Population-based cell classification

Abstract

Cell classification may be performed using a system that receives a set of images comprising representations of a plurality of cells and performs various actions of each of the cells. The actions may include determining imaging parameters for each of the cells, generating a cluster assignment for each of the cells, and providing a type of each of the cells based on the cluster assignment.

Description

[0001] Cross - reference to related applications

[0002] This application claims priority to U.S. Provisional Patent Application No. 63 / 434,658, entitled "Population - Based Cell Classification", filed on December 22, 2022, with the United States Patent and Trademark Office, and is a non - provisional patent application thereof. This application is incorporated herein by reference in its entirety. Background of the Invention

[0003] Blood cell analysis is one of the most commonly performed medical tests for providing a profile of a patient's health. A blood sample can be extracted from a patient and stored in a test tube containing an anticoagulant to prevent clotting. A whole - blood sample typically contains three main categories of blood cells, and the above - mentioned three main categories of blood cells include red blood cells (erythrocytes), white blood cells (leukocytes), and platelets (thrombocytes). Each category can be further divided into sub - classes of members. For example, the five main types or sub - classes of white blood cells (WBCs) have different shapes and functions. White blood cells can include neutrophils, lymphocytes, monocytes, eosinophils, and basophils. There are also sub - classes of red blood cell types. The appearance of the particles in the sample can vary depending on pathological conditions, cell maturity, and other reasons. Red blood cell sub - classes can include reticulocytes and nucleated red blood cells.

[0004] Traditionally, particle classification systems have utilized non - imaging techniques, such as measuring light scattering from laser illumination or changes in impedance when a particle passes through a pore. While these techniques can be effective, they do have drawbacks. Therefore, there is a need for improved classification systems, such as systems that can classify particles based on information extracted from images of those particles. Summary of the Invention

[0005] Described herein are devices, systems, and methods for classifying objects such as cells in images captured by an analyzer, such as a biometric system that captures images of blood cells from a blood sample.

[0006] Illustrative implementations of this technique involve a system that includes a processor and a non-transitory computer-readable medium. This medium can store instructions that are operable to perform a set of actions when executed by the processor. Such actions can include receiving a set of images, where the set of images includes representations of a plurality of cells. Such actions can also include a set of actions performed for each cell from the plurality of cells. The set of actions can include determining one or more imaging parameters for the cell based on the application of an image analysis algorithm. The set of actions can also include generating a cluster assignment for the cell by assigning the cell to a cluster from a plurality of clusters based on at least one imaging parameter of one or more imaging parameters for the cell and for a population of cells including the plurality of cells. In other words, the set of actions can include generating a cluster assignment for the cell by assigning the cell to a cluster from a plurality of clusters based on at least one imaging parameter of one or more imaging parameters for the cell and at least one imaging parameter of one or more imaging parameters for each cell of a population of cells including the plurality of cells. Such a set of actions can also include providing a type of the cell based on the cluster assignment of the cell.

[0007] While multiple examples are described herein, other examples of the described subject matter will become apparent to those skilled in the art from the following detailed description and the accompanying drawings that illustrate and describe illustrative examples of the disclosed subject matter. As will be recognized, the disclosed subject matter is capable of modification in various aspects, all of which do not depart from the spirit and scope of the described subject matter. Accordingly, the drawings and detailed description are to be regarded as illustrative rather than restrictive. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] While this specification concludes with claims that particularly point out and distinctly claim the invention, it is believed that the invention will be better understood from the following description of certain examples in conjunction with the accompanying drawings, in which like reference numerals represent like elements and in which:

[0009] Figure 1 is a schematic illustration, partly in section and not to scale, showing an exemplary flow cell and operational aspects of a high optical resolution imaging device for sample image analysis using digital image processing.

[0010] Figure 2 shows a slide-based visual inspection system in which aspects of the disclosed technology can be used.

[0011] Figure 3 shows a process that can be used to classify cells.

[0012] Figure 4 shows a process that can be used to assign cells to clusters using a gating algorithm.

[0013] Figure 5 Shows a process that can be used to assign cells to clusters using a gating algorithm.

[0014] Figure 6 Shows corresponding to Figure 7 An exemplary nuclear mask of the cell image.

[0015] Figure 7 Shows a slide-based image of a cell including a cell nucleus.

[0016] Figure 8 Shows for Figure 7 A cell mask of the image of the cells in.

[0017] Figure 9 Shows Figure 7 A histogram of each pixel value of the image of the cells in.

[0018] The drawings are not intended to be limiting in any way, and it is contemplated that various embodiments of the present invention may be implemented in various other ways, including those not necessarily depicted in the drawings. The drawings, which are incorporated in and form a part of the specification, illustrate several aspects of the present invention and, together with the description, serve to explain the principles of the present invention; however, it should be understood that the present invention is not limited to the exact arrangements shown. Detailed Description

[0019] The present disclosure relates to devices, systems, compositions, and methods for analyzing samples containing particles. In one embodiment, the present invention relates to an automated particle imaging system that includes an analyzer, which can be, for example, a vision analyzer. In some embodiments, the vision analyzer may further include a processor for facilitating the automated analysis of images.

[0020] According to some aspects of the present disclosure, a system for obtaining an image of a sample including particles suspended in a liquid may be provided that includes a vision analyzer. Such a system may be useful, for example, in characterizing particles in biological fluids, such as detecting and quantifying red blood cells, reticulocytes, nucleated red blood cells, platelets, and white blood cells, including white blood cell differential counts, classification, and subclassification, and analysis. Other similar uses are also contemplated, such as characterizing blood cells from other fluids.

[0021] Although other types of body fluid samples can be used, the classification of blood cells in a blood sample is an exemplary application particularly suitable for this subject matter. For example, aspects of the disclosed technology can be used to analyze non-blood body fluid samples that include blood cells (e.g., white blood cells and / or red blood cells), such as serum, bone marrow, lavage fluid, effusion, exudate, cerebrospinal fluid, pleural effusion, peritoneal fluid, and amniotic fluid. The sample can also be a solid tissue sample, such as a biopsy sample that has been processed to produce a cell suspension. The sample can also be a suspension obtained from processing a fecal sample. The sample can also be a laboratory sample or production line sample that includes particles, such as a cell culture sample. The term sample can be used to refer to a sample obtained from a patient or laboratory or any fragment, portion, or aliquot thereof. In some processes, the sample can be diluted, divided into multiple parts, or stained.

[0022] In some aspects, the sample is presented, imaged, and analyzed in an automated manner. In the case of a blood sample, a suitable diluent or salt solution can be used to substantially dilute the sample, which reduces the extent to which the view of some cells may be hidden by other cells in an undiluted or less diluted sample. Reagents that enhance the contrast of certain cellular aspects can be used to process the cells, such as using a permeabilizing agent to make the cell membrane permeable and allowing a histological stain to adhere thereto and reveal features such as granules and cell nuclei. In some cases, it may be desirable to stain an aliquot of the sample to count and characterize particles including reticulocytes, nucleated red blood cells, and platelets, and for white blood cell classification, characterization, and analysis. In other cases, a sample containing red blood cells can be diluted before and / or while being imaged in a flow cell or otherwise.

[0023] Details of sample preparation devices and methods for sample dilution, permeabilization, and histological staining can generally be accomplished using precision pumps and valves operated by one or more programmable controllers. Examples can be found in patents such as U.S. Patent No. 7,319,907. Similarly, techniques for differentiating certain cell classes and / or subclasses by attributes such as relative size and color can be found in U.S. Patent No. 5,436,978 related to white blood cells. The disclosures of these patents are incorporated herein by reference in their entirety.

[0024] I. Imaging System

[0025] Now turning to the drawings, Figure 1An exemplary flow cell 22 is schematically shown which is used to convey a sample liquid through an observation region 23 of a high optical resolution imaging device 24 in a configuration for imaging microscopic particles in a sample flow stream 32 using digital image processing. The flow cell 22 is coupled to a source 25 of the sample liquid which may have been subjected to processing such as contact with a particulate contrast agent composition and heating. The flow cell 22 is also coupled to one or more sources 27 of a particulate and / or intracellular organelle alignment liquid (PIOAL) (such as a transparent glycerol solution having a viscosity greater than that of the sample liquid).

[0026] The sample liquid is injected through a flat opening at the distal end 28 of a sample feed tube 29 and into the interior of the flow cell 22 at a point where the PIOAL flow has been substantially established, thereby resulting in a stable and symmetric laminar flow of the PIOAL surrounding / encircling (e.g., circumferentially surrounding / encircling in a circular cross-section arrangement, or surrounding / encircling multiple sides in a non-circular (e.g., rectangular) cross-section arrangement) a ribbon-like sample stream. The sample stream and the PIOAL stream may be supplied by a precision metering pump which moves the PIOAL along with the injected sample liquid through a substantially narrowed flow channel. The PIOAL encapsulates and compresses the sample liquid in the narrowed region 21 of the flow channel. Thus, the reduction in the flow channel thickness at region 21 can contribute to the geometric focusing of the sample stream 32. The sample liquid band 32 is encapsulated and carried with the PIOAL downstream of the narrowed region 21 and passes in front of or otherwise through the observation region 23 of the high optical resolution imaging device 24 where an image is collected, for example, using a CCD 48. Flow imaging is performed in this manner in which images are collected from the flowing sample stream and the cellular material contained therein. The processor 18 may receive pixel data from the CCD 48 as input. The sample liquid band flows into the discharge port 33 together with the PIOAL.

[0027] As shown herein, the narrowed region 21 may have a proximal flow channel portion 21a having a proximal thickness PT and a distal flow channel portion 21b having a distal thickness DT such that the distal thickness DT is less than the proximal thickness PT. Thus, the sample liquid may be injected through the distal end 28 of the sample tube 29 at a location distal to the proximal portion 21a and proximal to the distal portion 21b. Thus, when the PIOAL flow is compressed by region 21, the sample liquid may enter the PIOAL sheath where the sample liquid injection tube has a distal outlet port through which the sample liquid is injected into the flowing sheath liquid, the distal outlet port being defined by a reduction in the flow channel dimensions of the flow cell.

[0028] The digital high optical resolution imaging device 24 with the objective lens 46 is oriented along the optical axis intersecting the strip-shaped sample stream 32. The relative distance between the objective lens 46 and the flow cell 33 is varied by the operation of the motor driver 54 for resolving and collecting focused digital images on the photoelectric sensor array. Additional information regarding the construction and operation of an exemplary flow cell such as Figure 1 shown is provided in U.S. Patent No. 9,322,752, entitled "Flowcell Systems and Methods for Particle Analysis in Blood Samples", filed on March 17, 2014, the disclosure of which is incorporated herein by reference in its entirety.

[0029] Aspects of the disclosed technology may also be applied in contexts other than flow cell systems such as Figure 1 shown. For example, Figure 2 a slide-based visual inspection system 200 is shown in which aspects of the disclosed technology may be used. In the Figure 2 system shown, a slide 202 containing a sample such as a blood sample is placed in a slide holder 204. The slide holder 204 may be adapted to hold multiple slides or only one slide, as Figure 2 shown. An image capture device 206 including an optical system 208 and an image sensor 210 is adapted to capture image data depicting the sample in the slide 202.

[0030] The image data captured by the image capture device 206 can be transmitted to the image processing device 212. The image processing device 112 can be an external device such as a personal computer connected to the image capture device 206. Alternatively, the image processing device 212 can be incorporated into the image capture device 206. The image processing device 212 can include a processor 214 associated with a memory 216, and the processor 214 is configured to determine the changes required to determine the difference between the actual focus and the correct focus of the image capture device 206. When the difference is determined, an instruction can be transmitted to the steering motor system 218. The steering motor system 218 can change the distance z between the slide 202 and the optical system 208 based on the instruction from the image processing device 212. Descriptions of methods that can be used for focusing using this type of setup are provided in U.S. Provisional Patent Application 63 / 291,044, titled "Autofocusing Through Multi-Layer Processing," filed on December 17, 2021, U.S. Patent 9,857,361, titled "Flowcell, sheath fluid, and autofocus systems and methods for particle analysis in urine samples," filed on March 17, 2014, U.S. Patent 10,705,008, titled "Autofocus systems and methods for particle analysis in blood samples," filed on March 17, 2014, U.S. Patent 10,705,011, titled "Dynamic focus system and methods," filed on October 5, 2017, and International Application WO2023 / 150064, titled "Measure image quality of blood cell images," filed on January 27, 2023. The disclosures of each of the above applications are incorporated herein by reference in their entirety.

[0031] II. Data Processing

[0032] Data captured by systems such as Figure 1 and Figure 2 shown in Figure 3 can be subjected to various types of processing. Advanced methods that can be performed in such processing are shown in

[0033] A. Image Analysis

[0034] Initially, in Figure 3During the process, a representation of more than 301 cells will be received. This can include, for example, a processor receiving one or more images, each image including a representation of a plurality of cells (e.g., as can be captured in a slide-based system such as Figure 2 shown). Alternatively, receiving a representation of more than 301 cells can include a processor receiving a plurality of images, each of the plurality of images including a representation of only a single cell (e.g., as can be captured by a flow imaging system based on a flow cell such as Figure 1 shown). Once the representation has been received at 301, the method can proceed to perform 302 image analysis to obtain data that can be used in subsequent data processing. Exemplary data that can be obtained as a result of performing 302 image analysis is provided in Table 1 below.

[0035]

[0036]

[0037] Table 1: Exemplary data that can be obtained through image analysis

[0038] Other types of image analysis can also be included when performing 302 image analysis, such as using a cell separation algorithm (e.g., an algorithm for thresholding an image captured by a flow cell-based system to identify portions of the image that do not represent cells) to separate the representation of cells and / or using a method such as that described in U.S. Patent No. 4,538,299, titled "method and apparatus for locating the boundary of an object," issued on August 27, 1985, the disclosure of which is incorporated herein by reference in its entirety.

[0039] Additionally, in some embodiments, performing 302 image analysis can also include generating 303 masks. For example, the following masks can potentially be generated at 303.

[0040]

[0041]

[0042] Table 2: Exemplary masks

[0043] To illustrate what may be involved in generating 303 masks, Figure 7 is shown such as using as Figure 1 or Figure 2The apparatus shown in can capture an image of cells 700. In some embodiments, the image can include a background portion 710 and a foreground portion 720. The foreground portion 720 can represent blood cells that can be further segmented into cell parts (e.g., cytoplasm 722 and nucleus 724). These components of the blood cell image 700 (e.g., cytoplasm 722 and nucleus 724) can be described or defined using corresponding masks. For example, briefly referring to Figure 8 , an illustrative example of a cell mask 800 (corresponding to cell_mask from Table 2) is shown. In some embodiments, the mask can include binary values. For example, and as shown, the cell mask 800 can be approximately the same size as the original image and uses a series of 0s and / or 1s to represent each pixel. In some embodiments, the value 1 can indicate that the corresponding pixel belongs to the cell 810, while the value 0 can indicate that the corresponding pixel does not belong to the cell (e.g., background mask 820). For example, such a cell mask can be created by normalizing the pixel values in an RGB image and using the normalized values to create a histogram, and then using the histogram to define a threshold that separates cells from non-cell pixels. An example of such a histogram is shown in Figure 9 , Figure 9 depicts a graphical representation 900 of the normalized value 910 of the pixels (e.g., the minimum of the red value, green value, and blue value smoothed by averaging the values of adjacent pixels within a smoothing window and projecting onto the 0:1 range) plotted against the total number 920 of pixels in the image having the corresponding value.

[0044] Other masks can be generated in a manner similar to that discussed above for the cell mask, but the details of a particular mask will vary depending on the nature of the mask itself. For example, the left threshold 930 value in the histogram of Figure 9 can be used to define as shown in Figure 6The nuclear mask 600 (corresponding to nucleus_mask in Table 2) shown in [FIGURE] reflects the fact that pixels depicting the nucleus may be darker than pixels depicting other parts of the cell. As another example, in some cases, another mask and / or parameters of another mask can be used as input to generate a mask. For example, a dark mask can be a mask of pixels within the cell mask that have values below a first darkness threshold, while a black mask can be a mask of pixels within the cell mask that are below a second darkness threshold, which is lower than the first darkness threshold. It is also possible that, in some cases, in addition to or as an alternative to the above masks, one or more sets of other pixels can be generated. These can include, for example, boundary pixels and / or pixels corresponding to one or more of the listed masks (e.g., cytoplasmic pixels, black pixels). Additionally, in some cases, generating the 303 mask can include processing beyond pixel values. For example, in some cases, once a mask for a structure (e.g., a cell) that is expected to be continuous is created, an anomaly removal step can be performed, where small (e.g., less than a specified number of pixels) holes in the mask will be removed (i.e., considered to be included regardless of pixel value), such that mask creation is not adversely affected by imaging or other artifacts. Thus, the above discussion regarding generating the mask 603 should be understood as illustrative only and not restrictive.

[0045] B. Determine Parameters

[0046] In Figure 3 the method of, after completing the execution of 302 image analysis (which may include generating the 303 mask), the data obtained from the image analysis can be used to determine 304 imaging parameters. Examples of these types of parameters and how these values 304 can be determined are provided in Table 3 below:

[0047]

[0048]

[0049]

[0050] Table 3: Exemplary Parameters and Determination Methods

[0051] C. Assign Clusters

[0052] In the case where the 304 parameters are determined, a process such as Figure 3 shown in [FIGURE] can continue to assign 305 each cell in the cell to a cluster based on the imaging parameters. This can be done, for example, by applying a gating algorithm to the cell representation, where the previously determined parameters will be evaluated in sequence to assign the cells to various clusters. For illustration, consider Figure 4, which shows how this gating method can be applied to clusters of cells in a blood sample.

[0053] In Figure 4 's method, initially, an examination 401 can be applied to determine whether a cell is a white blood cell. This can be done using parameters such as the darkness of the cell, the ratio of blue to red of the cell, and light absorption caused by the presence of the cell. For example, histograms of the darkness, blue - red ratio, and / or light absorption values of the cells can be created, and the lowest point in each of these histograms can be defined as a threshold for determining whether a cell is a white blood cell. Using these thresholds, any cell that does not have a darkness value, blue - red ratio, or light absorption value higher than the cut - off value considered to be that of a white blood cell (e.g., a cell where the blue channel divided by the red channel in its cell mask is less than 1, and where the average value of the L channel of the pixels in its cell mask divided by the average value of the L values of its boundary pixels in the L*a*b* color space is less than 1) will be assigned 402 to the non - white - blood - cell cluster. Otherwise, the cell can be further analyzed to determine what type of white blood cell it is.

[0054] In Figure 4 's method, if it is determined 401 that the cell is a white blood cell, it can be further determined 403 whether the cell is an eosinophil. This can be done using parameters such as the darkness of the cell granules and the blue color of the cell granules and / or cytoplasm. For example, Gaussian mixture model clustering can be used to cluster the cells in a two - dimensional space defined by the blue color of the cell granules. Then, the cell with the fewest blue granules can be considered an eosinophil and assigned 404 to the eosinophil cluster. A similar process can be applied to identify neutrophils. That is, after the appropriate cells have been assigned 404 to the eosinophil cluster, it can be further determined 405 whether the remaining cells are neutrophils by applying Gaussian mixture model clustering to cluster those cells in a three - dimensional space defined by the size of the cells and the darkness and blue color of their cytoplasm. Then, cells that are relatively large with fewer blue granules or cytoplasm at the same darkness level (e.g., the difference between the V value in the HSV color space and 1 for the applicable cells) can be considered neutrophils and assigned 406 to the neutrophil cluster.

[0055] In Figure 4In the method, after appropriately allocating the cells 406 to the neutrophil cluster, the remaining cells can be further determined 407 as to whether they are basophils. This can be done, for example, by identifying the number of dark blue granules around the cell nucleus and using OPTICS clustering (as described in Kriegel, Hans-Peter; Kroger, Peer; Sander, Jorg; and Zimek, Arthur, Density-Based Clustering, Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery. 1(3): 231-240, the disclosure of which is incorporated herein by reference in its entirety) or a threshold algorithm to identify those cells that represent those including a relatively large number of such granules. Those cells can then be allocated 408 to the basophil cluster, and the remaining cells can be further determined 409 as to whether they are monocytes or lymphocytes. This can be accomplished by using k-means clustering to separate the cells based on the size of the cells and / or the size of their cell nuclei. Smaller cells (e.g., cells having a normalized_y value <= 0.42) can be allocated 410 to the lymphocyte cluster, while the remaining cells are allocated 411 to the monocyte cluster. In this way, based on the relationship of the relevant parameters of a particular cell to those of the entire cell population under consideration, a series of binary classifications are used to classify the cells in the sample as non-WBC, eosinophils, neutrophils, basophils, lymphocytes, and monocytes.

[0056] D. Classification

[0057] Now return to Figure 3In the discussion, after the cells have been assigned to clusters, the cluster assignment can be used to provide a classification of the cells. This can be done, for example, by providing an output to the user (e.g., on a display) that shows the cells on a scatter plot, where different colors are used to distinguish the types of cells from each other. Similarly, in some cases, the classification of the cells can be provided to an evaluation algorithm, which can use the classification to determine if there may be some anomalies in the image evaluation. For example, the evaluation algorithm can be configured with data representing the expected characteristics and relative positions of different clusters based on images captured under ideal conditions. In such cases, the cell classification can be provided to the evaluation algorithm, which can compare the relative positions and characteristics of the clusters created from the image with the expected relative positions and characteristics. Based on this comparison, the evaluation algorithm can indicate whether the cell representation was captured under less than ideal lighting (e.g., because the cells in the cluster appear darker than expected), whether the sample being imaged was incorrectly stained (e.g., because the cluster appears very light, which may cause the cluster to be offset along the axis representing the blue or darkness of the cells), or whether the image was captured at an incorrect focus (e.g., because the cell sizes in the various clusters appear larger than expected). The evaluation algorithm can also quantify the difference between the actual conditions and the ideal conditions, for example, by measuring the offset of various parameters across the entire population of cells represented by the cluster. This can then be used to provide various outputs, such as flagging samples that may need to be re-imaged due to anomalies exceeding a certain threshold magnitude, notifying the operator that he or she should adjust the imaging components he or she is using to improve future image quality. Other applications of this type of information, such as providing an assessment of the effectiveness of an analysis system when it is being developed, are also possible and will be readily apparent to those of ordinary skill in the art in light of the present disclosure. Accordingly, the examples provided should be understood as being merely illustrative and not restrictive.

[0058] The relative distribution profiles implemented based on the present disclosure can offer advantages over per-cell classification systems (e.g., identifying and classifying each cell one at a time). For example, the chance of misclassification due to poor imaging conditions (e.g., out-of-focus camera device, poor lighting, or staining problems) is reduced because the classification system can identify the relative population distribution and look for population clusters within the overall population distribution to classify the cells. In addition to the above examples, instead of comparing the observed population distribution with the expected population distribution, the classification system can be trained to look for pools or groupings of data and then segment the cell population in this way (e.g., without a step of comparing with an expected distribution profile).

[0059] III. Variations

[0060] Other variations and implementations of the disclosed technology are also possible. For example, in some cases, cell separation techniques such as those described in the context of flow imaging can be applied to images from a slide-based imaging system, and then individual cells can be grouped into populations (potentially along with other cells from different images of the same sample) as part of the received cell representation 301. Additionally, in some implementations, other types of classification can also be performed in some cases. For example, in some cases, before determining whether a cell still to be classified at 403 is an eosinophil, it can be determined ( Figure 4 not shown) whether the cell is a megakaryocyte based on whether the cell has a relatively low granularity and / or a blue-to-green ratio. Similarly, in some cases, it can be determined whether one or more groups of cells should be identified as immature granulocytes. This can be done, for example, by using a process such as that shown in Figure 5 . In this process, rather than simply determining at 405 whether a cell should be considered a neutrophil as shown in Figure 4 , it is determined at 505 whether the cell should be considered a potential neutrophil or an immature granulocyte. For example, the same Gaussian mixture module clustering described in the context of Figure 4 can be used for this determination at 505, and then an additional determination at 506 is made by separating the cluster into two sub-clusters based on darkness and blue whether the cells of the cluster are neutrophils. In this type of method, relatively less bright and / or bluer cells can be assigned at 507 to the cluster of neutrophils, while the remaining cells from the cluster identified in the previous determination at 505 can be assigned at 508 to the immature granulocyte (IG) cluster. Additional steps can also be included. For example, in some cases, certain pixels can be excluded from consideration, such as black pixels (which can be defined as having a low (e.g., less than 170) value in the blue channel of the RGB representation and a small (e.g., less than 35) difference between the red and blue and red and green channels) before the cell-to-cluster assignment 302.

[0061] As another example of a variant type, although different types of clustering were described above in the context of Figure 4 and Figure 5 , other methods can also be (or alternatively) used, such as agglomerative clustering or distribution-based clustering. Similarly, in some cases, rather than using, for example, that shown in Figure 4Rather than the gating method described in the context of assigning cells to clusters associated with a particular type of cell, cells can be clustered using parameters such as those described in Table 3. Then, after all cells are assigned to clusters, the clusters can be associated with different cell types based on their relative positions in the feature space. Based on the present disclosure, other variations, such as using sequential clustering and only determining the features of the cells (e.g., the masks shown in Table 2, or the features shown in Table 3), since those features used for the determination of the cells should be assigned to clusters, will be quite apparent to those of ordinary skill in the art and can be implemented without undue experimentation.

[0062] Another example of a type of variation is in the parameters, where the above parameters are determined 304 and / or used to assign 305 cells to clusters. For example, in some systems implemented based on the present disclosure, criteria for various color values or combinations of color values in an RGB image, or other types of values from other images (e.g., the L value of an L*a*b* image), may also be determined 304. These parameters (and in some cases other parameters) can be used for clustering and classification, which may be different from the clustering and classification described above. For example, in some embodiments, depending on the median BR-value (i.e., the blue channel divided by the red channel) and the median NV-value (i.e., the average of the L-channel values in the cell mask divided by the average of the L-channel values in the boundary mask), cells can be clustered into white blood cell (WBC) or non-WBC clusters. As another example, the system can cluster cells into eosinophil clusters by using a Gaussian mixture model (GMM) in a two-dimensional feature space, where the first dimension is the average of the blue channel values of the pixels in the black mask divided by the average of the blue channel values in the background mask, and the second dimension is the average of the blue channel minus the red channel of any pixel in the black mask where the blue channel is below a threshold (e.g., a value defined based on the percentile of the mask). As another example, a system implemented based on the present disclosure can cluster cells into basophil clusters by using a GMM in a two-dimensional feature space, where the first dimension is nucleus_contrast1 and the second dimension is nucleus_contrast2. In some cases, cells can be classified into lymphocyte-monocyte clusters by applying K-means clustering to a two-dimensional feature space, where the features are normalized_x and normalized_y with two target clusters. The cluster with a smaller normalized_x can be regarded as the lymphocyte-monocyte cluster. Then, by considering calls with normalized_y values <= 0.42 as lymphocytes and cells with normalized_y values > 0.42 as monocytes, this cluster can be further decomposed into lymphocytes and monocytes. Clustering can also be performed to separate neutrophils from immature granulocytes, where GMM is used to cluster neutrophils and immature granulocytes in a two-dimensional feature space, where the first dimension is nonblack_mask_Nv, and the second dimension is nonblack_mask_blueness2 (e.g., where nonblack_mask is used to help identify a specific cell type, such as immature granulocytes). As yet another example, in some cases, clusters can be generated in an n-dimensional space defined by the determined 304 features, and cells can be classified based on the relative positions of the clusters in this space.

[0063] The type of hardware used to implement the disclosed technology is also a possible variation. For example, in some cases, image analysis tools and other operations described in the context of Figures 3 to 9 can be applied using hardware incorporated into or local to the analyzer, which includes imaging components such as those shown in Figures 1 to 2 . For example, features can be extracted from an image captured by an analyzer using a computer and used to classify cells depicted in an image captured by an analyzer using a computer, which is connected to the analyzer using a USB cable or over a local area network. Alternatively, in some cases, an image captured by the analyzer (or information extracted from such an image) can be transmitted over a wide area network to remote processing equipment (such as a cloud server), and that equipment can be used to process the image (or information extracted from those images prior to transmission) and classify the cells shown in the image.

[0064] IV. Examples

[0065] As a further illustration of potential implementations and applications of the disclosed technology, the following examples are provided in non-exhaustive ways in which the teachings herein can be combined or applied. It should be understood that the following examples are not intended to limit the scope of any claims that may be presented at any time in this application or in subsequent filings of this application. The following examples are provided solely for illustrative purposes and are not intended to waive any rights. It is contemplated that the various teachings herein can be arranged and applied in many other ways. It is also contemplated that some variations may omit certain features mentioned in the following examples. Therefore, no aspect or feature mentioned below should be considered critical unless explicitly stated otherwise later by the inventor or the heirs of the inventor's interests. If any claims are presented in this application or in subsequent filings related to this application that include additional features beyond those mentioned below, those additional features should not be presumed to have been added for any reason related to patentability.

[0066] Example 1

[0067] A method for cell classification, comprising: receiving a set of images, wherein the set of images includes representations of a plurality of cells; for each cell from the plurality of cells: determining one or more imaging parameters for the cell based on the application of an image analysis algorithm; generating a cluster assignment for the cell by assigning the cell to a cluster from a plurality of clusters; and providing the type of the cell based on the cluster assignment of the cell.

[0068] Example 2

[0069] The method according to Example 1, wherein, for each cell from the plurality of cells, a gating algorithm is used to perform generation of the cluster assignment for the cell, and the gating algorithm is configured to: assign the cell to a cluster based on at least one imaging parameter among one or more imaging parameters for both the cell and a population including the plurality of cells.

[0070] Example 3

[0071] The method according to Example 2, wherein the gating algorithm is configured to assign cells to clusters corresponding to cell types by: first determining cells assigned to non-white blood cell clusters; next determining cells assigned to eosinophil clusters; next determining cells assigned to neutrophil clusters; next determining cells assigned to basophil clusters; next determining cells assigned to lymphocyte clusters; and next determining cells assigned to monocyte clusters.

[0072] Example 4

[0073] The method according to any one of Examples 2 to 3, wherein, for each cell from the plurality of cells, determining one or more imaging parameters for the cell based on the application of an image analysis algorithm includes: generating a plurality of masks for the cell; and for each mask of the plurality of masks, for each feature of the plurality of features, determining a value of the feature for the mask.

[0074] Example 5

[0075] The method according to Example 4, wherein, for each cell from the plurality of cells, the plurality of clusters includes a first cluster, and the gating algorithm is configured to determine whether to assign the cell to the first cluster based on: a first imaging parameter of pixel values in the RGB color space of the cell mask from the cell; and a second imaging parameter of pixel values in the L*a*b color space of the cell mask of the cell.

[0076] Example 6

[0077] The method according to any one of Examples 4 to 5, wherein, for each cell from the plurality of cells, the plurality of masks of the cell includes: a cell mask; a nuclear mask; a cytoplasmic mask; a dark mask; a black mask; and an IG mask.

[0078] Example 7

[0079] The method according to any one of claims 1 to 6, wherein the one or more imaging parameters include at least two of the following: size, shape, darkness, color, and internal structure.

[0080] Example 8

[0081] The method according to any one of claims 1 to 7, wherein the set of actions includes: extracting representations of a plurality of cells from the set of images based on the application of a cell separation algorithm.

[0082] Example 9

[0083] The method according to any one of Examples 1 to 8, wherein the method includes capturing the set of images by performing actions that include: establishing an alignment fluid flow from an alignment fluid reservoir into a flow cell; generating a sample flow that includes a sample fluid and an alignment fluid sheath on an opposite side of the sample fluid based on injecting the sample fluid from a channel into the alignment fluid flow; and capturing the set of images using an imaging device focused on an observation region of the flow cell while the sample flow is flowing through the observation region of the flow cell.

[0084] Example 10

[0085] The method according to any one of Examples 1 to 9, wherein the plurality of cells includes blood cells.

[0086] Example 11

[0087] The method according to Example 10, wherein: the plurality of cells includes a first white blood cell and a second white blood cell; and the type of the first white blood cell is different from the type of the second white blood cell.

[0088] Example 12

[0089] A system for cell classification, comprising: one or more processors; and a non-transitory computer-readable medium having instructions stored thereon that are operable to perform a set of actions when executed by the one or more processors, the set of actions including: receiving a set of images, wherein the set of images includes representations of a plurality of cells; for each cell from the plurality of cells: determining one or more imaging parameters for the cell based on the application of an image analysis algorithm; generating a cluster assignment for the cell by assigning the cell to a cluster from a plurality of clusters based on at least one imaging parameter of one or more imaging parameters for the cell and for a population including the plurality of cells; and providing a type of the cell based on the cluster assignment of the cell.

[0090] Example 13

[0091] The system according to Example 12, wherein for each cell from the plurality of cells, the instructions are operable to generate a cluster assignment for the cell using a gating algorithm when executed by the one or more processors, the gating algorithm being configured to: assign the cell to a cluster based on at least one imaging parameter of one or more imaging parameters for the cell and for a population including the plurality of cells.

[0092] Example 14

[0093] The system according to Example 13, wherein the gating algorithm is configured to assign cells to clusters corresponding to cell types by: first determining cells assigned to non-white blood cell clusters; next determining cells assigned to eosinophil clusters; next determining cells assigned to neutrophil clusters; next determining cells assigned to basophil clusters; next determining cells assigned to lymphocyte clusters; and next determining cells assigned to monocyte clusters.

[0094] Example 15

[0095] The system according to any one of Examples 13 to 14, wherein for each cell from a plurality of cells, determining one or more imaging parameters for the cell based on the application of an image analysis algorithm includes: generating a plurality of masks for the cell; and for each mask of the plurality of masks, for each feature of a plurality of features, determining a value of the feature for the mask.

[0096] Example 16

[0097] The system according to Example 15, wherein for each cell from a plurality of cells, the plurality of clusters includes a first cluster, and the gating algorithm is configured to determine whether to assign the cell to the first cluster based on: a first imaging parameter based on pixel values in the RGB color space of the cell mask of the cell; and a second imaging parameter based on pixel values in the L*a*b color space of the cell mask of the cell.

[0098] Example 17

[0099] The system according to any one of Examples 15 to 16, wherein for each cell from a plurality of cells, the plurality of masks of the cell includes: a cell mask; a nuclear mask; a cytoplasmic mask; a dark mask; a black mask; and an IG mask.

[0100] Example 18

[0101] The system according to any one of Examples 12 to 17, wherein the one or more imaging parameters include at least two of the following: size, shape, darkness, color, and internal structure.

[0102] Example 19

[0103] The system according to any one of Examples 12 to 18, wherein the set of actions includes: extracting representations of a plurality of cells from the set of images based on the application of a cell separation algorithm.

[0104] Example 20

[0105] The system according to any one of Examples 12 to 19, wherein: the system includes a flow cell and an imaging device configured to capture the set of images as a sample flow passes through an observation region of the flow cell via flow imaging of the sample flow; and each image from the set of images includes a representation of a single cell from a plurality of cells.

[0106] Example 21

[0107] The system according to Example 20, wherein the system includes: an alignment fluid reservoir in fluid communication with an observation region of the flow cell, and a channel adapted to inject a sample fluid into an alignment fluid stream to form a sample flow including the sample fluid and an alignment fluid sheath surrounding the sample fluid.

[0108] Example 22

[0109] The system according to any one of Examples 12 to 21, wherein the plurality of cells includes blood cells.

[0110] Example 23

[0111] The system according to Example 22, wherein: the plurality of cells includes a first white blood cell and a second white blood cell; and the type of the first white blood cell is different from the type of the second white blood cell.

[0112] Example 24

[0113] A machine, comprising: an imaging device; and means for classifying cells in an image captured by the imaging device.

[0114] Example 25

[0115] A system for cell classification, comprising: a flow cell configured to allow a sample flow to pass therethrough; an imaging device configured to capture one or more images of a plurality of cells from the sample flow; a processor; and a non-transitory computer-readable medium having instructions stored thereon that are operable, when executed by the processor, to perform a set of actions including: determining, using an image analysis algorithm, one or more imaging parameters for the plurality of cells; identifying a plurality of data clusters within a population distribution of at least one imaging parameter of the one or more imaging parameters; and assigning a cell type to each cell of the plurality of cells based on data clusters from the plurality of data clusters, wherein the cell is included in the data cluster from the plurality of data clusters.

[0116] Example 26

[0117] A system for cell classification, comprising: a processor; a non-transitory computer-readable medium storing computer-executable code, the computer-executable code including instructions that, when executed by the processor, perform a set of actions, the set of actions including: receiving one or more images depicting a plurality of cells; determining one or more imaging parameters for the plurality of cells; identifying a plurality of data clusters within a population distribution, the population distribution being established based on at least one of the one or more imaging parameters; and assigning a corresponding cell type for the data cluster to each of the plurality of data clusters from the plurality of data clusters.

[0118] Example 27

[0119] A system for cell classification, comprising: a flow cell configured to allow a sample stream to flow therethrough; an imaging device configured to capture one or more images of a plurality of cells in the sample stream; a processor; a non-transitory computer-readable medium storing computer-executable code, the computer-executable code including instructions that, when executed by the processor, perform a set of actions, the set of actions including: determining one or more imaging parameters for the plurality of cells; identifying a plurality of data clusters within a population distribution, the population distribution being established based on at least one of the one or more imaging parameters; and assigning a corresponding cell type for the data cluster to each of the plurality of data clusters from the plurality of data clusters.

[0120] Example 28

[0121] A method for cell classification, comprising: allowing a sample stream to flow through a flow cell; when the sample stream flows through an observation region of the flow cell, capturing one or more images of a plurality of cells from the sample stream using an imaging device; determining one or more imaging parameters for the plurality of cells using an image analysis algorithm; identifying a plurality of data clusters within a population distribution of at least one of the one or more imaging parameters; and assigning a cell type to each of the plurality of cells based on data clusters from the plurality of data clusters, wherein the cell is included in a data cluster from the plurality of data clusters.

[0122] Example 29

[0123] A method for cell classification, comprising: using an imaging device to capture one or more images of a plurality of cells; determining one or more imaging parameters for the plurality of cells; identifying a plurality of data clusters within a population distribution, the population distribution being established based on at least one of the one or more imaging parameters; and assigning a corresponding cell type for the data cluster to each of the plurality of data clusters from the plurality of data clusters.

[0124] Example 30

[0125] A method of cell classification, comprising: flowing a sample stream through a flow cell; when the sample stream flows through an observation region of the flow cell, capturing one or more images of a plurality of cells in the sample stream using an imaging device; determining one or more imaging parameters for the plurality of cells; identifying a plurality of data clusters within a population distribution, the population distribution being established based on at least one of the one or more imaging parameters; and assigning a corresponding cell type for the data cluster to each of the plurality of data clusters.

[0126] V. Explanation

[0127] Each of the calculations or operations described herein can be performed using a computer or other processor having hardware, software, and / or firmware. The various method steps can be performed by modules, and the modules can include any of a variety of digital and / or analog data processing hardware and / or software arranged to perform the method steps described herein. Optionally, the module includes data processing hardware adapted to perform one or more of these steps by having appropriate machine programming code associated therewith, and modules for two or more steps (or parts of two or more steps) are integrated into a single processor board or separated into different processor boards in any of a variety of integrated and / or distributed processing architectures. These methods and systems will generally employ a tangible medium that bears machine-readable code having instructions for performing the above-described method steps. Suitable tangible media can include memory (including volatile memory and / or non-volatile memory), storage media (such as magnetic recording on floppy disks, hard disks, magnetic tapes, etc.; magnetic recording on optical memories such as CDs, CD-R / W, CD-ROM, DVDs, etc.; or any other digital or analog storage media), and the like.

[0128] All patents, patent publications, patent applications, journal articles, books, technical references, etc. discussed in the instant disclosure are hereby incorporated by reference in their entirety for all purposes.

[0129] Different arrangements of the components depicted in the accompanying drawings or described above, as well as components and steps not shown or described, are possible. Similarly, some features and sub-combinations are useful and can be employed without reference to other features and sub-combinations. Embodiments of the present invention have been described for illustrative rather than restrictive purposes, and alternative embodiments will be apparent to the reader of this patent. In some cases, method steps or operations may be performed or implemented in a different order, or operations may be added, deleted, or modified. It is understood that in certain aspects of the present invention, a single component may be replaced by multiple components, and multiple components may be replaced by a single component, to provide an element or structure or to perform a given one or more functions. Such replacements are considered to be within the scope of the present invention, except where such replacement would be inoperable for practicing certain embodiments of the present invention. Accordingly, the claims should not be regarded as limited to the examples, drawings, embodiments, and descriptions provided above, but should be understood to have the scope provided when their terms are given the broadest reasonable interpretation provided by a general dictionary, unless a term or phrase is indicated to have a specific meaning under the heading "Explicit Definition," in which case it should be understood to have that meaning when used in the claims.

[0130] Explicit Definition

[0131] It should be understood that in the above examples and claims, a statement that something is "based on" something else should be understood to mean that it is determined at least in part by the thing it is indicated as being based on. To indicate that something must be determined entirely based on something else, it is described as being "entirely based on" whatever it must be entirely determined by.

[0132] It should be understood that in the above examples and claims, the phrase "means for classifying cells in an image captured by an imaging device" is a means-plus-function limitation provided in 35 U.S.C. § 112(f), where the function is "classifying cells in an image captured by an imaging device," and the corresponding structure is a computer configured to use an algorithm as shown in 、 Figure 3 and Figure 4 Figure 5 and described in the accompanying specification.

[0133] It should be understood that in the above examples and claims, the term "group" should be understood to mean one or more things grouped together.

Claims

1. A method for cell classification, comprising: Receiving a set of images, wherein the set of images includes representations of a plurality of cells; For each cell from the plurality of cells: Determining one or more imaging parameters for the cell based on the application of an image analysis algorithm; Generating a cluster assignment for the cell by assigning the cell to a cluster from a plurality of clusters; and Providing the type of the cell based on the cluster assignment of the cell.

2. The method according to claim 1, wherein For each cell from the plurality of cells, a gating algorithm is used to perform generating the cluster assignment for the cell, the gating algorithm being configured to: assign the cell to a cluster based on at least one imaging parameter of the one or more imaging parameters for both the cell and a population including the plurality of cells.

3. The method according to claim 2, wherein The gating algorithm is configured to assign cells to clusters corresponding to cell types by: First determining cells assigned to non - white blood cell clusters; Next determining cells assigned to eosinophil clusters; Next determining cells assigned to neutrophil clusters; Next determining cells assigned to basophil clusters; Next determining cells assigned to lymphocyte clusters; And Next determining cells assigned to monocyte clusters.

4. The method according to claim 2, wherein For each cell from the plurality of cells, determining one or more imaging parameters for the cell based on the application of the image analysis algorithm includes: Generating a plurality of masks for the cell; and For each mask of the plurality of masks, for each of a plurality of features, determining a value of the feature for the mask.

5. The method according to claim 4, wherein For each cell from the plurality of cells, the plurality of clusters includes a first cluster, and the gating algorithm is configured to determine whether to assign the cell to the first cluster based on: A first imaging parameter based on pixel values in the RGB color space of the cell mask of the cell; and A second imaging parameter based on pixel values in the L*a*b color space of the cell mask of the cell.

6. The method according to claim 4, wherein For each cell from the plurality of cells, the plurality of masks of the cell includes: A cell mask; A nuclear mask; A cytoplasmic mask.

7. The method according to claim 1, wherein The one or more imaging parameters include at least two of the following: size, shape, darkness, color, and internal structure.

8. The method according to claim 1, wherein The set of actions includes: extracting representations of the plurality of cells from the set of images based on the application of a cell separation algorithm.

9. The method according to claim 1, wherein The method includes capturing the set of images by performing actions, the actions including: Establishing an alignment fluid flow from an alignment fluid reservoir into a flow cell; Generating a sample flow including a sample fluid and an alignment fluid sheath surrounding the sample fluid based on injecting the sample fluid from a channel into the alignment fluid flow; and When the sample flow is flowing through an observation region of the flow cell, using an imaging device focused on the observation region of the flow cell to capture the set of images.

10. The method according to claim 1, wherein The plurality of cells includes blood cells.

11. The method according to claim 10, wherein: The plurality of cells includes a first white blood cell and a second white blood cell; and The type of the first white blood cell is different from the type of the second white blood cell.

12. A system for cell classification, comprising: One or more processors; And A non-transitory computer-readable medium having instructions stored thereon that are operative to perform a set of actions when executed by the one or more processors, the set of actions including: Receiving a set of images, wherein the set of images includes representations of a plurality of cells; For each cell from the plurality of cells: Determining one or more imaging parameters for the cell based on the application of an image analysis algorithm; Generating a cluster assignment for the cell by assigning the cell to a cluster from a plurality of clusters based on at least one of the one or more imaging parameters for the cell and for a population including the plurality of cells; and Providing a type of the cell based on the cluster assignment of the cell.

13. The system according to claim 12, wherein, For each cell from the plurality of cells, the instructions are operative to generate a cluster assignment for the cell when executed by the one or more processors using a gating algorithm configured to: assign the cell to a cluster based on at least one of the one or more imaging parameters for the cell and for a population including the plurality of cells.

14. The system according to claim 13, wherein, The gating algorithm is configured to assign cells to clusters corresponding to cell types by: First determining cells assigned to a non-white blood cell cluster; Next determining cells assigned to an eosinophil cluster; Next determining cells assigned to a neutrophil cluster; Next determining cells assigned to a basophil cluster; Next determining cells assigned to a lymphocyte cluster; And Next determining cells assigned to a monocyte cluster.

15. The system according to claim 13, wherein, For each cell from the plurality of cells, determining one or more imaging parameters for the cell based on the application of the image analysis algorithm includes: Generating a plurality of masks for the cell; and For each mask of the plurality of masks, for each of a plurality of features, determining a value of the feature for the mask.

16. The system according to claim 15, wherein, For each cell from the plurality of cells, the plurality of clusters includes a first cluster, and the gating algorithm is configured to determine whether to assign the cell to the first cluster based on: A first imaging parameter based on pixel values in the RGB color space of a cell mask of the cell; and A second imaging parameter based on pixel values in the L*a*b color space of the cell mask of the cell.

17. The system according to claim 15, wherein For each cell from the plurality of cells, the plurality of masks of the cell include: A cell mask; A nuclear mask; A cytoplasmic mask.

18. The system according to claim 12, wherein The one or more imaging parameters include at least two of the following: size, shape, darkness, color, and internal structure.

19. The system according to claim 12, wherein, The set of actions includes: extracting representations of the plurality of cells from the set of images based on the application of a cell separation algorithm.

20. The system according to claim 12, wherein: The system includes a flow cell and an imaging device configured to capture the set of images via flow imaging of a sample stream as the sample stream flows through an observation region of the flow cell; And Each image from the set of images includes a representation of a single cell from the plurality of cells.

21. The system according to claim 20, wherein, The system includes: an alignment fluid reservoir in fluid communication with an observation region of the flow cell, and a channel adapted to inject a sample fluid into an alignment fluid stream to form a sample stream including the sample fluid and an alignment fluid sheath surrounding the sample fluid.

22. The system according to claim 12, wherein, The plurality of cells includes blood cells.

23. The system according to claim 22, wherein: The plurality of cells includes a first white blood cell and a second white blood cell; and The type of the first white blood cell is different from the type of the second white blood cell.

24. A machine, comprising: An imaging device; And A device for classifying cells in an image captured by the imaging device.

Images