribution curve of the third region of interest at the first and last non-overlapping areas formed outside the fluorescence intensity distribution curve of the third region of interest are the same; if not, overflow occurs; if so, no overflow occurs.

[0040] A second classification module 110 is configured to perform a second classification on the type C cells in the biological sample; In some embodiments, the second classification method includes: The ratio of the area of ​​the second region of interest to the area of ​​the third region of interest in the C-type cells is analyzed to determine whether it is abnormal, so as to perform a second classification on the C-type cells in the biological sample.

[0041] In some embodiments, if the ratio of the area of ​​the second region of interest to the third region of interest in the C-type cells is abnormal, the classification of the C-type cells is adjusted to B-type cells; if it is not abnormal, the classification of the C-type cells is adjusted to A-type cells.

[0042] In some embodiments, when the ratio of the area of ​​the second region of interest to the area of ​​the third region of interest shows an increasing trend, it is considered abnormal.

[0043] The analysis module is configured to calculate the number and / or ratio of the type A cells and the type B cells in the biological sample, and then evaluate the functional status of the tissue or organ corresponding to the biological sample.

[0044] In some embodiments, the biological sample includes a biological sample treated with a drug and a biological sample not treated with a drug.

[0045] In some embodiments, the analysis module is further configured to analyze an average diameter of spots formed by the dense fiber component marker in the B-type cells; when the average diameter is less than a first threshold, the biological sample is assessed to be in a malignant proliferation state.

[0046] In some embodiments, the first threshold comprises 0.5-2 μm.

[0047] Compared with the prior art, the beneficial effects of the present invention include at least the following aspects: Conventional single nucleolar geometric morphology indicators are difficult to sensitively reflect the changing trends of nucleoli. Therefore, existing technologies usually require combining multiple nucleolar geometric morphology features to classify cells in biological samples. Even so, existing cell classification methods based on nucleolar morphology can usually only achieve qualitative analysis.

[0048] The present invention discovered that the relative position and distribution of dense fiber component markers and granular component markers in the nucleolus will change with the functional state of the tissue or organ corresponding to the biological sample, and the line scan data of the second region of interest and the third region of interest delineated based on the dense fiber component marker and the granular component marker (i.e., the distribution change of fluorescence intensity) can sensitively reflect the change trend of the nucleolus (for example, from regular to irregular), and thus can accurately classify the cell subpopulations in the biological sample through the first classification method.

[0049] Furthermore, the present invention has discovered that although the nucleolar morphology of non-tumor and tumor biological samples exhibits different characteristics, when their nucleolar function is "abnormal," the changing trend of the ratio of the area of ​​the second region of interest to the third region of interest is similar. Therefore, based on this changing trend, the present invention conducts a second classification of the type A and type C cells obtained in the first classification, based on the first classification, to obtain a more accurate number and / or ratio of type A cells to type B cells. This enables more effective quantitative and qualitative analysis of cells, accurately and sensitively reflecting and evaluating the functional status of the tissue or organ corresponding to the biological sample. This method is particularly suitable for biological samples with high heterogeneity that lack specific markers.

[0050] The method and system provided by the present invention effectively overcome the limitations of traditional methods in their dependence on markers, and can be combined with machine learning technology to improve the analysis speed while ensuring the accuracy and efficiency of cell classification. In the early stages of the disease, they can accurately and sensitively reflect changes in nucleolus-related phenotypes to predict cell function (such as hyperproliferative activity or cell failure) and characterize changes in tissue health status (such as vascular intimal hyperplasia, tumors or degeneration, etc.) and the pathological state of diseased tissues, thereby providing support for early diagnosis of the disease.

[0051] In addition, since the method and system provided by the present invention can accurately classify cell subpopulations in biological samples with high repeatability, quantitative analysis can be performed based on the cell classification results of the present invention, which is suitable for application scenarios such as high-throughput drug screening and drug efficacy evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the embodiments or the description of the prior art. In all drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the various elements or parts are not necessarily drawn according to the actual scale. Obviously, the drawings described below are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings without inventive work.

[0053] Figure 1 is a flow chart of the method of embodiment 1; Figure 2 This is a module diagram of the system of Example 2; Figure 3 This is a graph showing the results of testing rat vascular smooth muscle cell culture samples in Example 3; Figure 4 This is a graph showing the results of testing a culture sample of the neuroblastoma cell line SK-N-SH in Example 4; Figure 5 This is a diagram showing the results of detecting neuroblastoma tumor tissue in Example 5.

[0054] 100 is a cell classification and function evaluation system, 102 is an image acquisition module, 104 is a first processing module, 106 is a second processing module, 108 is a first classification module, 110 is a second classification module, and 112 is an analysis module. DETAILED DESCRIPTION

[0055] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0056] Herein, suffixes such as "module," "component," or "unit" used to represent elements are only used to facilitate description of the present invention and have no specific meaning. Therefore, "module," "component," or "unit" may be used interchangeably.

[0057] As used herein, terms such as "upper," "lower," "inner," "outer," "front," "back," "one end," and "the other end" indicate positions or locations based on those shown in the accompanying drawings. These terms are intended solely to facilitate and simplify the description of the present invention and are not intended to indicate or imply that the devices or components referred to must have, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limitations on the present invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0058] As used herein, unless otherwise expressly specified or limited, the terms "installed," "provided with," and "connected" should be understood broadly. For example, "connected" may refer to a fixed connection, a detachable connection, or an integral connection; it may refer to a mechanical connection, a direct connection, an indirect connection via an intermediate medium, or internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention on a case-by-case basis.

[0059] As used herein, "and / or" includes any and all combinations of one or more of the associated listed items.

[0060] Herein, "plurality" means two or more than two, ie, it includes two, three, four, five, etc.

[0061] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0062] As used in this specification, the term "about" typically means + / - 5% of the stated value, more typically + / - 4% of the stated value, more typically + / - 3% of the stated value, more typically + / - 2% of the stated value, even more typically + / - 1% of the stated value, and even more typically + / - 0.5% of the stated value.

[0063] In this specification, certain embodiments may be disclosed in a format that is within a range. It should be understood that this description of "within a range" is merely for convenience and brevity and should not be interpreted as a rigid limitation on the disclosed range. Therefore, the description of a range should be considered to have specifically disclosed all possible subranges and individual numerical values ​​within this range. For example, the description of a range of 1-6 should be considered to have specifically disclosed subranges such as from 1 to 3, from 1 to 4, from 1 to 5, from 2 to 4, from 2 to 6, from 3 to 6, etc., as well as individual numbers within this range, such as 1, 2, 3, 4, 5, and 6. Regardless of the breadth of the range, the above rules apply.

[0064] Example 1 like Figure 1 As shown, this embodiment provides a cell classification and function evaluation method based on nucleolus morphology, comprising the following steps: S101 acquires microscopic image data of biological samples; In some embodiments, the microscopic image data includes optical microscope images with an objective lens of 40 times or more and / or super-resolution microscopic image data.

[0065] In some embodiments, immunofluorescence images of biological samples can be collected in batches using a high-throughput imaging system to obtain the microscopic image data. In some embodiments, imaging requirements include using a 40x objective lens or greater to image glass-bottomed or other ultra-thin bottom plates (e.g., 0.17-0.19 mm).

[0066] In some embodiments, the biological sample comprises a tissue or cell culture containing at least one cell. In some embodiments, the biological sample can be any solid or fluid sample obtained from, excreted by, or secreted by any organism. In some embodiments, the organism includes a healthy or apparently healthy human subject or a human patient affected by a condition or disease (e.g., cancer) to be diagnosed or studied. In some embodiments, the biological sample can be a biological fluid obtained from blood, plasma, serum, urine, bile, ascites, saliva, cerebrospinal fluid, aqueous humor, or vitreous humor, or any bodily secretion, exudate, or exudate (e.g., fluid obtained from an abscess or any other site of infection or inflammation), or fluid obtained from a joint (e.g., a normal joint or a joint affected by disease). In some embodiments, the biological sample can also be a sample obtained from any organ or tissue (including biopsy or autopsy specimens, such as tumor biopsies). In some embodiments, the biological sample can include cells (whether primary or cultured) or media mediated by any cell, tissue, or organ.

[0067] It should be understood that the same biological sample may be a mixture of different cell subpopulations with strong heterogeneity. For example, the same biological sample may contain cells with both "cancerous" and "non-cancerous" cell phenotypes. Therefore, it is very challenging to correctly classify each single cell in the biological sample.

[0068] In some embodiments, the biological sample is labeled with nucleolar markers, and the nucleolar markers include at least a dense fiber component marker and a granular component marker.

[0069] In some embodiments, the dense fiber component marker comprises fibrinogen (FBL).

[0070] In some embodiments, the granule component marker comprises nucleophosmin 1 (NPM1).

[0071] In this example, FBL and NPM1 are used to simultaneously label the nucleolus, wherein FBL is used as a marker of the dense fiber component (the middle layer of the nucleolus) and NPM1 is used as a marker of the granular component protein (the outer layer of the nucleolus) to label the middle and outer layers of the nucleolus of the cells in the biological sample.

[0072] In some embodiments, the biological sample is further labeled with a nuclear marker.

[0073] In some embodiments, the nuclear marker comprises 4',6-diamidino-2-phenylindole (DAPI).

[0074] In some embodiments, the above markers in cells can be labeled by immunofluorescence staining or other suitable methods.

[0075] S102: Acquire a first region of interest, a second region of interest, and a third region of interest in the microscopic image data.

[0076] In some embodiments, the first region of interest is a cell nucleus region delineated by a fluorescent channel corresponding to the cell nucleus marker.

[0077] In some embodiments, the second region of interest is a nucleolus region delineated using a fluorescent channel corresponding to the dense fiber component marker.

[0078] In some embodiments, the third region of interest is a nucleolus region delineated using a fluorescent channel corresponding to the particle component marker.

[0079] S103 extracting nucleolus feature data of the second region of interest and the third region of interest, wherein the nucleolus feature data includes line scan data and region area data; S104: performing a first classification on cells in the biological sample based on the line scan data; In some embodiments, the line scan data includes fluorescence intensity distribution curves for the second and third regions of interest. Line scanning is an image analysis method used to measure changes in fluorescence intensity in a specific region. For example, line scanning can be used to measure fluorescence intensity in a specific region by drawing a line and analyzing the fluorescence values ​​along the line to quantitatively analyze the fluorescence intensity distribution in the specific region.

[0080] Existing techniques typically use line scanning to determine the relative position of an analyzed protein within a specific region. For example, a specific marker is selected as a reference, and the degree of colocalization between the analyzed protein and the selected marker is assessed through line scanning. It should be understood that in this application scenario, the relative position and distribution of the selected marker within the specific region are assumed to be constant, thus enabling the relative position of the analyzed protein to be determined.

[0081] This embodiment uses dense fiber component markers and granular component markers to simultaneously mark the nucleolus, which not only can more accurately locate the position of the nucleolus, but also this embodiment finds that, unlike conventional thinking, the relative position and distribution of dense fiber component markers and granular component markers in the nucleolus will change with the functional state of the tissue or organ corresponding to the biological sample, and the line scan data of the second region of interest and the third region of interest delineated based on the dense fiber component marker and the granular component marker (i.e., the distribution change of fluorescence intensity) can sensitively reflect the change trend of the nucleolus (for example, from regular to irregular), thereby helping to achieve preliminary classification of cells in the biological sample.

[0082] It should be emphasized that conventional single nucleolar geometric morphological characteristics are difficult to sensitively reflect the changing trend of the nucleolus (for example, the change from regular to irregular). Therefore, existing technologies generally require combining multiple nucleolar geometric morphological characteristics to classify cells in biological samples. This embodiment uses the first classification method to quickly and accurately determine the changing trend of the nucleolus and then classify cells into type A and type C cells.

[0083] In some embodiments, the first classification method includes: When the fluorescence intensity distribution curve of the third region of interest surrounds the fluorescence intensity distribution curve of the second region of interest, classifying the cell as a type A cell; When the fluorescence intensity distribution curve of the second region of interest overflows outside the fluorescence intensity distribution curve of the third region of interest, the cell is classified as a type C cell.

[0084] In some embodiments, the outside of the fluorescence intensity distribution curve of the third region of interest refers to a side of the fluorescence intensity distribution curve having a higher fluorescence intensity than that of the third region of interest.

[0085] In some embodiments, "spillover" means that part of the segment of the fluorescence intensity distribution curve of the second region of interest is distributed outside the fluorescence intensity distribution curve of the third region of interest, and forms one or more non-overlapping areas with the fluorescence intensity distribution curve of the third region of interest, and the fluorescence intensity change trends of the segments of the fluorescence intensity distribution curve of the second region of interest corresponding to the non-overlapping area are different from those of the fluorescence intensity distribution curve of the third region of interest.

[0086] In some embodiments, the segment of the fluorescence intensity distribution curve of the second region of interest corresponding to the overflowed non-overlapping area generally presents a "peak"-shaped change trend (i.e., in this segment, the fluorescence intensity first increases to a peak value and then decreases), while the segment of the fluorescence intensity distribution curve of the corresponding third region of interest generally presents a non-"peak"-shaped change trend (for example, in this segment, the fluorescence intensity first decreases and then increases, only increases, or only decreases, etc.).

[0087] In some embodiments, a "peak" refers to a point on a fluorescence intensity distribution curve where the curve changes direction from top to bottom along a mountain-shaped portion as the curve rises and falls in a vertical direction.

[0088] In some embodiments, the spillover assessment method includes: Analyze whether the fluorescence intensity change trends of the fluorescence intensity distribution curve of the second region of interest and the fluorescence intensity distribution curve of the third region of interest at the first and last non-overlapping areas formed outside the fluorescence intensity distribution curve of the third region of interest are the same; if not, overflow occurs; if so, no overflow occurs.

[0089] In some embodiments, "the fluorescence intensity distribution curve of the third region of interest surrounds the fluorescence intensity distribution curve of the second region of interest (which can also be understood as the fluorescence intensity distribution curve of the third region of interest substantially surrounds the fluorescence intensity distribution curve of the second region of interest)" can be: the fluorescence intensity distribution curve of the third region of interest completely surrounds the fluorescence intensity distribution curve of the second region of interest; it can also be: the fluorescence intensity distribution curve of the second region of interest and the fluorescence intensity distribution curve of the third region of interest mostly overlap, the fluorescence intensity distribution curve of the second region of interest and the fluorescence intensity distribution curve of the third region of interest have a non-overlapping area, but the fluorescence intensity change trend of the fluorescence intensity distribution curve of the second region of interest and the fluorescence intensity distribution curve of the third region of interest corresponding to the non-overlapping area is the same (i.e., both are "peak" shaped) and / or the fluorescence intensity of the fluorescence intensity distribution curve of the second region of interest and the fluorescence intensity distribution curve of the third region of interest corresponding to the non-overlapping area is substantially the same.

[0090] In some embodiments, when the fluorescence intensity distribution curve of the second ROI does not have an overflowing non-overlapping area outside the fluorescence intensity distribution curve of the third ROI, it can be evaluated that the fluorescence intensity distribution curve of the third ROI substantially surrounds the fluorescence intensity distribution curve of the second ROI.

[0091] In some embodiments, a trained cell classification model may be used to perform a first classification (and a subsequent second classification) on cells in the biological sample.

[0092] In some embodiments, the cell classification model can be obtained by training a corresponding model using a set of nucleolar morphological features associated with each cell type, such as by using labeled learning data from cells of known categories. The morphological differences between nucleoli in different cell types can be analyzed using machine learning to perform linear relationship analysis, thereby improving classification accuracy.

[0093] In some embodiments, the cell classification model can perform image recognition on the microscopic image data to determine the cell type (eg, vascular smooth muscle cell, neuroblastoma cell, etc.) of the cells in the biological sample and its corresponding nucleolar features.

[0094] To further ensure the accuracy of cell classification, a second classification can be performed on the C-type cells: In some embodiments, the method may further include: S105 performing a second classification on the type C cells in the biological sample; In some embodiments, the second classification method includes: The ratio of the area of ​​the second region of interest to the area of ​​the third region of interest in the C-type cells is analyzed to determine whether it is abnormal, so as to perform a second classification on the C-type cells in the biological sample.

[0095] In some embodiments, if the ratio of the area of ​​the second region of interest to the third region of interest in the C-type cells is abnormal, the classification of the C-type cells is adjusted to B-type cells; if it is not abnormal, the classification of the C-type cells is adjusted to A-type cells.

[0096] In some embodiments, when the ratio of the area of ​​the second region of interest to the area of ​​the third region of interest shows an increasing trend, it is considered abnormal.

[0097] This embodiment found that when the nucleolar function is "abnormal" (for example, the nucleolar function is impaired), the area of ​​the second region of interest and the third region of interest increases, and the ratio shows an increasing trend. This increasing trend can be used to accurately perform a second classification of type C cells, thereby obtaining a more accurate number and / or ratio of type A cells and type B cells, effectively performing quantitative and qualitative analysis of cells, and accurately and sensitively reflecting and evaluating the functional status of the tissue or organ corresponding to the biological sample. It is particularly suitable for biological samples with strong heterogeneity that lack specific markers.

[0098] In some embodiments, type A cells are cells with normal nucleolar function relative to cells of that type.

[0099] In some embodiments, a B-type cell is a cell that has abnormal nucleolar function relative to that type of cell itself.

[0100] In some embodiments, type C cells refer to cells that may have abnormal nucleolar function relative to the type of cells themselves.

[0101] In other words, in some special application scenarios (for example, when the accuracy requirements for the classification results are relatively low or when there are higher requirements for the classification speed), the subsequent S106 can be directly performed based on the results of the first classification (i.e., class C cells are regarded as class B cells) without performing step S105.

[0102] S106 Calculate the number and / or ratio of the type A cells to the type B cells in the biological sample, and then evaluate the functional status of the tissue or organ corresponding to the biological sample.

[0103] It is important to emphasize that this Example demonstrates that this upward trend applies not only to non-tumor biological samples but also to tumor biological samples. Specifically, while the nucleolar morphology of non-tumor and tumor biological samples exhibits different characteristics, when nucleolar function is "abnormal," the trend in the ratio of the area of ​​the second region of interest to the area of ​​the third region of interest is similar. Therefore, the method provided in this Example does not rely on specific cell activity markers; rather, it can achieve highly accurate and precise cell classification based solely on nucleolar-related phenotypic characteristics, providing a new detection method for diseases that lack specific markers.

[0104] In addition, the method provided in this embodiment can accurately and sensitively reflect changes in nucleolus-related phenotypes in the early stages of the disease to predict cell function (such as hyperactivity of cell proliferation or cell failure) and characterize changes in tissue health status (such as vascular intimal hyperplasia, tumors or degeneration, etc.) and the pathological state of diseased tissues, thereby providing support for early diagnosis of the disease.

[0105] In some embodiments, the biological sample includes a biological sample that has been treated with a drug and a biological sample that has not been treated with a drug. It should be understood that, unlike existing cell classification methods based on nucleolar morphology, which generally can only perform qualitative analysis, the method provided in this embodiment can accurately classify cell subpopulations in a biological sample with high reproducibility. Therefore, quantitative analysis can be performed based on the cell classification results of this embodiment. This is particularly reflected in the fact that the method provided in this embodiment can also be used for high-throughput drug screening and drug efficacy assessment.

[0106] In some embodiments, the effectiveness of a drug (e.g., whether it has a tumor-suppressing effect) can be determined by comparing the number and / or ratio of type A cells to type B cells in a drug-treated biological sample with that in an untreated biological sample. In some embodiments, high-throughput drug screening can also be performed by comparing the number and / or ratio of type A cells to type B cells in biological samples treated with different drug concentration gradients and / or biological samples treated with different drugs.

[0107] In some embodiments, when the biological sample is a tumor sample (e.g., neuroblastoma), the method may further include: S107 analyzing the average diameter of the spots formed by the dense fiber component marker in the second region of interest in the B-type cells; when the average diameter is less than a first threshold, the biological sample is assessed to be in a malignant proliferation state.

[0108] In some embodiments, the first threshold comprises 0.5-2 μm.

[0109] Example 2 like Figure 2As shown, this embodiment provides a cell classification and function evaluation system 100 based on nucleolus morphology corresponding to the method of embodiment 1, which includes: An image acquisition module 102 is configured to acquire microscopic image data of a biological sample; In some embodiments, the microscopic image data includes optical microscope images with an objective lens of 40 times or more and / or super-resolution microscopic image data.

[0110] In some embodiments, immunofluorescence images of biological samples can be collected in batches using a high-throughput imaging system to obtain the microscopic image data. In some embodiments, imaging requirements include using a 40x objective lens or greater to image glass-bottomed or other ultra-thin bottom plates (e.g., 0.17-0.19 mm).

[0111] In some embodiments, the biological sample comprises a tissue or cell culture comprising at least one cell.

[0112] In some embodiments, the biological sample is labeled with nucleolar markers, and the nucleolar markers include at least a dense fiber component marker and a granular component marker.

[0113] In some embodiments, the dense fiber component marker comprises fibrinogen (FBL).

[0114] In some embodiments, the granule component marker comprises nucleophosmin 1 (NPM1).

[0115] In some embodiments, the biological sample is further labeled with a nuclear marker.

[0116] In some embodiments, the nuclear marker comprises 4',6-diamidino-2-phenylindole (DAPI).

[0117] The first processing module 104 is configured to obtain a first region of interest, a second region of interest, and a third region of interest from the microscopic image data.

[0118] In some embodiments, the first region of interest is a cell nucleus region delineated by a fluorescent channel corresponding to a cell nucleus marker.

[0119] In some embodiments, the second region of interest is a nucleolus region delineated by a fluorescent channel corresponding to a dense fiber component marker.

[0120] In some embodiments, the third region of interest is a nucleolus region delineated using a fluorescent channel corresponding to a particle component marker.

[0121] A second processing module 106 is configured to extract nucleolus feature data of the second region of interest and the third region of interest, wherein the nucleolus feature data includes line scan data and region area data; A first classification module 108 is configured to perform a first classification on cells in the biological sample based on the line scan data; In some embodiments, the line scan data includes fluorescence intensity distribution curves of the second region of interest and the third region of interest.

[0122] In some embodiments, the first classification method includes: When the fluorescence intensity distribution curve of the third region of interest surrounds the fluorescence intensity distribution curve of the second region of interest, classifying the cell as a type A cell; When the fluorescence intensity distribution curve of the second region of interest overflows outside the fluorescence intensity distribution curve of the third region of interest, the cell is classified as a type C cell.

[0123] In some embodiments, the spillover assessment method includes: Analyze whether the fluorescence intensity change trends of the fluorescence intensity distribution curve of the second region of interest and the fluorescence intensity distribution curve of the third region of interest at the first and last non-overlapping areas formed outside the fluorescence intensity distribution curve of the third region of interest are the same; if not, overflow occurs; if so, no overflow occurs.

[0124] A second classification module 110 is configured to perform a second classification on the type C cells in the biological sample; In some embodiments, the second classification method includes: The ratio of the area of ​​the second region of interest to the area of ​​the third region of interest in the C-type cells is analyzed to determine whether it is abnormal, so as to perform a second classification on the C-type cells in the biological sample.

[0125] In some embodiments, if the ratio of the area of ​​the second region of interest to the third region of interest in the C-type cells is abnormal, the classification of the C-type cells is adjusted to B-type cells; if it is not abnormal, the classification of the C-type cells is adjusted to A-type cells.

[0126] In some embodiments, when the ratio of the area of ​​the second region of interest to the area of ​​the third region of interest shows an increasing trend, it is considered abnormal.

[0127] In some embodiments, the first classification module 108 and the second classification module 110 can be based on a trained cell classification model.

[0128] In some embodiments, the cell classification model can be obtained by training a corresponding model using a set of nucleolar morphological features associated with each cell type, such as by using labeled learning data from cells of known categories. The morphological differences between nucleoli in different cell types can be analyzed using machine learning to perform linear relationship analysis, thereby improving classification accuracy.

[0129] The analysis module 112 is configured to calculate the number and / or ratio of the type A cells and the type B cells in the biological sample, and then evaluate the functional status of the tissue or organ corresponding to the biological sample.

[0130] In some embodiments, the biological sample includes a biological sample treated with a drug and a biological sample not treated with a drug.

[0131] In some embodiments, the analysis module 112 is further configured to analyze the average diameter of the spots formed by the dense fiber component marker in the B cells. When the average diameter is less than a first threshold, the biological sample is assessed to be in a malignant proliferation state.

[0132] In some embodiments, the first threshold comprises 0.5-2 μm.

[0133] Example 3 The method of Example 1 was used to detect rat vascular smooth muscle cell cultures. Figure 3 Figure (a) shows super-resolution microscopic images and line scans of normal and drug-treated rat vascular smooth muscle cells fluorescently labeled with FBL and NPM1. The results show that in the normal rat vascular smooth muscle cells, the fluorescence distribution curve corresponding to NPM1 essentially encompasses the fluorescence distribution curve corresponding to FBL, representing the well-defined outer and middle nucleoli of rat vascular smooth muscle cells. Drug intervention causes the fluorescence distribution curve corresponding to FBL to overflow outside the fluorescence distribution curve corresponding to NPM1 (typically indicated by the arrow).

[0134] Figure 3 Figure b shows the areas of FBL and NPM1 fluorescently labeled vascular smooth muscle cells in the normal group and drug-treated groups. The areas of FBL and NPM1 and the ratio of NPM1 to FBL in the drug-treated group increased. Figure 3 Figure c shows that the method and system provided by the present invention can be combined with machine learning (deep learning). Through the high efficiency of machine learning algorithms, complex cell image data (microscopic image data) can be quickly processed and analyzed, thereby improving the accuracy and precision of cell classification. Figure 3As shown in c, the nucleoli of type A cells are regular, while the nucleoli of type B cells are irregular.

[0135] Figure 3 Figure d shows the changes in the number and ratio of type A cells and type B cells in rat vascular smooth muscle cell culture samples before and after drug intervention. The results showed that after drug intervention (such as Figure 3 (Indicated by the yellow column in d) in rat vascular smooth muscle cell culture samples, the number and proportion of type A cells decreased, while the number and proportion of type B cells increased. These results suggest that the drug's intervention inhibits nucleolar function, leading to impaired cellular function.

[0136] Example 4 The method of Example 1 was used to detect the culture of the neuroblastoma cell line SK-N-SH. Figure 4 Figure a shows super-resolution microscopic images of neuroblastoma cells in the normal group and drug-treated group that were fluorescently labeled with FBL and NPM1. Figure 4 The corresponding line scan results are shown in panel b. The results show that the fluorescence distribution curve corresponding to NPM1 in normal neuroblastoma cells essentially encompasses the fluorescence distribution curve corresponding to FBL, indicating the clear outer and middle compartments of the normal nucleolus in neuroblastoma cells. Drug intervention causes the fluorescence distribution curve corresponding to FBL to overflow outside the fluorescence distribution curve corresponding to NPM1 (typically indicated by the arrow).

[0137] Figure 4 Figure c shows the areas of neuroblastoma cells in the normal group and drug-treated group fluorescently labeled with FBL and NPM1. The areas of FBL and NPM1 and the ratio of NPM1 to FBL areas in the neuroblastoma cells of the drug-treated group increased.

[0138] Figure 4 Figure d shows the changes in the number and ratio of type A and type B cells in SK-N-SH culture samples before and after drug intervention. The results show that after drug intervention, the number and ratio of type A cells in SK-N-SH culture samples decreased, while the number and ratio of type B cells (cells with "necklace" nucleoli) increased. These results suggest that the drug inhibits nucleolar function, thereby suppressing neuroblastoma cell proliferation.

[0139] Example 5 The method of Example 1 was used to detect neuroblastoma tumor tissue. The cell proliferation marker Ki67 was used to classify the tumor tissue into three levels of cell proliferation (~0%, 5.743%, 38.227%). Figure 5As shown in Figure 2, the diameter of FBL spots is significantly reduced in highly proliferating neuroblastoma cells. These results indicate that the malignancy of a tumor can be determined based on the diameter of the FBL spots.

[0140] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better embodiment. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), including a number of instructions for enabling a computer terminal (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present invention.

[0141] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which are protected by the present invention.

Claims

1. A method for cell classification and function assessment based on nucleolar morphology, characterized in that: The following steps are involved: S101: Acquire microscopic image data of a biological sample, wherein the biological sample comprises a tissue or cell culture comprising at least one cell; the biological sample is labeled with a nucleolar marker, and the nucleolar marker comprises at least a dense fiber component marker and a granular component marker; S102: Acquiring a first region of interest, a second region of interest, and a third region of interest from the microscopic image data; wherein the first region of interest is a cell nucleus region delineated using a fluorescence channel corresponding to a cell nucleus marker; the second region of interest is a nucleolus region delineated using a fluorescence channel corresponding to a dense fiber component marker; and the third region of interest is a nucleolus region delineated using a fluorescence channel corresponding to a granular component marker; S103 extracting nucleolus feature data of the second region of interest and the third region of interest, wherein the nucleolus feature data includes line scan data and region area data; S104: Perform a first classification of cells in the biological sample based on the line scan data; the line scan data includes fluorescence intensity distribution curves of the second region of interest and the third region of interest; the first classification method includes: classifying the cells as type A cells when the fluorescence intensity distribution curve of the third region of interest surrounds the fluorescence intensity distribution curve of the second region of interest; and classifying the cells as type C cells when the fluorescence intensity distribution curve of the second region of interest overflows outside the fluorescence intensity distribution curve of the third region of interest; wherein the type C cells are considered type B cells; S105 Calculate the number and / or ratio of the type A cells to the type B cells in the biological sample, and then evaluate the functional status of the tissue or organ corresponding to the biological sample.

2. The method according to claim 1, wherein The dense fiber component marker includes fibrin, and the granular component marker includes nucleophosmin 1.

3. The method according to claim 1, wherein The overflow evaluation method includes: analyzing whether the fluorescence intensity change trends of the fluorescence intensity distribution curve of the second region of interest and the fluorescence intensity distribution curve of the third region of interest in the non-overlapping area formed outside the fluorescence intensity distribution curve of the third region of interest are the same in the non-overlapping area where the fluorescence intensity distribution curve of the second region of interest appears for the first time and the non-overlapping area where the last time appears are the same; if not, overflow occurs; if so, no overflow occurs.

4. The method according to claim 1, wherein When the ratio of the area of ​​the second ROI to the third ROI shows an increasing trend, it is considered abnormal.

5. The method according to claim 1, wherein The biological samples include biological samples treated with drugs and biological samples not treated with drugs.

6. The method according to claim 1, wherein The method further includes S107 analyzing an average diameter of spots formed by the dense fiber component marker in the second region of interest in the B-type cells; when the average diameter is less than a first threshold, the biological sample is assessed as being in a malignant proliferation state.

7. A cell classification and function evaluation system based on nucleolar morphology, characterized in that: The system comprises: An image acquisition module configured to acquire microscopic image data of a biological sample; wherein the biological sample comprises a tissue or cell culture comprising at least one cell; and the biological sample is labeled with a nucleolar marker, wherein the nucleolar marker comprises at least a dense fiber component marker and a granular component marker; a first processing module configured to acquire a first region of interest, a second region of interest, and a third region of interest from the microscopic image data; wherein the first region of interest is a cell nucleus region delineated by a fluorescence channel corresponding to a cell nucleus marker; the second region of interest is a nucleolus region delineated by a fluorescence channel corresponding to a dense fiber component marker; and the third region of interest is a nucleolus region delineated by a fluorescence channel corresponding to a granular component marker; A second processing module is configured to extract nucleolus feature data of the second region of interest and the third region of interest, wherein the nucleolus feature data includes line scan data and region area data; a first classification module configured to perform a first classification of cells in the biological sample based on the line scan data; the line scan data including fluorescence intensity distribution curves of the second region of interest and the third region of interest; a method for the first classification comprising: classifying the cells as type A cells when the fluorescence intensity distribution curve of the third region of interest surrounds the fluorescence intensity distribution curve of the second region of interest; and classifying the cells as type C cells when the fluorescence intensity distribution curve of the second region of interest overflows outside the fluorescence intensity distribution curve of the third region of interest; wherein the type C cells are considered type B cells; The analysis module is configured to calculate the number and / or ratio of the type A cells and the type B cells in the biological sample, and then evaluate the functional status of the tissue or organ corresponding to the biological sample.

8. The system according to claim 7, wherein: The dense fiber component marker includes fibrin, and the granular component marker includes nucleophosmin 1.

9. The system according to claim 7, wherein: When the ratio of the area of ​​the second ROI to the third ROI shows an increasing trend, it is considered abnormal.

10. The system according to claim 7, wherein: The biological samples include biological samples treated with drugs and biological samples not treated with drugs.

Images