Flow imaging for rare cell types

Abstract

A blood imaging method for identification of circulating tumor cells may comprise drawing a first aliquot from one or more mixing chambers of an analyzer, the first aliquot including a plurality of cells. Such a method may also comprise, after drawing the first aliquot from the one or more mixing chambers, flowing the first aliquot through a flowcell of the analyzer. Such a method may also comprise utilizing an image capture device to capture an image of a cell as the first aliquot flows through the flowcell. Such a method may also comprise drawing a second aliquot from the one or more mixing chambers as the first aliquot flows through the flowcell. Such a method may also comprise analyzing the image of the cell using a classification algorithm, wherein the classification algorithm is configured to classify the cell according to a plurality of classes, wherein the plurality of classes includes a class for circulating tumor cells.

Description

FLOW IMAGING FOR RARE CELL TYPESPRIORITY

[0001] This claims the benefit of U. S. Prov. Pat. App. No. 63 / 751,761 entitled “Flow Imaging for Rare Cell Types,” filed on January 30, 2025, the disclosure of which is incorporated by reference herein, in its entirety.BACKGROUND

[0002] In biological analyzers (e.g., body fluid analyzers and hematology analyzers), biological samples are driven through a flowcell for analysis. These biological samples may include blood samples and / or other body fluid samples. The biological samples are typically surrounded by sheath fluid to hydrodynamically focus the biological sample into a thin stream through the analysis region of the flowcell. Flow imaging for blood analysis utilizes a flowcell and camera to capture images of cells / particles as they pass through an imaging region of the flowcell.

[0003] Biological sample analysis typically involves driving a single aliquot of a biological sample through a flowcell. This single aliquot is typically a low-volume sample and meant to provide clinically significant blood information for a patient (e.g., red or white blood cell count, platelet count, or white blood cell differential information). However, in certain circumstances such as rare cell detection, increased sensitivity may be required due to the limited presence of such rare cells in a blood sample which could necessitate a larger aliquot amount, or multiple aliquots being needed to identify such rare cells (e.g. circulating tumor cells / CTCs). Accordingly, there is a need for biological sample analysis technology which can efficiently provide increased sensitivity through extended measurements of particular biological samples.SUMMARY

[0004] Biological imaging systems and methods for identification of circulating tumor cells are described herein. In some embodiments, such systems and methods may be used to analyze blood. In some embodiments, such systems and methods may be used to analyze other body - 1 -0133788.0817287 4901-6446-8614v10fluids such as synovial fluid, cerebrospinal fluid, urine, bone marrow aspirate, or other bodily fluids / substances.

[0005] According to a first aspect, a blood imaging system for identification of circulating tumor cells is disclosed. The blood imaging system may comprise a flowcell; one or more mixing chambers; an image capture device; and one or more processors. The one or more processors may be configured to perform a set of acts comprising drawing a first aliquot from the one or more mixing chambers and, after drawing the first aliquot from the one or more mixing chambers, flowing the first aliquot through the flowcell. The acts the one or more processors may be configured to perform may also comprise, utilizing the image capture device to capture an image of a cell as the first aliquot flows through the flowcell, and drawing a second aliquot from the one or more mixing chambers as the first aliquot flows through the flowcell. The acts the one or more processors may be configured to perform may also comprise analyzing image of the cell using a classification algorithm, wherein the classification algorithm is configured to classify the cell according to a plurality of classes, wherein the plurality of classes includes a class for circulating tumor cells.

[0006] In some examples, drawing the first aliquot from the one or more mixing chambers comprises drawing the first aliquot into a first pending sample line; and drawing the second aliquot from the one or more mixing chambers comprises drawing the second aliquot into a second pending sample line.

[0007] In some examples, the set of acts the one or more processors may be configured to perform may further comprise: after the first aliquot is no longer flowing through the flowcell, flowing the second aliquot through the flowcell; and determining whether to clean the flowcell between flowing the first aliquot through the flowcell and flowing the second aliquot through the flowcell based on whether the first aliquot and the second aliquot are from a single blood sample.

[0008] In some examples, the system comprises a sheath fluid source; a diluent source; and one or more pairs of valves. In some such examples, each pair of valves from the one or more pairs - 2 -0133788.0817287 4901-6446-8614v10of valves is separated by a corresponding pending sample line and is operable to selectively provide fluid communication between: the sheath fluid source and the flowcell; and the diluent source and the flowcell.

[0009] In some examples, for each pair of valves from the one or more pairs of valves, that pair of valves providing fluid communication between the sheath fluid source and the flowcell comprises: at the first valve from that pair of valves, providing fluid communication between the flowcell and the pending sample line corresponding to that pair of valves; and at the second valve from that pair of valves, providing fluid communication between the pending sample line corresponding to that pair of valves and the sheath fluid source.

[0010] In some examples, for each pair of valves from the one or more pairs of valves, that pair of valves providing fluid communication between the sheath fluid source and the flowcell comprises: at the first valve from that pair of valves, blocking fluid communication between the pending sample line corresponding to that pair of valves and the one or more mixing chambers; and at the second valve from that pair of valves, blocking fluid communication between the pending sample line corresponding to that pair of valves and the diluent source.

[0011] In some examples, for each pair of valves from the one or more pairs of valves, that pair of valves providing fluid communication between the diluent source and the flowcell comprises: at the first valve from that pair of valves, providing fluid communication between the flowcell and the pending sample line corresponding to that pair of valves; and at the second valve from that pair of valves, providing fluid communication between the pending sample line for that pair of valves and the diluent source.

[0012] In some examples, for each pair of valves from the one or more pairs of valves, that pair of valves providing fluid communication between the diluent source and the flowcell comprises: at the first valve from that pair of valves, blocking fluid communication between the pending sample line corresponding to that pair of valves and the one or more mixing chambers; and at the second valve from that pair of valves, blocking fluid communication between the pending sample line corresponding to that pair of valves and the sheath fluid source.- 3 -0133788.0817287 4901-6446-8614v10

[0013] In some examples, each pair of valves from the one or more pairs of valves is configured to be used in drawing an aliquot from the one or more mixing chambers by performing acts comprising, at the first valve from that pair of valves, establishing fluid communication between the one or more mixing chambers and the pending sample line corresponding to that pair of valves.

[0014] In some examples, each pair of valves from the one or more pairs of valves is configured to be used in cleaning the flowcell by performing acts comprising: at the first valve from that pair of valves, providing fluid communication between the flowcell and the pending sample line corresponding to that pair of valves; and at the second valve from that pair of valves, providing fluid communication between the pending sample line corresponding to that pair of valves and the diluent source.

[0015] In some examples, each pair of valves from the one or more pairs of valves is configured to be used in cleaning the one or more mixing chambers by performing acts comprising: at the first valve from that pair of valves, providing fluid communication between at least one of the one or more mixing chambers and the pending sample line corresponding to that pair of valves; and at the second valve from that pair of valves, providing fluid communication between the pending sample line corresponding to that pair of valves and the diluent source.

[0016] In some examples, the first pending sample line is the pending sample line corresponding to a first pair of valves from the one or more pairs of valves; the second pending sample line is the pending sample line corresponding to a second pair of valves from the one or more pairs of valves; cleaning the flowcell between flowing the first aliquot through the flowcell and flowing the second aliquot through the flowcell comprises, after flowing the first aliquot through the flowcell, using the first pair of valves in cleaning the flowcell; and the set of acts comprises, before drawing the second aliquot from the one or more mixing chambers, using the second pair of valves in cleaning the one or more mixing chambers.

[0017] In some examples, the first pending sample line is the pending sample line corresponding to a first pair of valves from the one or more pairs of valves; the second pending sample line is the - 4 -0133788.0817287 4901-6446-8614v10pending sample line corresponding to a second pair of valves from the one or more pairs of valves; drawing the first aliquot from the one or more mixing chambers comprises drawing the first aliquot from the one or more mixing chambers using the first pair of valves; flowing the first aliquot through the flowcell comprises the first pair of valves providing fluid communication between the sheath fluid source and the flowcell; drawing the second aliquot from the one or more mixing chambers comprises drawing the second aliquot from the one or more mixing chambers using the second pair of valves; and the set of acts comprises, after drawing the second aliquot from the one or more mixing chambers, flowing the second aliquot through the flowcell by performing acts comprising the second pair of valves providing fluid communication between the sheath fluid source and the flowcell.

[0018] In some examples, the plurality of classes includes a first class which has a plurality of subclasses, wherein the plurality of subclasses includes the class for circulating tumor cells; and the classification algorithm is configured to classify an input cell image as a circulating tumor cell image by performing a set of classification acts. In some such examples, the set of classification acts may comprise: classifying the input cell image into the first class; and classifying the input cell image into the class for circulating tumor cells after classifying the input cell image into the first class.

[0019] In some examples, the classification algorithm is configured to classify an input cell image as a circulating tumor cell image based on a size of a cell depicted in the input cell image.

[0020] According to a second aspect, a biological imaging method for identification of circulating tumor cells is provided. Such a biological imaging method may comprise drawing a first aliquot from one or more mixing chambers of an analyzer, the first aliquot including a plurality of cells. Such a method may also comprise, after drawing the first aliquot from the one or more mixing chambers, flowing the first aliquot through a flowcell of the analyzer. Such a method may also comprise, utilizing an image capture device to capture an image of a cell as the first aliquot flows through the flowcell; and drawing a second aliquot from the one or more mixing chambers as the first aliquot flows through the flowcell. Such a method may also comprise - 5 -0133788.0817287 4901-6446-8614v10analyzing the image of the cell using a classification algorithm, wherein the classification algorithm is configured to classify the cell according to at least one of a plurality of classes, wherein the plurality of classes includes a class for circulating tumor cells.

[0021] In some examples, drawing the first aliquot from the one or more mixing chambers comprises drawing the first aliquot into a first pending sample line; and drawing the second aliquot from the one or more mixing chambers comprises drawing the second aliquot into a second pending sample line.

[0022] In some examples, the method comprises: after the first aliquot is no longer flowing through the flowcell, flowing the second aliquot through the flowcell; and determining whether to clean the flowcell between flowing the first aliquot through the flowcell and flowing the second aliquot through the flowcell based on whether the first aliquot and the second aliquot are from a single blood sample.

[0023] In some examples, the analyzer may comprise a sheath fluid source, a diluent source, and one or more pairs of valves. In some such examples, each pair of valves form the one or more pairs of valves may be separated by a corresponding pending sample line and may be operable to selectively provide fluid communication between the sheath fluid source and the flowcell, and the diluent source and the flowcell.

[0024] In some examples, for each pair of valves from the one or more pairs of valves, that pair of valves providing fluid communication between the sheath fluid source and the flowcell comprises: at the first valve from that pair of valves, providing fluid communication between the flowcell and the pending sample line corresponding to that pair of valves; and at the second valve from that pair of valves, providing fluid communication between the pending sample line corresponding to that pair of valves and the sheath fluid source.

[0025] In some examples, for each pair of valves from the one or more pairs of valves, that pair of valves providing fluid communication between the sheath fluid source and the flowcell comprises: at the first valve from that pair of valves, blocking fluid communication between - 6 -0133788.0817287 4901-6446-8614v10the pending sample line corresponding to that pair of valves and the one or more mixing chambers; and at the second valve from that pair of valves, blocking fluid communication between the pending sample line corresponding to that pair of valves and the diluent source.

[0026] In some examples, for each pair of valves form the one or more pairs of valves, that pair of valves providing fluid communication between the diluent source and the flowcell comprises: at the first valve from that pair of valves, providing fluid communication between the flowcell and the pending sample line corresponding to that pair of valves; and at the second valve from that pair of valves, providing fluid communication between the pending sample line for that pair of valves and the diluent source.

[0027] In some examples, for each pair of valves from the one or more pairs of valves, that pair of valves providing fluid communication between the diluent source and the flowcell comprises: at the first valve from that pair of valves, blocking fluid communication between the pending sample line corresponding to that pair of valves and the one or more mixing chambers; and at the second valve from that pair of valves, blocking fluid communication between the pending sample line corresponding to that pair of valves and the sheath fluid source.

[0028] In some examples, each pair of valves from the one or more pairs of valves is configured to be used in drawing an aliquot from the one or more mixing chambers by performing acts comprising, at the first valve from that pair of valves, establishing fluid communication between the one or more mixing chambers and the pending sample line corresponding to that pair of valves.

[0029] In some examples, each pair of valves from the one or more pairs of valves is configured to be used in cleaning the flowcell by performing acts comprising: at the first valve from that pair of valves, providing fluid communication between the flowcell and the pending sample line corresponding to that pair of valves; and at the second valve from that pair of valves, providing fluid communication between the pending sample line corresponding to that pair of valves and the diluent source.- 7 -0133788.0817287 4901-6446-8614v10

[0030] In some examples, each pair of valves from the one or more pairs of valves is configured to be used in cleaning the one or more mixing chambers by performing acts comprising: at the first valve from that pair of valves, providing fluid communication between at least one of the one or more mixing chambers and the pending sample line corresponding to that pair of valves; and at the second valve from that pair of valves, providing fluid communication between the pending sample line corresponding to that pair of valves and the diluent source.

[0031] In some examples, the first pending sample line is the pending sample line corresponding to a first pair of valves from the one or more pairs of valves; the second pending sample line is the pending sample line corresponding to a second pair of valves from the one or more pairs of valves; cleaning the flowcell between flowing the first aliquot through the flowcell and flowing the second aliquot through the flowcell comprises, after flowing the first aliquot through the flowcell, using the first pair of valves in cleaning the flowcell; and the set of acts comprises, before drawing the second aliquot from the one or more mixing chambers, using the second pair of valves in cleaning the one or more mixing chambers.

[0032] In some examples, the first pending sample line is the pending sample line corresponding to a first pair of valves from the one or more pairs of valves; the second pending sample line is the pending sample line corresponding to a second pair of valves from the one or more pairs of valves; drawing the first aliquot from the one or more mixing chambers comprises drawing the first aliquot from the one or more mixing chambers using the first pair of valves; flowing the first aliquot through the flowcell comprises the first pair of valves providing fluid communication between the sheath fluid source and the flowcell; drawing the second aliquot from the one or more mixing chambers comprises drawing the second aliquot from the one or more mixing chambers using the second pair of valves; and the set of acts comprises, after drawing the second aliquot from the one or more mixing chambers, flowing the second aliquot through the flowcell by performing acts comprising the second pair of valves providing fluid communication between the sheath fluid source and the flowcell.- 8 - 0133788.0817287 4901-6446-8614v10

[0033] In some examples, the plurality of classes includes a first class which has a plurality of subclasses, wherein the plurality of subclasses includes the class for circulating tumor cells; and the classification algorithm is configured to classify an input cell image as a circulating tumor cell image by performing a set of classification acts. In some such examples, the set of classification acts may comprise classifying the input cell image into the first class; and classifying the input cell image into the class for circulating tumor cells after classifying the input cell image into the first class.

[0034] In some examples, the classification algorithm is configured to classify an input cell image as a circulating tumor cell image based on a size of a cell depicted in the input cell image.BRIEF DESCRIPTION OF THE DRAWINGS

[0035] FIG. 1 depicts an exemplary system for imaging particles;

[0036] FIG. 2 depicts an exemplary sample fluid injection system;

[0037] FIGS. 3A-3B illustrate an exemplary sample analysis method;

[0038] FIGS. 4A-4C illustrate perspective views of an exemplary staining module; and

[0039] FIG. 5 illustrates a method which may be used to identify rare cell types.

[0040] FIG. 6 illustrates an exemplary architecture for a machine learning model that could be used in identifying rare cell types.

[0041] FIG. 7 illustrates exemplary details which could be included in an architecture such as illustrated in FIG. 6.

[0042] FIG. 8 illustrates an exemplary fluidics system.DETAILED DESCRIPTION- 9 -0133788.0817287 4901-6446-8614v10

[0043] The present disclosure relates to apparatus, systems and methods for analyzing biological samples. The apparatus, systems, and methods in some examples are optimized for imaging and analysis of blood samples, and for identifying rare cell types such as circulating tumor cells (CTC’s) which can indicate patient diseases such as cancers.

[0044] Many rare cells, such as CTC’s are present in low numbers within a patient blood sample. In practice, this means in order to successfully identify such rare cells, a large volume of patient blood may need to be analyzed to detect a relatively small amount of the rare cells. This can create a design challenge in designing a system that can analyze a larger amount of patient blood than is generally required for typical blood analysis (e.g., a complete blood count). Further, such rare cells need to be identified with some high level of specificity as misclassifying cells can affect a correct patient diagnosis (e.g., misclassifying a CTC as a normal cell, or classifying a normal cell as a CTC).

[0045] A. Exemplary System for Imaging Particles

[0046] FIG. 1 shows aspects of a system for imaging particles in a blood fluid sample, which can be used in a system to detect rare cells such as CTC’s. As shown, the system includes a sample fluid injection system 112, a flowcell 114, an image capture device 116, and a processor 118. The flowcell 114 provides a flowpath 120 that transmits a flow of the sheath fluid, optionally in combination with the sample fluid. The sample fluid injection system 112 can be in fluid communication with the flowpath 120 (e.g., via sample fluid entrance 124), and can operate to inject sample fluid 126 through a distal exit port 130 of the injection tube 122 and into a flowing sheath fluid 128 within the flowcell 114 so as to provide a sample fluid stream 132, also referred to as a ribbon. For example, the processor 118 may include or be in operative association with a storage medium having a computer application that, when executed by the processor, is configured to cause the sample fluid injection system 112 to inject sample fluid 126 into the flowing sheath fluid 128.

[0047] As shown, sheath fluid 128 may be introduced into the flowcell 114 by a sheath fluid injection system 134 (e.g., via sheath fluid entrance 136). For example, the processor 118 may include - 10 -0133788.0817287 4901-6446-8614v10or be in operative association with a storage medium having a computer application that, when executed by the processor, is configured to cause the sheath fluid injection system 134 to inject sheath fluid 128 into the flowcell 114. As depicted in FIG. 1, the distal exit port 130 of injection tube 122 can be positioned at a central location along the length of the narrowing transition zone 138. In some cases, the distal exit port 130 can be positioned more closely to the beginning (proximal portion) of the transition zone 138. In some cases, the distal exit port 130 can be positioned more closely to the end (distal portion) of the transition zone 138. In some cases, the distal exit port 130 can be positioned entirely outside of the transition zone 138 (where distal exit port 130 is disposed proximal to the narrowing transition zone).

[0048] With continued reference to FIG. 1, the sample fluid stream 132 has a first thickness T1 at the exit port 130. The flowpath 120 of the flowcell having a decrease in flowpath size such that the thickness of the sample fluid stream 132 decreases from the initial thickness T1 to a second thickness T2 adjacent an image capture site 140. The sample fluid 126 exits the exit port 130 where it merges and / or is surrounded with sheath fluid 128 at a pre-imaging region. In the pre- imaging region, the sample fluid stream 132 becomes sandwiched by the sheath fluid 128 prior to proceeding downstream to the image capture site 140. The image capture device 116 is aligned with the image capture site 140 so as to image a plurality of the particles from the first sample fluid at the image capture site 140 of the flowcell 114.

[0049] Accordingly, the disclosed technology may be used to implement a system 110 for imaging a plurality of particles in a sample fluid 126 having a sample fluid viscosity. The system 110 can include a flowcell 114 having a flowpath 120 and injection tube 122. The flowpath 120 can have a reduction in flowpath size or narrowing transition zone. Further, the system 110 can include a sheath fluid entrance 136 in fluid communication with the flowpath 120 of the flowcell 114 so as to transmit a flow of the sheath fluid along the flowpath 120 of the flowcell 114. The system 110 can also include a sample fluid entrance 124 in fluid communication with the injection tube 122 of the flowcell 114 so as to inject a sample fluid stream 132 of the blood fluid sample into the flowing sheath fluid within the flowcell 114. For example, the sample- 11 -0133788.0817287 4901-6446-8614v10fluid 126 can exit the distal exit port 130 of the injection tube 122 and into an envelope of the flowing sheath fluid 128 to form a sample ribbon 132 therein.

[0050] The sheath fluid 128 along with the sample fluid ribbon 132, formed from the sample fluid 126, flows through the transition zone 138 in flowpath size and toward an imaging site 140. As shown, the system 110 also includes an imaging device 116 that images the plurality of particles at the imaging site 140.

[0051] B. Exemplary Sample Fluid Injection System

[0052] FIG. 2 illustrates an exemplary implementation of a sample fluid injection system such as could be used for a sample fluid injection system 112 in the exemplary system for imaging particles of FIG. 1. In some examples, the fluidics system of FIG. 2 is optimized for a rare cell detection type system whereby a relatively large amount of patient blood sample needs to be imaged to identify rare cells. This optimization can be achieved, for instance, by having a lengthy and larger diameter tubing section to accommodate a large amount of patient blood prior to delivery to the flowcell, and / or allowing for multiple aliquots (e.g., from the same patient sample) through the use of one or more fluidics lines, and / or the inclusion of fluid valves to allow fluid routing between various sections of a fluid and flowcell delivery system.

[0053] As shown in FIG. 2, a sample fluid injection system may have three incoming lines through which fluids can be introduced into the system. First, there may be a sample line 201, which may be connected to a mixing chamber (not shown) in which a sample could be prepared for imaging, such as by mixing with staining and / or lysing agent(s) (e.g., as described in section D, infra, or in U. S. published patent application 2024 / 0027310, filed July 21, 2023 for “Biological Sample Staining Module and Biological Analysis Systems and Methods,” the disclosure of which is hereby incorporated by reference in its entirety). Second, a sample fluid injection system may include a sheath line 202 connected to a source of sheath fluid (e.g., a tank or reservoir of sheath fluid) which may be driven by a fluidics system such as that discussed in section C, infra. The sheath fluid on the sheath line 202 may be used to drive sample fluid from the sample line 201 into the flowcell so that the sample fluid could be - 12 -0133788.0817287 4901-6446-8614v10surrounded by flowing sheath fluid 128 (which may be drawn from its own source of sheath fluid, or may be drawn from a common source as described in U. S. published patent application 2024 / 0342715. filed June 21. 2024 for “Biological Sample Driving System and Method,” the disclosure of which is hereby incorporated by reference in its entirety) and stabilized for imaging as the sample fluid flows through the flowcell. Finally, a sample fluid injection system may include a diluent line 203. As described below, this line may be used to introduce diluent into the flowcell for cleaning between samples.

[0054] In addition to the sample, sheath and diluent lines, the sample fluid injection system of FIG. 2 includes a number of valves for controlling the entry of various fluids into the system as well as their ultimate injection into the flowcell. An exemplary sample analysis method which may be performed using those valves for fluid control is illustrated in FIGS. 3 A and 3B. In the illustrated method, an aliquot of sample fluid may be drawn 301 into a first pending line 204 of the sample fluid injection system. This may be done by actuating a first sample valve 205 and a first fluid valve 206 to establish fluid communication between a first sample line branch 207 and first diluent line branch 208, and using a pump connected to the diluent line 203 to pull sample fluid through the first sample line branch 207 and into the first pending line 204.

[0055] Once the aliquot is present in the first pending line 204, processing may proceed differently depending on whether the flowcell would need to be flushed before the aliquot could be imaged. For example, if the flowcell had been cleaned such that it may still contain a residual amount of diluent, then a portion of the aliquot in the first pending line 204 may be used to flush 302 the flowcell. This may include actuating the first sample valve 205 and the first fluid valve 206 to break fluid communication between the first sample line branch 207 and the first pending line 204, and to establish fluid communication between the first diluent line branch 208 and a first flowcell injection line 209. With fluid communication established, diluent from the diluent line 203 could be used to push a portion of the aliquot from the first pending line 204 though the first flowcell injection line 209 and the flowcell itself, thereby flushing out any diluent or other material which may be present from previous use and / or cleaning of the flowcell.- 13 -0133788.0817287 4901-6446-8614v10

[0056] When the sample fluid injection system is ready to provide sample fluid for imaging (e.g., after the flowcell had been flushed with sample fluid, if it was determined that such flushing was necessary) the aliquot from the first pending line can be flowed 303 through the flowcell. This may include actuating the first sample valve 205 and the first fluid valve 206 to establish fluid communication between the first flowcell injection line 209 and a first sheath line branch 210 while breaking fluid communication between the first pending line 204 and the first diluent line branch 208. With fluid communication established between the first sheath line branch 210 and the first flowcell injection line 209, sheath fluid could be pushed through the first pending line 204, thereby driving the sample fluid and flowing the sample fluid into the flowcell. Once in the flowcell, the sample fluid could be surrounded by an envelope of sheath fluid as described previously, and a plurality of cell images could be captured 304 as the sample flowed through a viewing area of the flowcell. The cell images could then be analyzed 305 (e.g., to identify circulating tumor cells or other rare cell types).

[0057] In the method illustrated in FIGS. 3 A and 3B, while the aliquot from the first pending line 204 is being flowed 303 through the flowcell, the components of the sample fluid injection system not involved in flowing 303 the aliquot could be used to prepare so that another aliquot could be processed from a second pending line 211 without unnecessary delay once the aliquot from the first pending line 204 is finished. This may include determining 306 if the system (as well as upstream components, such as the mixing chamber) needs cleaning. For example, if the aliquot which would be processed on the second pending line 211 is from a different patient sample than the aliquot being flowed 303 through the flowcell (meaning the blood sample is taken from different patient sample tubes), then the sample line 201 and first sample line branch 207 would be cleaned out to avoid carryover between different patient samples (though in some configurations, different aliquots taken from the same patient sample can utilize a cleaning step between imaging of the separate aliquots). Similarly, if the new aliquot would be prepared in the same mixing chamber as the previous aliquot (e.g., if there was only a single mixing chamber in the analyzer), then the mixing chamber may also be cleaned at this time. To perform this cleaning 307, a second sample valve 212 and a second fluid valve 213 could be actuated- 14 -0133788.0817287 4901-6446-8614v10so as to establish fluid communication between a second sample line branch 214 and a second diluent line branch 215. Diluent could then be flowed from the diluent line 203, washing out the tubing of the sample fluid injection system which wasn’t being used in flowing 303 the aliquot from the first pending line 204 through the flowcell, and potentially also washing out the mixing chamber which would be used to prepare the second aliquot. In this way, the disclosed technology could support scenarios where multiple aliquots (which may be from the same sample, or may be from different samples) could be drawn from a single mixing chamber, where the same sample could potentially be prepared in and then drawn out from multiple mixing chambers, and where there are multiple mixing chambers which would each be used to prepare fluid from a different sample.

[0058] In the method of FIG. 3B, after the cleaning 307 is completed, or after a determination 306 that cleaning was not necessary (e.g., if the aliquot which would be processed from the second pending line 211 was from the same patient sample as the aliquot which was then being flowed 303 through the flowcell), a patient sample aliquot could be drawn 308 into the second pending line 211. This may be performed in a manner similar to that described above for drawing 301 an aliquot into the first pending line 301. Specifically, if fluid communication between the second sample line branch 214 and the second diluent line branch 215 was not already present (e.g., due to cleaning 307), the second sample valve 212 and the second fluid valve 213 could be actuated to establish it. A pump connected to the diluent line 203 could then be used to pull the sample fluid of the new aliquot through the second sample line branch and into the second pending line 211.

[0059] Just as similar processes could be used to draw aliquots into the first and second pending lines, processing of sample fluid in the second line 211 could proceed in a manner similar to that described above for processing sample fluid in the first pending line 204. That is, if the flowcell needed to be flushed (e.g., if it had just been used for imaging an aliquot from a different patient’s sample) the second sample valve 212 could be actuated to break communication between the second pending line 211 and the second sample line branch 214 and establish it between the second pending line 211 and establish it between the second pending line 211 and - 15 -0133788.0817287 4901-6446-8614v10a second flowcell injection line 216, and a portion of the sample fluid from the second pending line 211 could be used to flush 309 out the flowcell. Fluid communication with the second pending line 211 could then be switched from the second diluent line branch 215 to the second sheath line branch 217, and the sample fluid from the second pending line 211 could be flowed 310 through the flowcell for imaging using sheath fluid from the second sheath line branch 217. While the sample fluid from the second pending line 211 was flowing through the flowcell, preparation steps for imaging another aliquot on the first pending line 204 (e.g., cleaning if necessary, drawing an aliquot into the first pending line 204) could be taking place. This process of alternating utilization of first and second pending lines could then proceed iteratively while there was still more sample fluid to be analyzed.

[0060] Other actions may also be performed when processing patient sample aliquots using a sample fluid injection system such as that illustrated in FIG. 2. For example, in a case where a second aliquot is drawn 308 into a second pending line 211 while a first aliquot from a different patient’s sample is being flowed 303 through the flowcell, after imaging is completed for the first aliquot but before the second aliquot is moved out of the second pending line, the flowcell may be subjected to a cleaning step. This may include switching fluid communication with the first pending line 204 to the first diluent line branch 208 from the first sheath line branch 217, and then flowing diluent through the flowcell via the first pending line 204. Other modifications, including removing or changing implementation of one or more actions described the context of FIGS. 3A and 3B, are also possible, and could be implemented without undue experimentation by one of skill in the art in light of this disclosure. However, regardless of particular implementation details, alternating utilization such as described above may increase efficiency by removing downtime between when processing of one aliquot has ended and processing of another aliquot can begin. Additionally, since the amount of sample fluid which would need to be processed to identify circulating tumor cells or other rare cell types may be more than can be addressed in a single aliquot (e.g., because it is more than the amount of fluid which can be held in a pending line), streamlined multi-aliquot processing such as could be performed using a method such as described in the context of FIGS. 3 A and 3B can- 16 -0133788.0817287 4901-6446-8614v10be particularly useful when the analysis to be performed includes identification of rare cell types.

[0061] As referenced earlier and herein, rare cell detection necessitates a relatively larger amount of a patient blood sample to be analyzed compared to a conventional CBC test since these rare cells (e.g., CTC’s) may have a very low concentration in blood. By way of example, the fluidics concepts discussed herein can allow about 20% of an initially aliquoted patient sample to be imaged (e.g., if 100 μL of blood is aspirated, after preparation and other steps, about 20% of this would end up being imaged) - in some examples this number is about 20-40%, or about 20-30%. Certain portions of the prepared sample can be used to flush the flowcell, and - for instance to avoid boundary edge detection issues - only a portion of the imaged flow-stream may be analyzed (e.g., the more medial section). In comparison, a CBC may require a much smaller number such as 1% of the blood since the test is looking for common cells that are equivalently expressed throughout a blood column.

[0062] C. Sheath Fluid Fluidics

[0063] FIG. 8 shows an exemplary fluidics system 10 which could be used to convey sheath fluid from a common source such as a sheath fluid reservoir 12 to a sheath fluid line 202 of a sample fluid injection system such as shown in FIG. 2 as well as to a sheath fluid entrance 136 such as shown in FIG. 1 so that it could form a flow enveloping the sample stream 132 from the sample fluid injection system 112. In the exemplary fluidics system, the sheath fluid reservoir 12 may be configured to provide sheath fluid at a bulk flow rate to a first flow path 16 and a second flow path 18. This may be done by connecting the sheath reservoir to a source of sheath fluid 34 via a pump 17. as well as to an air compressor 26 and air vent 30 via corresponding valves 28, 32. In this configuration, the pump 17 could supply additional sheath fluid as it was used (e.g., during imaging, as described above) with the valves 28, 32 to the air compressor 26 and vent 30 being selectively opened and closed to maintain appropriate pressure for driving the sheath fluid through the fluidics system.- 17 -0133788.0817287 4901-6446-8614v10

[0064] After exiting the reservoir 12 the sheath fluid may pass through a T junction 15, followed by first and second restrictors 38, 44 and first and second flow rate sensors 40, 46 on, respectively, the first and second flow paths 16, 18. These restrictors 38, 44 and flow rate sensors 40, 46 may be connected to a controller 64, which could modify the restrictors (e.g., by expanding or contracting a programmable inner diameter) based on data from the flow rate sensors to dynamically maintain a particular flow rate on the first flow path 16 and the second flow path 18. In some cases, this may allow the flow to be established at a controlled ratio between the first and second flow paths in which the flow rate through the second flow path 18 is higher than the flow rate through the first flow path 16. With the appropriate ratio established, the sheath fluid from the first flow path 16 could enter a sample fluid injection system 53 (e.g., through the sample line 202 of the sample fluid injection system of FIG. 2). From there, it could entrain or push an aliquot of a biological sample 22 to be injected as a sample stream 54 into a flowcell 14 (e.g., injected into flowcell 114 of FIG. 1 via sample fluid entrance 124). Meanwhile, the higher velocity sheath fluid from the second flow path 18 could be injected into a second passageway 56 (e.g., sheath fluid entrance 136 in the flowcell 114 of FIG. 1) to envelope the sample stream and stabilize it for imaging by an imaging device 24, before the fluid exited the flowcell and was discharged to waste 62.

[0065] While a fluidics system such as discussed above in the context of FIG. 8 may be used to convey sheath fluid through a sample fluid injection system and flowcell, it should be understood that that fluidics system is intended to be illustrative only, and that other approaches may be used in different implementations of the disclosed technology. For example, in some cases, rather than having a common source of sheath fluid such as the reservoir 12 of FIG. 8, first and second flow lines may be fed independently from their own sheath fluid sources. Similarly, in some cases, rather than having flow restrictors 38, 44 which could be actuated by a controller, in some cases static flow restrictors (e.g., portions of a flow line with decreased inner diameters) may be used to maintain a constant ratio between flows on the first and second flow lines. Additionally, in some cases additional components (e.g., valves capable of shutting off the flow of sheath fluid on the first and / or second flow lines) may also be included in a fluidics- 18 -0133788.0817287 4901-6446-8614v10system used when implementing the disclosed technology. Accordingly, the discussion of the exemplary fluidics system of FIG. 8 should be understood as being illustrative only, and should not be treated as limiting on the scope of protection provided by this application or by any other application claiming the benefit of this disclosure.

[0066] D. Exemplary Staining Module

[0067] Turning next to FIGS. 4A-4C, those figures depict a staining module (also referred to as a staining device) 400, such as could be used as a mixing chamber from which sample fluid may be dispensed into a sample fluid injection system such as discussed above in the context of FIGS. 2 and 3A-3B. The staining module can be used to stain cells such as rare cells (e.g., in the case where such rare cells such as CTC’s contain a nucleus). In some examples, the staining module facilitates a lysing step (to remove red blood cells), and a staining step (to stain the remaining material - such as white blood cells, platelets, and / or rare cells such as CTC’s). In some examples, the staining module facilitates staining of a nucleus region of cells (e.g., CTC’s, white blood cells) where an imaging algorithm can then analyze the stained image to determine a particular cell type of the cell which is imaged. In some examples, this application of a staining model may involve mixing about 50 μL of sample aspirated from a sample container with about 200 μL of staining and / or lysing agent(s), with essentially all of the 250 μL of prepared sample fluid being pulled out for imaging or routed through the flowcell in a process such as described above in the context of FIGS. 3A-3B.

[0068] In a staining module such as illustrated in FIGS. 6A-6C, a sample may be both mixed with a staining agent and incubated via heating prior to cells from the sample being imaged by a camera such as image capture device 116. The staining agent may include any suitable composition. For example, the staining agent may be composed in accordance with any one or more teachings of U. S. Pat. No. 9,279,750, entitled “Method and Composition for Staining and Sample Processing,” issued on March 8, 2016, the disclosure of which is hereby incorporated by reference in its entirety; and / or U. S. Pat. No. 9,322,753, entitled “Method and Composition for Staining and Processing a Urine Sample,” issued on April 26, 2016, the disclosure of which - 19 -0133788.0817287 4901-6446-8614v10is hereby incorporated by reference in its entirety; and / or US Pub. No. 2021 / 0108994, entitled “Method and Composition for Staining and Sample Processing,” published on April. 15, 2021, the disclosure of which is hereby incorporated by reference in its entirety.

[0069] In the embodiment shown, the staining module 400 includes a housing 410, a pair of ferromagnetic sheets 412, and a heater in the form of a heating coil 414 (FIG. 4C). In various embodiments, heating coil 414 can comprise a resistive coil, or alternatively an inductive coil. As best shown in FIG. 4A, the housing 410 includes a plurality of (e.g.. four) sidewalls 420 which collectively define an interior chamber 422 (also referred to as a sample reservoir or a mixing chamber) for receiving the staining agent and the sample, mixing the staining agent and the sample to form a sample mixture, and incubating the sample mixture. The housing 410 also includes a top wall 424 and a port 426 extending through the top wall 424 to the interior chamber 422. The port 426 may permit a stain dispenser (not shown) to deliver the staining agent to the interior chamber 422, and / or may permit a sample dispenser (not shown) to deliver the sample to the interior chamber 422 so as to be added to the staining agent.

[0070] In some embodiments, the housing 410 may comprise a metallic material having relatively high thermal conductivity, such as aluminum, in order to promote uniform heating of the housing 410 and likewise uniform heating of the contents of the interior chamber 422. In the embodiment shown, the sidewalls 420 of the housing 410 are laminated with respective ferromagnetic sheets 412 to improve the efficiency of the heating (e.g., resistive heating, or alternatively inductive heating) performed by staining module 400 (e.g., due to the relatively low ferromagnetic properties of aluminum). More particularly, each ferromagnetic sheet 412 is secured to the outer surfaces of a corresponding pair of sidewalls 420. It will be appreciated that any suitable number of ferromagnetic sheets 412 may be used to laminate the sidewalls 420. In the embodiment shown, a thermally conductive compound 430 is deposited on the outer surfaces of the sidewalls 420 for adhering the ferromagnetic sheets 412 to the sidewalls 420. As best shown in FIG. 4B, an adhesive tape 432 is tightly wrapped about the ferromagnetic sheets 412 to securely engage the inner surfaces of the ferromagnetic sheets 412 with the outer surfaces of the sidewalls 420.- 20 -0133788.0817287 4901-6446-8614v10

[0071] As best shown in FIG. 4C, the heating coil 414 includes a wire 440 wound about the sidewalls 420 of the housing 410 (and about ferromagnetic sheets 412). The wire 440 may comprise a metallic material having relatively high electrical conductivity, such as copper. The wire 440 may have any suitable cross-sectional area and / or thickness, and may be wound to define any suitable number of turns for the heating coil 414. The heating coil 414 in one embodiment functions as an inductor or induction coil, and is operatively coupled to a power unit 450, which may be configured to drive the heating coil 414 to a frequency at which the heating coil 414 behaves as a resonant circuit that under excitation produces an alternating current thereby producing an alternating magnetic field at or near the heating coil 414. This field may generate an electromagnetic field (EMF) on the outer surfaces of the sidewalls 420, which may in turn cause an alternating current. This current, in conjunction with the resistivity of the housing 410, may yield power dissipation and heat up the outer surfaces of the sidewalls 420. Such heat may be transferred to the contents of the chamber 422, such as the staining agent and / or the sample. It will be appreciated that such induction heating may be performed using relatively low input power, and / or may achieve homogeneous heating of the contents of the chamber 422 and thereby improve staining and / or lysing performance. In this regard, exciting the circuit at the resonant frequency may deliver maximum power, and exciting the circuit at an increasing frequency may effectively adjust the power delivery. Alternative embodiments can utilize a resistive heater / resistance heating coil for heater coil 414.

[0072] In some embodiments, a temperature sensor such as a thermistor (not shown) may be configured to continuously sense the temperature of the contents of the chamber 422. The temperature sensor may be configured to send feedback signals indicative of the sensed temperatures to a controller (not shown) which may in turn be configured to send control signals to the power unit 450 for selectively driving the heating coil 414. In this manner, the controller may cease heating of the contents of the chamber 422 upon reaching a threshold temperature. In one example, the controller utilizes heating control algorithms and the feedback signals are incorporated into elements of the algorithms or computer-driven instructions provided to the power unit 450 and / or heating coil 414 to optimally regulate- 21 -0133788.0817287 4901-6446-8614v10temperature. In some embodiments, the controller may be configured to send control signals to a maintenance heater (not shown) for maintaining the contents of the chamber 422 at the threshold temperature.

[0073] In one embodiment, a plurality of staining modules are contemplated, each utilizing the structure of Figures 4A-4C (i.e., a plurality of structural elements 400). In this way, a plurality of samples can be stained, incubated, or otherwise prepared at a similar time. In one example, each staining module has its own unique heating element. In one example a staining module has a plurality of chambers 422, each capable of receiving a sample, and a common heating structure connected to the entire module (e.g., a single housing 410 with a plurality of chambers 422 and a common heating coil 414 surrounding housing 410).

[0074] E. Exemplary Rare Cell Type Identification

[0075] Turning next to FIG. 5, that figure illustrates a method which may be used to identify rare cell types. As shown in that figure, this identification may begin by applying 501 a first classifier to a cell image from a sample being analyzed for the presence of circulating tumor cells or other type of rare cell (e.g., an image of a cell from the sample after it has been stained and lysed in a staining module such as described in the preceding section). This first classifier may be an exclusion criteria which could separate cell images into a first class for images which did not depict the relevant rare cell type, or a second class for images which may or may not depict the rare cell type. For example, in the case where a sample was being analyzed to identify circulating tumor cells, applying 501 the first classifier may comprise comparing the size of the cell depicted in the cell image with a size threshold (e.g., 15 micrometers). Cells which were below the size threshold could then be treated as not being circulating tumor cells, while cells above that threshold may be circulating tumor cells, but may also be other types of imaged objects (e.g., fatty cells, large air bubbles).

[0076] In the method of FIG. 5, after the first classifier had been applied 501, one or more additional classification acts may be performed to classify the cell image into a specific class for the rare cell type which was subject of the analysis. This may include applying additional image - 22 -0133788.0817287 4901-6446-8614v10processing steps which could determine the image depicts an object with characteristics consistent with it being the rare cell type. For example, in a case where applying the first classifier comprises applying an object size cutoff, the one or more additional classification acts may include analyzing the cell image to determine if the object it depicts has a nucleus, as the presence or absence of a nucleus could help distinguish objects (e.g., air bubbles) which would satisfy the first classifier other than the rare cell type in question. Similarly, depending on the rare cell type and the potential confounding objects in question, various image characteristics may be evaluated to determine if the object depicted in a cell image may be classified as the rare cell type (e.g., a circulating tumor cell may be distinguishable in terms of color and / or shape from various other types of similarly sized objects). In some examples, circulating tumor cells are generally large and have relatively large nucleuses, so characteristics pertaining to cell features such as cell size and nucleus size, among others, can be useful in identifying circulating tumor cells (or differentiating circulating tumor cells from other cell types). Exemplary characteristics which may be used in this type of evaluation are listed below in table 1.Characteristic Name Characteristic DescriptionArea Foreground region area (e.g., in pixels)Perimeter Foreground region perimeter, which may be measured in pixels, or may provide use various techniques (e.g., interpolation or curve fitting) to provide a length with subpixel accuracy.Circularity A circularity metric for the foreground area (e.g., ratio of the radius of a circle inscribed in the foreground area to the radius of a circle circumscribed around the foreground area) Major Axis Dimension Major axis length of the foreground region (e.g., in pixels) Minor Axis Dimension Minor axis length of the foreground rection (e.g., in pixels) Red Min Minimum, maximum, mean and standard deviations of red Red Max channel values (e.g., the R channel in an RGB format image)- 23 -0133788.0817287 4901-6446-8614v10Red Mean of pixels in the foreground region area, potentially after image Red StDev processing such as illumination correction and / or debayering. Green Min Minimum, maximum, mean and standard deviations of green Green Max channel values (e.g., the G channel in an RGB format image) Green Mean of pixels in the foreground region area, potentially after image Green StDev processing such as illumination correction and / or debayering. Blue Min Minimum, maximum, mean and standard deviations of blue Blue Max channel values (e.g., the B channel in an RGB format image) Blue Mean of pixels in the foreground region area, potentially after image Blue StDev processing such as illumination correction and / or debayering. Value Min The minimum, maximum, mean, and standard deviation of the Value Max pixel values in the value component of an HSV rendition of Value Mean the foreground region.Value StDevEdge Count The count of Canny edge image pixels in the center (e.g., within half a radius) of the foreground regionB oundary_Point_Count The number of pixels on the foreground region boundary. Peak Count Number of peaks (convex ranges) on the boundary of the foreground region.Valley Count Number of valleys (concave ranges) on the boundary of the foreground region.Peakl Valley # Index A location for the specified peak or valley on the boundary of the foreground region (e.g., Peak N Index would be the location of the Nthpeak on the boundary of the foreground region, while Valley N Index would be the location of the Nthvalley on the boundary of the foreground region).Peakl Valley # Angle The included angle (e.g., in radians) of the specified peak or valley on the boundary of the foreground region.- 24 -0133788.0817287 4901-6446-8614v10PeakIValley # Span The number of pixels on either side of the center are included in the specified peak or valley?Peakl Valley # Location The ratio of the partial perimeter (to the center of the specified Ratio peak or valley) to the full foreground region perimeter.PeakIValley # Direction X The appropriate component of the direction vector for the PeakIValley # Direction Y specified peak or valley.PeakIValley # Width The width, in pixels, of the specified peak or valley.PeakIValley # Height or The height of the specified peak or depth of the specified Depth valley.PeakIValley # Fill Ratio Solidity of the specified peak or valley.Angle_std Standard deviation of boundary angles (e.g., an array of boundary angles as illustrated in FIG. 5).1st_horn_pos The index of the first horn in a boundary point array 2nd_horn_pos The index of the second horn in a boundary point array 1st_horn_mag The boundary angle of the first horn.2nd_horn_mag The boundary angle of the second horn.Horn_pos_delta A value equal to abs(lst_horn_pos 2“d_horn_pos) / Boundary_Point_Count, except if abs(1st_horn_pos - 2nd_horn_pos) / Boundary_Point_Count is greater than 0.5, in which case Horn_pos_delta is equal to 1. Horn_mag_delta A value equal to 2nd_horn_mag – 1st_horn_mag.Hl_large_mass Ratio of pixel count in the larger part of foreground region to the pixel count in the entire foreground region when the foreground region is separated by the primary line between the first and second horns.H2_large_mass Ratio of pixel count in the larger part of foreground region to the pixel count in the entire foreground region when the foreground region is separated by a secondary line which is- 25 -0133788.0817287 4901-6446-8614v10perpendicular to and bisects the primary line between the first and second horns.Depth The maximum point distance for any point in the foreground region, where the point distance for a point in the foreground region is the distance between that point and the nearest point on the boundary of the foreground region.Area_bounding_box The area (e.g., in pixels) of a bounding box of the foreground region.Table 1: Illustrative image characteristics

[0077] Another approach which may be used in determining whether a cell image should be classified as depicting a particular rare cell type is to use a machine learning model which has been trained to identify the particular rare cell type in question. An example architecture which may be used for this purpose is illustrated in FIG. 6. In that architecture, an input image 601 would be analyzed in a series of stages 602a-602n, each of which may be referred to as a “layer,” and which are illustrated in more detail in FIG. 7. As shown in FIG. 7, an input 701 (which, in the initial layer 702a of FIG. 7 would be the input image 601, and otherwise would be the output of the preceding layer) is provided to a layer 702 where it would be processed to generate one or more transformed images 703a-703n. This processing may include convolving the input 701 with a set of filters 704a-704n, each of which would identify a type of feature from the underlying image that would then be captured in that filter’s corresponding transformed image. For instance, as a simple example, convolving an image with the filter shown in table 2 could generate a transformed image capturing the edges from the input image 701.[ -1 -1 -1 ][ -1 8 -1 ][ -1 -1 -1 ]Table 2

[0078] As shown in FIG. 7, in addition to generating transformed images 703a-703n a layer may also generate a pooled image 705a-705n for each of the transformed images 703a-703n. This may be done, for example, by organizing the appropriate transformed image into a set of regions,- 26 -0133788.0817287 4901-6446-8614v10and then replacing the values in that region with a single value, such as the maximum value for the region or the average of the values for the region. The result would be a pooled image whose resolution would be reduced relative to its corresponding transformed image based on the size of the regions it was split into (e.g., if the transformed image had NxN dimensions, and it was split into 2x2 regions, then the pooled image would have size (N / 2)x(N / 2)). These pooled images 705a-705n could then be combined into a single output image 706, in which each of the pooled images 705a-705n is treated as a separate channel in the output image 706. This output image 706 can then be provided as input to the next layer as shown in FIG. 6.

[0079] Returning to the discussion of FIG. 6, after a final output image 603 has been created through the various stages 602a-602n of processing, the final output image 603 could be provided as input to a neural network 604. This may be done, for example, by providing the value of each channel of each pixel in the output image 603 to an input node of a densely connected single layer network. The output of the neural network 604 could then be treated as a classification of the original input image 601. For example, in the case such as shown in FIG. 6, where a neural network 604 has a single output node, that output node may be treated as a probability, with values greater than 0.5 indicating that the input image should be classified as the rare cell type, and values less than 0.5 indicating that the input image should not be classified as the rare cell type. Alternatively, in a case where there are multiple output nodes, each of the output nodes may be assigned to a particular classification (e.g., circulating tumor cell, fatty cell, air bubble, etc.), and the classification for the output node with the highest value could be treated as the classification for the input image.

[0080] A machine learning model such as a model following the architecture discussed in the context of FIGS. 6 and 7 can be trained to make classification determinations using the types of images it would be expected to see in production (e.g., images of various types of cells captured using a flowcell based analyzer) annotated with labels indicating the correct type of particle that those training images should be identified as depicting (e.g., labels as added by a human annotator). This training may include comparing labels applied by the model being trained with the ground truth labels provided by the annotation, and adjusting the values of the machine - 27 -0133788.0817287 4901-6446-8614v10learning model’s parameters to minimize a loss function (e.g., cross entropy loss) for that comparison. The training can also include splitting the annotated images into multiple subsets, or folds, and then training and evaluating the model multiple times, with a different fold of training images being held back as a validation set each time (i.e., K-fold cross validation). In this way, performance metrics from each training instance can be averaged to verify the model’s generalization performance and, assuming the performance is acceptable, a final trained version of the model (e.g., whichever trained model had the best individual performance) can be used to make inferences (i.e., classify cell images) in production.

[0081] Combinations of the foregoing approaches may also be used in some cases. For example, it is possible that cell images may initially be analyzed using a size qualification and one or more image characteristics, and then images which couldn’t be classified as other than a rare cell type could be analyzed using a machine learning model for a final classification. Alternatively, it is also possible that all images may initially be analyzed using a machine learning model, with other evaluations (e.g., size thresholding) being applied subsequently, if at all. It is also possible that, in some cases, the classification process may assign multiple classifications to an image, such as by first classifying an image into a “large object” type, and subsequently making a more fine grained classification into a “circulating tumor cell” type. Other alternatives are also possible, with the particular approach used in practice being determined based on relevant requirements (e.g., required turnaround times) and resources (e.g., available compute power for evaluating machine learning models) available in a particular context. Accordingly, the examples provided above for potentially classifying a cell image as depicting a rare cell type should be understood as being illustrative only, and should not be treated as limiting.

[0082] All patents, patent publications, patent applications, journal articles, books, technical references, and the like discussed in the instant disclosure are incorporated herein by reference in their entirety for all purposes.- 28 -0133788.0817287 4901-6446-8614v10

[0083] Different arrangements of the components depicted in the 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 may be employed without reference to other features and subcombinations. Embodiments of the invention have been described for illustrative and not restrictive purposes, and alternative embodiments will become apparent to readers of this patent. In certain cases, method steps or operations may be performed or executed in differing order, or operations may be added, deleted or modified. It can be appreciated that, in certain aspects of the 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 function or functions. Except where such substitution would not be operative to practice certain embodiments of the invention, such substitution is considered within the scope of the invention. Accordingly, the claims should not be limited to the embodiments described above or depicted in the drawings, and instead should be understood as having the scope provided when their terms which are explicitly defined below are given their explicit definitions and the remaining terms are given their broadest reasonable interpretations as provided by a general purpose dictionary.

[0084] Explicit Definitions

[0085] In the claims, an “aliquot” should be understood as a portion of a larger whole, and would include a portion which is subjected to one or more processing steps after being separated from the larger whole. For example, an “aliquot” of a sample could refer to a portion of a sample which is extracted from a sample tube and placed in a mixing chamber, and could also be used to refer to fluid in the mixing chamber after it has been subjected to one or more processing steps (e.g., staining and / or lysing).

[0086] In the 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 that it is indicated as being based on. To indicate that something must be completely determined based on something else, it is- 29 -0133788.0817287 4901-6446-8614v10described as being “based EXCLUSIVELY on” whatever it must be completely determined by.

[0087] In the claims, the term “set” should be understood as one or more things which are grouped together.- 30 - 0133788.0817287 4901-6446-8614v10

Claims

What is claimed is:

1. A blood imaging system for identification of circulating tumor cells, comprising:a) one or more mixing chambers;b) a flowcell;c) an image capture device; andd) one or more processors configured to perform a set of acts comprising:i) drawing a first aliquot from the one or more mixing chambers;ii) after drawing the first aliquot from the one or more mixing chambers, flowing the first aliquot through the flowcell;iii) utilizing the image capture device to capture an image of a cell as the first aliquot flows through the flowcell;iv) drawing a second aliquot from the one or more mixing chambers as the first aliquot flows through the flowcell: andv) analyzing the image of the cell using a classification algorithm, wherein the classification algorithm is configured to classify the cell according to a plurality of classes, wherein the plurality of classes includes a class for circulating tumor cells.

2. The system of claim 1, wherein:a) drawing the first aliquot from the one or more mixing chambers comprises drawing the first aliquot into a first pending sample line: andb) drawing the second aliquot from the one or more mixing chambers comprises drawing the second aliquot into a second pending sample line.

3. The system of any of claims 1-2, wherein the set of acts further comprises:a) after the first aliquot is no longer flowing through the flowcell, flowing the second aliquot through the flowcell; and- 31 -0133788.0817287 4901-6446-8614v10b) determining whether to clean the flowcell between flowing the first aliquot through the flowcell and flowing the second aliquot through the flowcell based on whether the first aliquot and the second aliquot are from a single blood sample.

4. The system of any of claims 1-3, further comprising:a) a sheath fluid source;b) a diluent source; andc) one or more pairs of valves, wherein each pair of valves from the one or more pairs of valves is separated by a corresponding pending sample line and each pair of valves from the one or more pairs of valves is operable to selectively provide fluid communication between:i) the sheath fluid source and the flowcell; andii) the diluent source and the flowcell.

5. The system of claim 4, wherein, for each pair of valves from the one or more pairs of valves, that pair of valves providing fluid communication between the sheath fluid source and the flowcell comprises:a) at the first valve from that pair of valves, providing fluid communication between the flowcell and the pending sample line corresponding to that pair of valves; and b) at the second valve from that pair of valves, providing fluid communication between the pending sample line corresponding to that pair of valves and the sheath fluid source.

6. The system of any of claims 4-5, wherein, for each pair of valves from the one or more pairs of valves, that pair of valves providing fluid communication between the sheath fluid source and the flowcell comprises:a) at the first valve from that pair of valves, blocking fluid communication between the pending sample line corresponding to that pair of valves and the one or more mixing chambers; and- 32 -0133788.0817287 4901-6446-8614v10b) at the second valve from that pair of valves, blocking fluid communication between the pending sample line corresponding to that pair of valves and the diluent source.

7. The system of any of claims 4-6, wherein, for each pair of valves from the one or more pairs of valves, that pair of valves providing fluid communication between the diluent source and the flowcell comprises:a) at the first valve from that pair of valves, providing fluid communication between the flowcell and the pending sample line corresponding to that pair of valves; and b) at the second valve from that pair of valves, providing fluid communication between the pending sample line for that pair of valves and the diluent source.

8. The system of any of claims 4-7, wherein, for each pair of valves from the one or more pairs of valves, that pair of valves providing fluid communication between the diluent source and the flowcell comprises:a) at the first valve from that pair of valves, blocking fluid communication between the pending sample line corresponding to that pair of valves and the one or more mixing chambers; andb) at the second valve from that pair of valves, blocking fluid communication between the pending sample line corresponding to that pair of valves and the sheath fluid source.

9. The system of any of claims 4-8, wherein each pair of valves from the one or more pairs of valves is configured to be used in drawing an aliquot from the one or more mixing chambers by performing acts comprising, at the first valve from that pair of valves, establishing fluid communication between the one or more mixing chambers and the pending sample line corresponding to that pair of valves.

10. The system of any of claims 4-9, wherein each pair of valves from the one or more pairs of valves is configured to be used in cleaning the flowcell by performing acts comprising:- 33 -0133788.0817287 4901-6446-8614v10a) at the first valve from that pair of valves, providing fluid communication between the flowcell and the pending sample line corresponding to that pair of valves; and b) at the second valve from that pair of valves, providing fluid communication between the pending sample line corresponding to that pair of valves and the diluent source.

11. The system of any of claims 4-10, wherein each pair of valves from the one or more pairs of valves is configured to be used in cleaning the one or more mixing chambers by performing acts comprising:a) at the first valve from that pair of valves, providing fluid communication between at least one of the one or more mixing chambers and the pending sample line corresponding to that pair of valves; andb) at the second valve from that pair of valves, providing fluid communication between the pending sample line corresponding to that pair of valves and the diluent source.

12. The system of claim 2 in combination with claims 3, 10 and 11, wherein:a) the first pending sample line is the pending sample line corresponding to a first pair of valves from the one or more pairs of valves;b) the second pending sample line is the pending sample line corresponding to a second pair of valves from the one or more pairs of valves;c) cleaning the flowcell between flowing the first aliquot through the flowcell and flowing the second aliquot through the flowcell comprises, after flowing the first aliquot through the flowcell, using the first pair of valves in cleaning the flowcell; andd) the set of acts comprises, before drawing the second aliquot from the one or more mixing chambers, using the second pair of valves in cleaning the one or more mixing chambers.- 34 -0133788.0817287 4901-6446-8614v1013. The system of claim 2 in combination with claims 5-9 and optionally any of claims 10-12, wherein:a) the first pending sample line is the pending sample line corresponding to a first pair of valves from the one or more pairs of valves;b) the second pending sample line is the pending sample line corresponding to a second pair of valves from the one or more pairs of valves;c) drawing the first aliquot from the one or more mixing chambers comprises drawing the first aliquot from the one or more mixing chambers using the first pair of valves; d) flowing the first aliquot through the flowcell comprises the first pair of valves providing fluid communication between the sheath fluid source and the flowcell; e) drawing the second aliquot from the one or more mixing chambers comprises drawing the second aliquot from the one or more mixing chambers using the second pair of valves; andf) the set of acts comprises, after drawing the second aliquot from the one or more mixing chambers, flowing the second aliquot through the flowcell by performing acts comprising the second pair of valves providing fluid communication between the sheath fluid source and the flowcell.

14. The system of any of claims 1-13, wherein:a) the plurality of classes includes a first class which has a plurality of subclasses, wherein the plurality of subclasses includes the class for circulating tumor cells; b) the classification algorithm is configured to classify an input cell image as a circulating tumor cell image by performing a set of classification acts comprising: i) classifying the input cell image into the first class; andii) classifying the input cell image into the class for circulating tumor cells after classifying the input cell image into the first class.- 35 -0133788.0817287 4901-6446-8614v1015. The system of any of claims 1-14, wherein the classification algorithm is configured to classify an input cell image as a circulating tumor cell image based on a size of a cell depicted in the input cell image.

16. A biological imaging method for identification of circulating tumor cells, comprising: a) drawing a first aliquot from one or more mixing chambers of an analyzer, the first aliquot including a plurality of cells;b) after drawing the first aliquot from the one or more mixing chambers, flowing the first aliquot through a flowcell of the analyzer;c) utilizing an image capture device to capture an image of a cell as the first aliquot flows through the flowcell;d) drawing a second aliquot from the one or more mixing chambers as the first aliquot flows through the flowcell;e) analyzing the image of the cell using a classification algorithm, wherein the classification algorithm is configured to classify the cell according to at least one of a plurality of classes, wherein the plurality of classes includes a class for circulating tumor cells.

17. The method of claim 16 wherein:a) drawing the first aliquot from the one or more mixing chambers comprises drawing the first aliquot into a first pending sample line; andb) drawing the second aliquot from the one or more mixing chambers comprises drawing the second aliquot into a second pending sample line.

18. The method of any of claims 16-17, wherein the method further comprises:a) after the first aliquot is no longer flowing through the flowcell, flowing the second aliquot through the flowcell; and- 36 -0133788.0817287 4901-6446-8614v10b) determining whether to clean the flowcell between flowing the first aliquot through the flowcell and flowing the second aliquot through the flowcell based on whether the first aliquot and the second aliquot are from a single blood sample.

19. The method of any of claims 16-18, wherein the analyzer comprises:a) a sheath fluid source;b) a diluent source; andc) one or more pairs of valves, wherein each pair of valves from the one or more pairs of valves is separated by a corresponding pending sample line, and each pair of valves from the one or more pairs of valves is operable to selectively provide fluid communication between:i) the sheath fluid source and the flowcell; andii) the diluent source and the flowcell.

20. The method of claim 19, wherein, for each pair of valves from the one or more pairs of valves, that pair of valves providing fluid communication between the sheath fluid source and the flowcell comprises:a) at the first valve from that pair of valves, providing fluid communication between the flowcell and the pending sample line corresponding to that pair of valves; and b) at the second valve from that pair of valves, providing fluid communication between the pending sample line corresponding to that pair of valves and the sheath fluid source.

21. The method of any of claims 19-20, wherein, for each pair of valves from the one or more pairs of valves, that pair of valves providing fluid communication between the sheath fluid source and the flowcell comprises:- 37 -0133788.0817287 4901-6446-8614v10a) at the first valve from that pair of valves, blocking fluid communication between the pending sample line corresponding to that pair of valves and the one or more mixing chambers; andb) at the second valve from that pair of valves, blocking fluid communication between the pending sample line corresponding to that pair of valves and the diluent source.

22. The method of any of claims 19-21, wherein, for each pair of valves from the one or more pairs of valves, that pair of valves providing fluid communication between the diluent source and the flowcell comprises:a) at the first valve from that pair of valves, providing fluid communication between the flowcell and the pending sample line corresponding to that pair of valves; and b) at the second valve from that pair of valves, providing fluid communication between the pending sample line for that pair of valves and the diluent source.

23. The method of any of claims 19-22, wherein, for each pair of valves from the one or more pairs of valves, that pair of valves providing fluid communication between the diluent source and the flowcell comprises:a) at the first valve from that pair of valves, blocking fluid communication between the pending sample line corresponding to that pair of valves and the one or more mixing chambers; andb) at the second valve from that pair of valves, blocking fluid communication between the pending sample line corresponding to that pair of valves and the sheath fluid source.

24. The method of any of claims 19-23, wherein each pair of valves from the one or more pairs of valves is configured to be used in drawing an aliquot from the one or more mixing chambers by performing acts comprising, at the first valve from that pair of valves, establishing fluid communication between the one or more mixing chambers and the pending sample line corresponding to that pair of valves.- 38 -0133788.0817287 4901-6446-8614v1025. The method of any of claims 19-24, wherein each pair of valves from the one or more pairs of valves is configured to be used in cleaning the flowcell by performing acts comprising: a) at the first valve from that pair of valves, providing fluid communication between the flowcell and the pending sample line corresponding to that pair of valves; and b) at the second valve from that pair of valves, providing fluid communication between the pending sample line corresponding to that pair of valves and the diluent source.

26. The method of any of claims 19-25, wherein each pair of valves from the one or more pairs of valves is configured to be used in cleaning the one or more mixing chambers by performing acts comprising:a) at the first valve from that pair of valves, providing fluid communication between at least one of the one or more mixing chambers and the pending sample line corresponding to that pair of valves; andb) at the second valve from that pair of valves, providing fluid communication between the pending sample line corresponding to that pair of valves and the diluent source.

27. The method of claim 17 in combination with claims 18, 25 and 26, wherein:a) the first pending sample line is the pending sample line corresponding to a first pair of valves from the one or more pairs of valves;b) the second pending sample line is the pending sample line corresponding to a second pair of valves from the one or more pairs of valves;c) cleaning the flowcell between flowing the first aliquot through the flowcell and flowing the second aliquot through the flowcell comprises, after flowing the first aliquot through the flowcell, using the first pair of valves in cleaning the flowcell; and- 39 -0133788.0817287 4901-6446-8614v10d) the set of acts comprises, before drawing the second aliquot from the one or more mixing chambers, using the second pair of valves in cleaning the one or more mixing chambers.

28. The method of claim 17 in combination with claims 20-24 and optionally any of claims 25-27, wherein:a) the first pending sample line is the pending sample line corresponding to a first pair of valves from the one or more pairs of valves;b) the second pending sample line is the pending sample line corresponding to a second pair of valves from the one or more pairs of valves;c) drawing the first aliquot from the one or more mixing chambers comprises drawing the first aliquot from the one or more mixing chambers using the first pair of valves; d) flowing the first aliquot through the flowcell comprises the first pair of valves providing fluid communication between the sheath fluid source and the flowcell; e) drawing the second aliquot from the one or more mixing chambers comprises drawing the second aliquot from the one or more mixing chambers using the second pair of valves; andf) the set of acts comprises, after drawing the second aliquot from the one or more mixing chambers, flowing the second aliquot through the flowcell by performing acts comprising the second pair of valves providing fluid communication between the sheath fluid source and the flowcell.

29. The method of any of claims 16-28, wherein:a) the plurality of classes includes a first class which has a plurality of subclasses, wherein the plurality of subclasses includes the class for circulating tumor cells; b) the classification algorithm is configured to classify an input cell image as a circulating tumor cell image by performing a set of classification acts comprising: i) classifying the input cell image into the first class; and- 40 -0133788.0817287 4901-6446-8614v10ii) classifying the input cell image into the class for circulating tumor cells after classifying the input cell image into the first class.

30. The method of any of claims 16-29, wherein the classification algorithm is configured to classify an input cell image as a circulating tumor cell image based on a size of a cell depicted in the input cell image.- 41 -0133788.0817287 4901-6446-8614v10

Images